THE VIETNAM PROVINCIAL - The Asia Foundationasiafoundation.org/resources/pdfs/PCI08Fullreport2.pdfMr. Nguyen Ngoc Phi Chairman, People’s Committee of Vinh Phuc Mr. Nguyen Van Tu

THE VIETNAM PROVINCIALCOMPETITIVENESS INDEX 2008MEASURING ECONOMIC GOVERNANCEFOR PRIVATE SECTOR DEVELOPMENT

Primary Author and Lead Researcher:Dr. Edmund Malesky

Research Team:Tran Huu HuynhDau Anh TuanLe Thanh HaLe Thu HienNguyen Ngoc Lan

FOREWORD

The Provincial Competitiveness Index (PCI) is designed to assess and rank the performance, capacity andwillingness of provincial governments to develop business-friendly regulatory environments for private sectordevelopment.The fourth iteration, PCI 2008, once again validates that economic governance does matter. Ateach level of initial conditions, better-governed provinces are able to not only use their endowments moreefficiently but also influence business performance and income in subsequent years.

Due to its widespread acceptance, there have been many practical applications of the PCI reports andanalyses. Provincial governments find it a reflection of their respective provinces' strengths and weaknesses, apressure for reforms and a source of best practices for replication. In the past year, over 40 provinces andstate agencies have used the PCI to engage in public-private dialogue and diagnostic analysis with the localcommunities to better understand the competitive factors that drive economic growth. Investors use theindex as reference for their investment decision-making and advocacy for local initiatives.

This PCI 2008 report also introduces a new provincial analysis of infrastructure which is a critical factoraffecting access to markets, transaction costs, and economic growth and the overall competitiveness of theinvestment environment in Vietnam. Its policy implications offer for government planners, provincial andbusiness leaders important insights and recommendations on how to tackle the major obstacles created byout-dated and inadequate infrastructure that affects all enterprises.

The next phase of the PCI initiative should focus on using this valuable tool to stimulate dialogue and focuspriority actions on national and provincial strategies to create higher value-added industries and clusters ofgoods and services in different parts of the country based on their natural comparative advantages.

We hope this publication will serve as a useful resource for business associations, government and businessleaders, academic and research institutes, donor agencies, and the media as they seek to improve the qualityof economic governance throughout Vietnam's provinces.Vietnam Competitiveness Initiative and VietnamChamber of Commerce & Industry are ready to be your partners to improve the socio-economicdevelopment of Vietnam.

JJaammeess PPaacckkaarrdd WWiinnkklleerr,, PPhh..DD

DirectorVietnam Competitiveness Initiative

VVuu TTiieenn LLoocc,, PPhh..DD

ChairmanVietnam Chamber of Commerce & Industry

iTHE VIETNAM PROVINCIAL COMPETITIVENESS INDEX 2008

iiiTHE VIETNAM PROVINCIAL COMPETITIVENESS INDEX 2008

ACKNOWLEDGEMENTS

The Provincial Competitiveness Index 2008 is the result of a major, ongoing collaborative effort between theVietnam Chamber of Commerce and Industry (VCCI) and the U.S. Agency for International Development(USAID)-funded Vietnam Competitiveness Initiative (VNCI), managed by DAI, with a substantial contributionby VNCI partner The Asia Foundation (TAF).

Edmund Malesky of the University of California - San Diego led the development of the PCI's researchmethodology and authored the presentation of its analytical findings.

Professor Malesky was supported by a PCI research team that included Tran Huu Huynh, Deputy SecretaryGeneral and Director of the Legal Department at VCCI; Dau Anh Tuan, Le Thanh Ha and Nguyen Le Ha ofVCCI; Le Thu Hien, Nguyen Ngoc Lan, and Trinh Thi Hang of VNCI; Nina Merchant and Nguyen Thu Hangof TAF; and Paul Schuler of the University of California - San Diego. David Brunell, Jr., and Anne Le providedvaluable inputs to the PCI construction and analysis.

Do Hoang Anh,VNCI Deputy Project Director provided comments and Steve Parker of DAI served as thereport's expert reviewer.

The PCI was developed under the overall leadership of Vu Tien Loc, Chairman of VCCI, James Packard Winkler,VNCI Project Director, and benefited from the assistance and input provided by David Brunell, Director ofEconomic Growth for USAID, and Kim Ngoc Bao Ninh,Vietnam Representative of The Asia Foundation.

Suggested citation when referencing report or data: Malesky, Edmund. 2008. "The Vietnam ProvincialCompetitiveness Index: Measuring Economic Governance for Private Sector Development. 2008 Final Report."Vietnam Competitiveness Initiative Policy Paper #13. Vietnam Chamber of Commerce and Industry (VCCI) andUnited States Agency for International Development's Vietnam Competitiveness Inititative (VNCI): Ha Noi,Vietnam.

iv THE VIETNAM PROVINCIAL COMPETITIVENESS INDEX 2008

The PCI research team extends special thanks to our expert advisory panel, who provided valuable insightson methodology and indicator selection.

Mme. Pham Chi Lan Member, Institute of Development Studies

Mr.Vu Quoc Tuan President, Association of Vietnamese Trade Villages

Mr. Nguyen Dinh Cung Director of Macroeconomics, Central Institute for Economic Management

Mr. Nguyen Minh Man Director of State Administration and Civil Service Department, Office of theGovernment

Mr. Le Dang Doanh Member, Institute of Development Studies

Mr.Vu Duy Thai President, Ha Noi Union Association of Industry and Commerce

Mr. Cao Sy Kiem President, Association of Vietnamese Small and Medium Sized Enterprises

Mr. Nguyen Van Thoi National Assembly Member and Director of the Thai Nguyen Joint-StockExport Company

Mr.Vu Thanh Tu Anh Director of Research, Harvard Fulbright Economics Teaching Program, Ho ChiMinh City

Mr.Vo Hung Dung Director,VCCI Can Tho

Mr. Nguyen Ngoc Phi Chairman, People’s Committee of Vinh Phuc

Mr. Nguyen Van Tu Vice Director of Planning and Investment, Ha Tay (Ha Noi)

Mr. Pham Phuong Bac Vice Director of Planning and Investment, Bac Ninh

Mr. Nguyen Thai Son Senior Expert of the Government and Member of the Committee forImplementation of the Enterprise Law and Investment Law

Mr. Cao Ba Khoat Economist, Director of the K & Associates Consulting

Mr. Phan Duc Hieu Economist, Central Institute for Economic Management

Mr.Vu Quoc Huy Economist, National Economics University and Support for Trade AccelerationVietnam (STAR) Project

Mme. Pham Thi Thu Hang Director, Center for Supporting Small and Medium Sized Enterprises,VCCI

Mr. Pham Duy Nghia Legal Department, National University of Ha Noi

Mr. Le Duy Binh Economist, German Technical Cooperation

Mr. Nguyen Van Lan Economist, Mekong Private Sector Development Facility

Mr.Tran Du Lich National Assembly Member and Director of Ho Chi Minh City Institute ofEconomics.

Mr.Vu Xuan Tien Director, VFAM Consulting

Mr.Vu Quang Thinh General Director, MCG Consulting

vTHE VIETNAM PROVINCIAL COMPETITIVENESS INDEX 2008

TABLE OF CONTENTS

FORE WORD...... ...................................................................................................................i

ACKNOWLEDGEMENTS.................................................................................................iii

TABLES AND FIGURES ....................................................................................................vii

ABBREVIATIONS AND ACRONYMS............................................................................ix

EXECUTIVE SUMMARY..................................................................................................xiiiOverall Rankings and Scores ....................................................................................................xiiiGovernance Improvements .....................................................................................................xiiiReform challenges...........................................................................................................................xivEconomic Impact of Improving Governance ..................................................................xivInfrastructure Index .......................................................................................................................xivInformation and Communication Technology Index....................................................xv

CHAPTER ONE BACKGROUND AND KEY FINDINGS ........................................................................3

1.1 A Picture of the PCI Respondents...............................................................................31.2 The Final 2008 Provincial Competitiveness Ranking..........................................6

Robustness of the Rankings ...........................................................................................121.3 Changes over Time .............................................................................................................15

Positive Signs...........................................................................................................................15Worries......................................................................................................................................15

1.4 Impact of Governance Improvements on Private Sector Growth and Welfare .........................................................................16

CHAPTER TWO INDIVIDUAL POLICY EVALUATION USING THE PCI DATASET ...............................23

2.1 One-Stop Shop Business Registration.....................................................................232.2 Publication of Provincial Legal Gazettes .................................................................272.3 Legal Cases ............................................................................................................................30

CHAPTER THREE INFLUENCE OF PCI AND POLICY IMPLICATIONS .............................................35

3.1 Catalyst for reform.................................................................................................................353.2 Uses of the PCI..........................................................................................................................................36

3.3 Influence of the PCI ....................................................................................................37

3.4 The Way forward ..........................................................................................................37

vi THE VIETNAM PROVINCIAL COMPETITIVENESS INDEX 2008

CHAPTER FOUR ANALYSIS OF CHANGING EXPECTATIONS AND THE DECLINE IN PCI SCORES ....................................................................................41

CHAPTER FIVE INFRASTRUCTURE AND INFORMATION AND COMMUNICATIONTECHNOLOGY ...............................................................................................................53

5.1 Core Elements of the PCI Infrastructure Index.................................................545.2 Industrial Zone Quality.....................................................................................................585.3 Road Quality and Transport Costs............................................................................595.4 Utilities (Telecommunications and Energy)...........................................................605.5 Major Infrastructure............................................................................................................615.6 The PCI Infrastructure Index and Economic Governance ..........................625.7 Infrastructure as a Constraint on Fast Growing Regions .............................635.8 Information and Communication Technology .....................................................655.9 Conclusions .............................................................................................................................71

CHAPTER SIXDETAILED DESCRIPTION OF INDIVIDUAL SUB-INDICES.................................75

6.1 Entry Costs..............................................................................................................................756.2 Land Access and Security of Tenure .........................................................................79

6.2.1 Land Access...............................................................................................................796.2.2 Security of Land Tenure......................................................................................80

6.3 Transparency...........................................................................................................................836.3.1 Access...........................................................................................................................866.3.2 Equity and Consistency of Application......................................................876.3.3 Predictability ..............................................................................................................886.3.4 Openness....................................................................................................................88

6.4 Time Costs of Regulatory Compliance ..................................................................906.5 Informal Charges ..................................................................................................................936.6 State-Owned Enterprise Bias and Competition Environment ................96

6.6.1 Perceptions Indicators .........................................................................................966.6.2 Hard Indicators........................................................................................................96

6.7 Proactivity of Provincial Leadership........................................................................1016.8 Private Sector Development Policies ....................................................................1046.9 Labor Training .....................................................................................................................1076.10 Legal Institutions ................................................................................................................110

DATASETS USED IN ANALYSIS ..................................................................................115

APPENDICES ...................................................................................................................119

viiTHE VIETNAM PROVINCIAL COMPETITIVENESS INDEX 2008

TABLES

Table 1.1: Who Answers the PCI Survey? ..................................................................................4

Table 1.2: Sub-Index Weightings........................................................................................................9

Table 1.3: Provinces with Greatest Improvement in Weighted PCI Rankings .....14

Table 2.1: History of Private Business Entry in Vietnam...................................................25

Table 2.2: Access to Business Documentation ......................................................................28

Table 4.1: Comparison of Overall Scores and Sub-Indices (2006-2008) ..............42

Table 4.2: Assessment of Provincial Services and Infrastructure.................................46

Table 4.3: Firms' Evaluations of Infrastructure and Public Service,

By Month of PCI Response and Type of Firm Ownership ......................47

Table 5.1: Indicators Used in Infrastructure Index...............................................................56

Table 5.2: Infrastructure Growth Versus Demand for Usage ........................................64

Table 5.3: Indicators used in ICT Index......................................................................................66

Table 5.4: Bivariate Correlations of Final Indices and Outcome Variables.............71

Table 6.1: Comparison of Entry Costs Sub-Index ...............................................................77

Table 6.2: Comparison of Land Access and Tenure Security Sub-Index

(2005-2008).........................................................................................................................82

Table 6.3: Comparison of Transparency Sub-Index (2005-2008) ...............................83

Table 6.4: Factor Analysis of Planning and Legal Documents........................................86

Table 6.5: Scoring System for Provincial Websites...............................................................89

Table 6.6: Comparison of Time Costs of Regulatory Compliance

(2005-2008).........................................................................................................................91

Table 6.7: Comparison of Informal Charges (2005-2008)..............................................94

Table 6.8: Comparison of SOE Bias and Competition Environment

(2005-2008).........................................................................................................................98

Table 6.9: Comparison of Proactivity (2005-2008)..........................................................102

Table 6.10: Comparison of Private Sector Development Policies (2005-2008)105

Table 6.11: Comparison of Labor Policies (2006-2008) ..................................................108

Table 6.12: Comparison of Legal Institutions (2006-2008) ............................................111

TABLES AND FIGURES

viii THE VIETNAM PROVINCIAL COMPETITIVENESS INDEX 2008

FIGURES

Figure 1.1: Weighted Provincial Competitiveness Index 2008............................................7Figure 1.2: PCI 2008 Map of Vietnam ..............................................................................................8Figure 1.3: Province Performance by Sub-Index .......................................................................11Figure 1.4: Stability of PCI Ranking Over Time..........................................................................12Figure 1.5: Improvement in PCI Ranking Between 2006 and 2008..............................13Figure 1.6: Governance Premium ....................................................................................................17Figure 2.1: Access to Business Documentation, By Online Availability........................30Figure 2.2: Changes in Usage of Court System Over Time...............................................31Figure 2.3: Increasing Court Usage and Confidence in the System..............................32Figure 4.1: Average Firm Evaluation of Education versus Actual Educational Quality..49Figure 5.1: PCI Infrastructure Index..................................................................................................55Figure 5.2: Industrial Zone Quality in Province .........................................................................58Figure 5.3: Percentage of Road Covered by Asphalt, By Responsible Authority .........59Figure 5.4: Relationship Between Infrastructure and Governance.................................62Figure 5.5: Relationship Between ICT Infrastructure and Email Usage in PCI ........68Figure 5.6: Relationship between ICT Application and the PCI Webpage Score .......69Figure 5.7: Final ICT Readiness Index..............................................................................................70Figure 5.8: Total Investment Environment.....................................................................................72Figure 6.1: Entry Costs Sub-Index.....................................................................................................78Figure 6.2: Land Access and Security of Tenure Sub-Index ................................................81Figure 6.3: Transparency Sub-Index..................................................................................................85Figure 6.4: Access to Planning and Legal Documents by Province..............................87Figure 6.5: Time Costs of Regulatory Compliance Sub-Index .........................................92Figure 6.6: Informal Charges Sub-Index.........................................................................................95Figure 6.7: Bias Toward State-Owned Enterprises Sub-Index........................................100Figure 6.8: Proactivity Sub-Index .....................................................................................................103Figure 6.9: Private Sector Development Policies Sub-Index...........................................105Figure 6.10: Labor Policies Sub-Index..............................................................................................109Figure 6.11: Legal Policy and Institutions Sub-Index ...............................................................112

APPENDICESAppendix 1A: Multiple Regression Analysis of Governance and

Key Private Sector Outcome Variables ...............................................................119Appendix 1B: Multiple Regression Analysis of Governance and

Provincial Gross Domestic Product ......................................................................120Appendix 1C: Panel Analysis of Governance and Key Private Sector Outcome

Variables (2006–2008)..................................................................................................121Appendix 1D: Panel Analysis of GDP Growth (2005–2007).................................................122Appendix 1E: Predicted Effects of One-Point Change in

Sub-Indices on Dependent Variables ....................................................................123Appendix 1F: Predicted Effects of One-Point Change in

Sub-Indices on Dependent Variables ....................................................................124Appendix 2A: Impact of One-Stop Shop (OSS) Implementation on Business

Registration Waiting Periods......................................................................................125Appendix 2B: Determinants of Change in Economic Court Usage .................................127Appendix 3: Determinants of Transport Costs (VND) .........................................................128

ixTHE VIETNAM PROVINCIAL COMPETITIVENESS INDEX 2008

ABBREVIATIONS AND ACRONYMS

ASMED Agency for SME DevelopmentBRVT Ba Ria-Vung Tau provinceDPI Department of Planning and InvestmentGDP Gross Domestic ProductGSO General Statistics OfficeHCMC Ho Chi Minh CityICT Information and communication technologyIZ Industrial zoneLND Legal normative documentLURC Land Use Rights CertificateMPI Ministry of Planning and InvestmentOSS One-stop shopPAR Public Administration ReformPCI Provincial Competitiveness Index PSD Private sector developmentRIA Regulatory impact assessmentSME Small and medium-sized enterpriseSOE State-owned enterpriseSTAR Support for Trade Acceleration projectTAF The Asia FoundationVAT Value-added taxVCCI Vietnam Chamber of Commerce and IndustryVNCI Vietnam Competitiveness InitiativeWTO World Trade Organization

EXECUTIVE SUMMARY

The Provincial Competitiveness Index (PCI) wasdeveloped in 2005 by the Vietnam Chamber ofCommerce and Industry and the U.S. Agency forInternational Development-funded VietnamCompetitiveness Initiative. Since that time, the PCI iswidely viewed as a critical tool for measuring andassessing the standards of economic governance inVietnam's 64 provinces from the perspective ofprivate sector businesses 1. This year marks thefourth iteration of the PCI analysis.

Overall Rankings and ScoresThe 2008 version of the PCI brought somesurprising findings. First, Binh Duong, the defendingPCI champion for the past three years, lost itscrown to Da Nang, which had ranked second in allprevious iterations. Both provinces still remain in theExcellent tier and their final scores (72.18 and 71.76)are statistically indistinguishable.

Second, despite the strong stability in PCI rankingsover time, 2008 scores are generally lower acrossevery level of the PCI rankings.The medianprovince received a score about 2.4 points lowerthan in 2007, dropping from 55.6 to 53.2. Medianscores remain higher than in 2006 (52.41), but theimpressive governance improvements experiencedbetween 2006 and 2007 have reversed somewhat.

While most PCI sub-indices experienced moderateincreases, two sub-indices, Labor and Private SectorDevelopment, show dramatic declines. Because ofthe high weights of these indices, the declinesaffected the overall PCI scores significantly.Weexplore the causes of the lower scores in ChapterFour, finding that they are predominantly due to

either : 1) evidence of actual deterioration in publicservice delivery, or 2) increasing firm expectationsthat have not been met by proportionateimprovements in the quality of governmentservices. Negativity, due to Vietnam'smacroeconomic instability, also had a measurableinfluence on firms' perceptions.

Governance ImprovementsIn terms of individual indicators, there are a numberof reasons for celebration:

! Waiting periods for business registration andother formalities of business entry are athistoric lows. One-stop shop (OSS)implementation has had a positive impact onentry costs. In the median province,registration now takes only 12.5 days andfewer than 6 percent of respondents waitedmore than three months to finish all businessentry procedures.These regulatory changescontinue to be reflected in greenfield businessstart-up and formalization of householdenterprises. Chapter Two takes a closeempirical look at OSS, finding that the greatestdecline in waiting periods occurred inprovinces with more thorough implementationof the official procedures.

! Property rights are at historic highs.This year,81 percent of PCI respondents in the medianprovince have formal Land Use RightsCertificates, up from 75 percent in 2007 and55 percent in 2006.This startling growth willyield positive dividends in future investmentand business expansion as firms feel moresecure about their business prospects.

! As a result of the proliferation of provinciallegal gazettes (Công báo), access to businessdocumentation also continues to improve.Over 65 percent of firms believe they canaccess legal documents in their province, up

xiiiTHE VIETNAM PROVINCIAL COMPETITIVENESS INDEX 2008

EXECUTIVE SUMMARY

1. Ha Tay has now been merged with Ha Noi, but was a separate entityat the time of the survey.We assess it as a separate unit becausefirms would have been responding to questions with Ha Tay and notHa Noi authorities in mind. A similar approach is used for thedistricts of Vinh Phuc and Hoa Binh provinces that have now beenmerged partially with Ha Noi.

from 61 percent in 2007. Increasing access ishighest for People's Committee Resolutions,which we trace directly to the publication oflegal gazettes.We find that access to legaldocumentation is especially high in provincesthat publish their gazettes online. Furthermore,the percentage of firms negotiating their taxpayments with local officials-another importantindicator of transparency-improved in 2008.Only 36 percent of firms admit to this activityin 2008, down from 41 percent last year and61 percent in 2006.

! Usage of Provincial Economic Courts toresolve business disputes has nearly doubled inthe past year.The number of total cases filedmore than doubled between 2006 and 2007,from 2,445 to 5,198, while the number ofcases filed by private entrepreneurs grew by130 percent in 2007 compared with 2006.Higher usage of courts by private actors isconsidered a positive development because itindicates that entrepreneurs have greater faithin legal institutions and are more willing tomove beyond social relations as a mechanismfor enforcing contracts 2. It also may reflectimprovement in contract and othercommercial laws in Vietnam over the past sixyears, due in part to the Bilateral TradeAgreement (BTA) with the US.These changeshave made the law clearer in many regardsand, thus, given firms greater confidence intheir rights and responsibilities.

! Bias toward local state-owned enterprises isno longer a worry. Equitization of theseenterprises, expansion of bank lending toprivate firms, and increasing growth of theprivate sector have diminished firms' worriesabout special favors being granted to localchampions. Improvement on this sub-index isso impressive that we are consideringdropping it in future PCI iterations, althoughfirms do still worry about bias in favor of largestate conglomerates.

Reform ChallengesOn the other hand, there is still cause for worry:

! A disconcerting deterioration in performanceis observable in the Time Costs of RegulatoryCompliance sub-index.The decline in themedian score from 6.2 to 5.8 is not nearly asdramatic as the precipitous falls in PrivateSector Development and Labor, nor is it astatistically significant change. Nevertheless,individual indicators within the indexdemonstrate some decline.The percentage oftime spent on bureaucratic procedurescontinues to increase annually. Currently, 23percent of firms spend more than 10 percentof their time on bureaucratic procedures, withfew firms detecting any improvements in theburden caused by administrative barriers.While entry costs have improved dramatically,it appears there is still much work to do toaddress the cumbersome post-registrationcosts of regulatory compliance faced by firms.Most policy efforts on post-registrationregulation have focused on inspections;consequently, these have been reduced tominimal levels and do not appear to be animportant obstacle faced by entrepreneurs.On the other hand, firms still complain ofcumbersome paperwork and otherprocedures in day-to-day interactions withprovincial bureaucrats.

! Indicators of informal charges show nodiscernable improvements. Frequency and sizeof bribes, and the obstacles posed by bribery,have remained the same for the past threeyears, despite the government's highly publicefforts to reign in such activity. One of thecornerstones of public administration reformwas to increase the salaries and improve thequality of civil servants so bribes wouldbecome a less important part of public servicedelivery.These ambitious goals do not appearto have been achieved at the local level. Anti-corruption efforts through administrativesimplification have the stated goal of reducingambiguity in regulations and therebyminimizing the discretion of civil servants that

xiv THE VIETNAM PROVINCIAL COMPETITIVENESS INDEX 2008

2. Bergling, Per, 1999 Legal Reform and Private Enterprise:TheVietnamese Experience, Umea Studies in Law No. 1, Department ofLaw, Umea University: 130

enables informal charges. The PCI can serve asan effective monitoring tool by governmentand private actors for analyzing the impact ofthese changes on informal charges and thetime cost of compliance.

In sum, tremendous improvements have beenrecorded, but more work remains.The continuingburden of cumbersome regulatory proceduresunnecessarily raises firms' costs and cuts into theirbottom lines, while the lack of real progress oninformal charges substantially raises the risks ofentrepreneurial activity.The continuing reformefforts of the Government are essential to removeadministrative burdens and red tape to reducethese costs and risks. Such notable efforts includethe Prime Minister's Master Plan on AdministrativeProcedures Simplification in all aspects of StateAdministration from 2007 to 2010 (Project 30), theNational Strategy on Combating and PreventingCorruption to 2020, and Civil Service Reformthrough the adoption of a new Law on PublicOfficials and Civil Servants.

Economic Impact of ImprovingGovernanceThis year's report on our analysis of the governancepremium varies from previous versions of the PCI.Past reports relying on comparisons amongprovinces within a single year have demonstratedthat provinces with above-average PCI scores makemore efficient use of initial conditions than poorlygoverned provinces. At every level of initialendowments, therefore, provinces with goodgovernance have higher levels of gross domesticproduct (GDP) per capita than their peers.Wetermed this gap in achievement the "governancepremium." This analysis was important, but it waslimited by its static nature. Time series data wasnecessary for a dynamic analysis of the impact ofpolicy changes over time.

The new analysis takes advantage of time seriesdata to show that improvements in PCI scoreswithin a single province are associated witheconomic growth over time and not just acrossgeographic space. All else being equal, a one point

improvement in the 100-point PCI score in agiven year, within a particular province, isassociated with a 6.9 percent increase in activeprivate companies, a 2.6 percent increase inaverage investment per capita, and a 1.6 percentincrease in GDP per capita in the subsequentyear. Governance improvements do not appear tobe associated with year-to-year changes in averagefirm profitability.

Transparency of business information, proactivity ofleadership, and labor policies have the mostconsistent and robust impact on private sectordevelopment. Simply enhancing access to legal andplanning documents through a one-pointimprovement on transparency is associated with a3.7 percent rise in the number of firms per 10,000citizens in the province.

Infrastructure IndexChapter Five of this year's report adds a newfeature to the PCI analysis-the Infrastructure Index.A series of recent, high-profile reports havedemonstrated that Vietnam's infrastructure isharming its competitive advantage internationally. Atthe same time, increased fiscal decentralization has,in theory, increased the opportunities for provincesto raise their own resources for infrastructureimprovements. Consequently, a number of our keystakeholders requested that we bring PCI data tobear on the subject. Longer time series of annualdata will be necessary for more rigorous analysis ofthe influence of changes in infrastructure.Nevertheless, the PCI Infrastructure Indexrepresents a baseline that can be used to track andanalyze future developments.

The Infrastructure Index ranks the relative levels ofinfrastructure across the country along fourdimensions: 1) industrial zone quality and capacity;2) transportation costs; 3) telecommunications andenergy costs and stability; and 4) majorinfrastructure (ports/airports).We find that DaNang, Binh Dinh, Ho Chi Minh City (HCMC), HaiPhong, Binh Duong, and Dong Nai have the bestinfrastructure in the country, while the ruralNorthwestern Uplands have the worst.

xvTHE VIETNAM PROVINCIAL COMPETITIVENESS INDEX 2008

It is immediately obvious that the absolute levels ofinfrastructure are associated with economicdevelopment, but it is not clear which way thecausal arrow points. Many of these provincesinherited stocks of infrastructure going back severalgenerations or more. At the same time, their abilityto attract early investment has bolstered theirrevenue, allowing them to spend more onsubsequent infrastructure improvements.

Despite their absolute advantage, we find that theinfrastructure in the high-performing provinces isnot keeping up with demands on its usage.Theprovinces in the North Southeast industrial hubhave the fastest population growth and the largestshare of private enterprises and economic output,yet they lag behind other regions in the percentageof road covered with asphalt and improvements inroad quality and they pay the highest averageenergy prices in the country. In essence,infrastructure does appear to be placing constraintson development in high-growth regions, while policymakers have used infrastructure spending as a toolfor redistribution.

It is critical that policy makers find the right balancebetween redistribution and growth-inducinginfrastructure investments. After all, slower nationalgrowth reduces the pool of monies available forredistribution. Fiscal decentralization and thecorresponding powers of finance it has brought toprovincial authorities may be a first step in allowinghigh-growth areas to raise the capital necessary tomeet their changing economic needs.The draftLaws on Public Debt and State Budget that areunder discussion in the current session of theNational Assembly should facilitate fiscaldecentralization but maintain centralized oversightto ward off irresponsible spending.

Information and CommunicationTechnology IndexIn addition to the traditional types of infrastructurehighlighted above, more and more firms in Vietnamrecognize the utility of information technology forfacilitating interactions with business partners andacquiring market and technology information. Infact, 27 percent of respondents recorded activeemail addresses on their PCI survey forms. In HaNoi and HCMC, over 60 percent of firms usedemail addresses for their contact information.Annual growth in information and communicationtechnology (ICT) penetration in all of Vietnam isestimated at 8 percent.

The ICT Index borrows from an excellent datasetassembled by Office of the Steering Committee forNational Information Technology Readiness of theMinistry of Information and Communications tobuild on five sub-indices: 1) hardware, or thephysical infrastructure for internet usage; 2) humanresources, including the training, skill sets, andpersonnel necessary for technology acquisition; 3)ICT applications, measuring the scope, scale, andefficiency of current use on existing ICT hardware;4) availability of private ICT providers as asubstitute for government provision; and 5) theconduciveness of the provincial regulatoryframework for ICT development, including whetherprovinces have local ICT committees anddevelopment policies.

HCMC, Ha Noi, Binh Duong,TT-Hue, Da Nang, BacNinh, and Vinh Phuc received the top overall scoreson the ICT Index. As with infrastructure, a numberof mountainous provinces in the Northwest ranknear the bottom.

xvi THE VIETNAM PROVINCIAL COMPETITIVENESS INDEX 2008

CHAPTER ONE

BACKGROUND ANDKEY FINDINGS

At its most basic level, the 2008 ProvincialCompetitiveness Index (PCI) is the collective voice of7,820 domestic private firms. Private entrepreneurs'opinions regarding economic governance in theirprovinces are collected in a large-scale survey.Theirresponses are adjusted with published data toaddress perception biases, and these answers areaggregated into provincial-level scores.The finaloutcome is a composite index ranking Vietnam's 64provinces according to their performance on 10dimensions of governance that are critical for privatesector development 3. As a result, the PCI providesthe most objective metric available for gauging theimpact of economic and administrative reforms atprovincial and national levels.

Four features of the PCI approach have enhancedits reception among Vietnamese policy makers andbusinesspersons since its initiation in 2005. First, bynormalizing the scores around best economicgovernance practices already found in Vietnam andnot against ideal and possibly unattainable standardsof good governance, the PCI encouragessubnational governments to improve theirperformance. For each indicator, a "star" or top-performing locality can be identified and any otherprovince can conceivably attain a perfect score of100 in future PCIs by emulating these Vietnamesebest practices.

Second, by separating out the economic growthgenerated by initial conditions (that is, thefundamental underlying factors that contribute toeconomic growth in a province but that are virtuallyimpossible to address in the short term, such aslocation, infrastructure, size of the domestic market,and human resources), the PCI identifies goodeconomic governance practices achieved at theprovincial level.

Third, by comparing economic governance practicesagainst actual economic performance, the PCI

estimates how important governance practices arein attracting investment and generating growth.Theresearch shows a strong association among: 1)business-friendly economic governance practices; 2)business operations in response to those practices;and 3) importantly, improvements in economicactivity such as output, investment, profits, andincome.This last connection is critical because itmakes clear that business-friendly policiesencourage firms to operate in ways that benefit notonly the executives and workers in each firm butalso society more broadly by creating jobs andraising incomes throughout the economy.

Fourth, the indicators that comprise the PCI areactionable in the sense that they are specific itemsthat provincial officials can target and monitorimprovement upon, and they are action-worthybecause they represent policies that private firmshave identified as critical to their business successand that previous research has shown to beinfluential for firm performance.

More detail is provided on specific indicators inChapter Six, but a province that performs well on all10 PCI sub-indices is one that has: 1) low entry costsfor business start-up; 2) uncomplicated access to landand security of business premises; 3) transparent andequitable legal and business information; 4) minimalinformal charges; 5) limited time wasted onbureaucratic procedures and inspections; 6) lack ofbias toward particular types of firms, especially state-owned enterprises (SOEs); 7) a proactive andcreative leadership; 8) special private sectordevelopment (PSD) initiatives to address marketfailures; 9) well-trained labor; and 10) fair andeffective legal procedures for dispute resolution.

1.1 A Picture of the PCI RespondentsThe claim that the PCI represents the collectivevoice of the private sector is somewhatpresumptuous.Who are these entrepreneurs whopurport to speak for the entire business

3THE VIETNAM PROVINCIAL COMPETITIVENESS INDEX 2008

3. See footnote 1.

BACKGROUND ANDKEY FINDINGS

4 THE VIETNAM PROVINCIAL COMPETITIVENESS INDEX 2008

community? After all, any bias in the selection ofoperations affects the value of the information thatcan be gleaned from the survey.

By delineating the PCI sample according to keyfactors of interest to readers,Table 1.1 shows thatfirms answering the PCI survey look much like thebusiness community as a whole, representing allsizes and flavors of entrepreneurial activity inVietnam.This, of course, is by design. Respondentsare randomly selected from a list of registeredprivate firms that is supplied by the National Tax

Authority and precautions are put in place toensure the representativeness of the sample4.

4. Results are based on a stratified random sample and mail-out surveyin each province, yielding a national response rate of 26 percent, upfrom 21 percent in 2007. Response rates are similar throughout thecountry, so non-response bias is likely systematic across jurisdictions.This year we followed up with non-responders in our survey, findingthat 21 percent of our mail-out went to firms that were no longer inexistence or had moved their operations, or where the Tax Authorityhad incorrect contact information.Taking these into account, our trueresponse rate is 30.2 percent.

Table 1.1:Who Answers the PCI Survey?(Composition of the 7820 Total Respondents)

Provincial Sample National Sample

LLeeggaall FFoorrmm

Sole Proprietorship

Limited Liability

Joint Stock

Joint Stock with Share Listed on Stock Exchange

Partnership/Other

SSeeccttoorr ww//MMaajjoorriittyy OOuuttppuutt

Manufacturing/Construction

Service/Commerce

Agriculture/Aquaculture/Natural Resources

Equal Output in Two Sectors

AAggee ooff FFiirrmm

Registered before Enterprise Law

Registered After Enteprise Law

PPCCII

41.1%

42.7%

15.4%

0.3%

0.2%

PPCCII

43.9%

50.6%

1.2%

4.2%

PPCCII

12.2%

87.8%

MMeeddiiaann TTaaxx AAuutthhoorriittyy

43.8%

44.3%

10.4%

NA

1.5%


44.1%

53.3%

2.6%

NA


7.2%

92.7%

WWeeiigghhtteedd PPCCII

26.3%

50.1%

21.6%

1.8%

0.0%


43.6%

55.4%

0.4%

0.6%


15.9%

84.1%

TToottaall TTaaxx AAuutthhoorriittyy

25.9%

56.8%

17.3%

NA

NA


42.6%

52.8%

3.6%

NA


10.40%

89.60%


Provincial Sample National Sample

SSiizzee ooff OOppeerraattiioonnss ((TToottaall AAsssseettss,, BBiilllliioonn VVNNDD))

Under 0.5

From 0.5 to under 1

From 1 to under 5

From 5 to under 10

From 10 to under 50

Over 50

HHiissttoorryy ooff CCoommppaannyy

Greenfield Private Company

Began Operation as Household Enterprise

Former Local State Owned Enterprise

Former Central State Owned Enterprise

PPrriimmaarryy CCuussttoommeerrss

Vietnamese Indivduals and Companies

State Owned Companies

Export Directly or Indirectly

Foreign Individuals or Companies in Vietnam

PPCCII

14.5%

17.8%

42.7%

13.3%

9.3%

2.4%

PPCCII

36.5%

57.7%

4.8%

1.0%

PPCCII

64.7%

25.2%

5.2%

3.5%

MMeeddiiaann GGSSOO

16.7%

18.7%

44.7%

8.4%

8.4%

3.1%


8.6%

16.5%

44.3%

18.1%

8.2%

4.2%


34.8%

54.9%

6.4%

3.9%


58.2%

21.0%

11.4%

9.4%

GGSSOO CCeennssuuss

13.5%

18.5%

54.0%

9.9%

7.5%

2.0%

PCI is the PCI survey sample, stratified at the provincial level.Weighted PCI is the PCI survey sample, but weighted by provincial share of enterprises to create a nationally representative sample.Median Tax Authority provides the values in the median province.Total Tax Authority shows the national level aggregate scores.GSO Census is the 2007 Enterprise Census of the General Statistical Office.(http://www.gso.gov.vn/default_en.aspx?tabid=479&idmid=4&ItemID=7184)


Forty-one percent of respondent firms are soleproprietorships, 43 percent are limited liabilitycompanies, and 16 percent of firms are registered asjoint-stock companies. By design, these proportionsreflect the average provincial patterns; the PCI uses astratified random sampling strategy at the provinciallevel. According to the Tax Authority, the medianprovince possesses 44 percent sole proprietorships,44 percent limited liability companies, 10 percentjoint-stock companies, and 2 percent other.Thus, oursample is perfectly adjusted to each province, butvaries slightly from national scores as a whole, wherelimited liability companies are better represented,because Ha Noi and Ho Chi Minh City (HCMC)represent a smaller share of the PCI sample thanthey do national-level private sector activity.Table 1.1also presents a national PCI sample (re-weighted bythe proportion of total firms in each province) andaggregate data from the Tax Authority to facilitatenational-level comparisons.

Nineteen of the joint-stock companies in the sampleare traded on either the Ha Noi (HaSTC) or Saigon(HOSE) stock exchanges, accounting for about 7percent of all listed firms and including some of themost important operations in the country.

Half of the respondents are engaged in the serviceor commerce sectors, and about 44 percent areinvolved in manufacturing (22 percent) orconstruction (22 percent). By way of comparison,according to the Tax Authority data, 53 percent ofregistered firms are involved in service andcommerce and 43 percent in manufacturing andconstruction nationally. Eighty-eight percent of PCIrespondents registered after the 2000 EnterpriseLaw; 12 percent precede that law and, therefore,registered under older, less efficient procedures.

Forty-three percent of firms have between VND 1billion (US$56,000) and VND 5 billion ($282,500) intotal assets.Thirty-two percent of firms have lessthan VND 1 billion in assets and 25 percent haveover VND 5 billion.Two percent of firms have overVND 50 billion ($2.8 million) in assets.This is not adramatic share of large firms, but it does mirrornational-level data. Over 40 percent of firms havebetween 10 and 50 employees. Six percent ofrespondents have more than 200 employees and49 companies have more than 1,000.

More than half of PCI respondents beganoperations as household firms before they decidedto formalize their activities by registering atprovincial Departments of Planning and Investment.This is a critical finding because it demonstrates thatVietnamese institutions are conducive toformalization 5. A total of 454 enterprises resultedfrom equitizations (the Vietnamese form ofprivatization) of local or central SOEs. An additional320 companies have owners who formerlymanaged SOEs, and more than 1,000 have ownerswho worked in some capacity for an SOE. Finally,about 37 percent of firms are greenfield entities,meaning that their owners established andregistered the firms at roughly the same time.

Most companies concentrate their business activityon the domestic market, selling either toVietnamese individuals and private companies (65percent) or to SOEs (25 percent). About 6 percentare actively engaged in exporting, either directly orindirectly through trading companies.

Finally, and quite importantly for the purposes ofthe PCI, all of Vietnam's 64 provinces arerepresented in the Index.The average number ofresponses per province was 122, with only oneprovince (Lai Chau) receiving fewer than 75.Nevertheless, the 52 firms that did answer in LaiChau account for about one-third of all operationsin the rural, Northwestern jurisdiction.

1.2 The Final 2008 ProvincialCompetitiveness RankingThe weighted 2008 PCI ranking is shown in Figure1.1.Three differences between this year and lastyear are immediately apparent. First, for the firsttime in four years, we have a new top-performingprovince-Da Nang (72.18) moved slightly past BinhDuong (71.76) province.The margin between thetwo top provinces, however, is so small that it isstatistically negligible. Both remain within theExcellent performance tier.

5. For a more detailed discussion of the formalization in Vietnam, seeMalesky, Edmund, and Markus Taussig. 2008a. "Out of the Gray:TheImpact of Institutions on Business Formalization." Presented at theAnnual Meeting of the American Political Science Association, Boston,Massachusetts, August 28.


Figure 1.1:Weighted Provincial Competitiveness Index 2008


Figure 1.2: PCI 2008 Map of Vietnam

The final 2008 PCI represents the weighted sum ofthe scores of the 10 sub-indices, based on theweights shown in Table 1.2. In 2006, weights werecalculated using a three-step statistical procedure 6.The ultimate goal of weighting is to ensure that PCIscores are calibrated to private sector performanceand, therefore, that the PCI relates the mostrelevant information to provincial officials regardingthe impact of their policies on private sector activity.Sub-indices that were shown to have the largestassociation with private sector growth, investment,and profitability received the highest weight class of15 percent. Correspondingly, those that are notstrongly correlated with private sector developmentoutcomes received the lowest weight class of 5percent. Medium weights of 10 percent werereserved for average correlations across the threeoutcome variables or a large substantive effect onone outcome (e.g., profitability), but a minimalrelationship with the other two.The 2008 PCIemploys the 2006 weights to allow for longitudinalcomparison. In future versions of the PCI, however,weights will be re-calibrated to capture the dynamicnature of the Vietnamese economic reform agenda.

The second difference between this year's PCI andlast year's is that the median weighted PCI score(53.2) is slightly lower than the 2007 median (55.6),yet it remains above the 52.4 median recorded in2006. Significant drops in scores were limited to twospecific areas: PSD and Labor Policy. Because these are

two of the four most heavily weighted sub-indices,their impact on the overall outcomes was quitesubstantial. Low scores on PSD and especially LaborPolicy are not anomalies; they are recorded amongmost provinces and represent what appears to be agrowing disappointment among the private sectorregarding the effectiveness of the contribution byboth local and national officials to economicdevelopment. In some cases, the disappointmentemanates from tangible declines in public servicedelivery. In other cases, negative firm perceptions aremore likely the result of increasing expectationsbrought about by a strong history of improvement inlegal and regulatory reform in Vietnam. Constantimprovement in government performance over thepast decade may have raised firm expectations to thepoint where no improvement or marginal declinesmay generate negative perceptions from respondents.We explore this issue more directly in Chapter Four.

As a result of declines in scoring, another overalldifference can be seen between the 2007 and 2008ratings-shifts in the tiers of provincial performance. Asmaller number of provinces are now ranked in theExcellent and High tiers in Figure 1.1. Once again, tofacilitate comparisons with previous years, we peggedthe six performance tiers (Excellent, High, Mid-High,Average, Mid-Low, and Low) to the break-points setin the 2006 analysis. For instance, this year both BinhDinh and Vinh Long dropped below the break-pointat 69, which differentiates Excellent (shaded inpurple) and High performers (shaded in dark blue),while Vinh Phuc rose above the threshold. At thebottom of the rankings, two new provinces slippedbelow 42 and into the realm of low-performers.


6. See Malesky, Edmund. 2008. A Peek under the Engine Hood:TheMethodology of The Asia Foundation's Subnational EconomicGovernance Indices. San Francisco:The Asia Foundation , p. 16.http://asiafoundation.org/publications/index.php?q=&searchType=country&country=3.

Sub-Index

PSD services

Transparency

Labor training

Proactivity

Time costs of regulatory compliance

Legal institutions

SOE bias (competition environment)

Informal charges

Land access and security

Entry costs

Table 1.2: Sub-Index Weightings

Rounded Weights

15%

15%

15%

15%

10%

10%

5%

5%

5%

5%

Weight Class

High

High

High

High

Medium

Medium

Low

Low

Low

Low

Weights are derived from regression analysis using three dependent variables (number of private firms, investment per capita, and profitper firm). See Appendix 2 for full regression results.


As in early reports, the PCI 2008 uses a range of indicators that are grouped together into 10composite sub-indices.These 10 sub-indices, illustrated in Figure 1.3, are:

1. Entry Costs: A measure of: i) the time it takes a firm to register and acquire land; ii) the timeto receive all the necessary licenses needed to start a business; iii) the number of licensesrequired to operate a business; and iv) the perceived degree of difficulty to obtain alllicenses/permits.

2. Land Access and Security of Tenure: A measure combining two dimensions of the landproblems confronting entrepreneurs: how easy it is to access land and the security of tenureonce land is acquired.

3. Transparency and Access to Information: A measure of whether firms have access to theproper planning and legal documents necessary to run their businesses, whether thosedocuments are equitably available, whether new policies and laws are communicated to firmsand predictably implemented, and the business utility of the provincial webpage.

4. Time Costs and Regulatory Compliance: A measure of how much time firms waste onbureaucratic compliance, as well as how often and for how long firms must shut theiroperations down for inspections by local regulatory agencies.

5. Informal Charges: A measure of how much firms pay in informal charges, how much of anobstacle those extra fees pose for their business operations, whether payment of thoseextra fees results in expected results or "services," and whether provincial officials usecompliance with local regulations to extract rents.

6. SOE Bias and Competition Environment : A measure focusing on the perceived bias ofprovincial governments toward state-owned enterprises, equitized firms, and other provincialchampions in terms of incentives, policy, and access to capital.

7. Proactivity of Provincial Leadership: A measure of the creativity and cleverness of provincesin implementing central policy, designing their own initiatives for private sector development,and working within sometimes unclear national regulatory frameworks to assist and interpretin favor of local private firms.

8. Private Sector Development Services: A measure of provincial services for private sectortrade promotion, provision of regulatory information to firms, business partner matchmaking,provision of industrial zones or industrial clusters, and technological services for firms.

9. Labor and Training: A measure of the efforts by provincial authorities to promote vocationaltraining and skills development for local industries and to assist in the placement of locallabor.

10. Legal Institutions: A measure of the private sector's confidence in provincial legal institutions;whether firms regard provincial legal institutions as an effective vehicle for dispute resolution,or as an avenue for lodging appeals against corrupt official behavior.

Box :The 10 Sub-Indices of the PCI


Figure 1.3: Province Performance by Sub-index


Robustness of the RankingsThe decline in the rankings cited above is systematicacross the country. No region or particular group ofprovinces was impacted disproportionately. As aresult, the 2008 rankings look remarkably similar tothose from previous years. Da Nang, Binh Duong,Vinh Phuc, Lao Cai, Dong Thap, and An Giang remainamong the very best performers in the country.Furthermore, the bivariate correlation between 2008and 2007 final scores is 0.9 (see Figure 1.4). In PanelB, we see that 2006 PCI scores are also stronglycorrelated with 2007 scores.Together, the plots showconvincingly that historical governance is an excellentpredictor of future governance.This makes sense;governance practices cannot change overnight. Ittakes time to plan and implement new initiatives, andthere is even a longer lag between implementationand firms experiencing their effects.

This consistency in the provincial rankings hasimportant methodological and policy implications.Policy-wise, the stability of the rankings, despitethe decline in scores, indicates that the negativitythat the PCI is picking up this year results fromnational-level factors that are being experiencedacross the country. It is quite likely that thedecline in firm perceptions is somehow connectedto the macroeconomic instability and subsequentinsecurity experienced by private actors at thebeginning of 2008. It may be that provincialleaders have insufficient levers to address thedecline in perception to the degree that problemsare associated with national-level policies.Alternatively, it could simply be a case of risingexpectations of respondents not being met byimprovements in performance.

Methodologically, the strong correlation over timeindicates that the indexing approach works quitewell.The index can consistently identify top

Figure 1.4: Stability of PCI ranking over time


performers, while allowing room for otherprovinces to improve and receive higher scores insubsequent years. Such stability is confirmationthat the PCI approach offers a consistent andaccurate measure of provincial economicgovernance over time.

"Stability" is not synonymous with "fixed." Over thepast two years, a number of provinces havegradually climbed up the PCI ladder. Figure 1.5shows the cumulative changes experienced by allprovinces over the past two-years, while Table 1.3highlights the provinces that have recorded the

greatest two-year and single-year improvements.There are several interesting things to note in Table1.3. First, 4 of the top 10 improvements wereachieved by provinces in the Mekong Delta, includingthe 2 provinces with the greatest improvements-CaMau and Long An. Long An now ranks among thetop-performing provinces in the country. Second,whereas last year the key drivers of improvementwere the Transparency and Transactions Costs sub-indices, this year, improvements have been led by lessBias toward SOEs and greater Proactivity ofProvincial Leadership.

Figure 1.5: Improvement in PCI Ranking Between 2006 and 2008


Tabl

e 1.

3:Pr

ovin

ces w

ith G

reat

est I

mpr

ovem

ent i

n W

eigh

ted

PCI R

anki

ngs

Ca

Mau

Long

An

TT-H

ue

Ha

Nam

Binh

Phu

oc

Ben

Tre

Qua

ng N

gai

Binh

Thu

an

Tien

Gia

ng

Ha T

inh

Med

ian

Prov

ince

Ca

Mau

Long

An

TT-H

ue

Ha

Nam

Binh

Phu

oc

Ben

Tre

Qua

ng N

gai

Binh

Thu

an

Tien

Gia

ng

Ha T

inh

Med

ian

Prov

ince

Proa

ctiv

ity/L

and

Labo

r/SO

E

Lega

l

Proa

ctiv

ity

SOE

Bias

Proa

ctiv

ity/S

OE

Entr

y/In

form

al

SOE

Bias

Tran

spar

ency

SOE/

Entr

y

38 36 30 24 22 18 16 14 12 11 0

18 6 10 26 32 7 41 17 21 49 32

58.6

4

63.9

9

60.7

1

55.1

3

53.7

1

62.4

2

50.0

5

58.7

5

57.2

7

47.4

8

53.5

1

20 17 15 15 11 10 8 8 8 7 1

Ca

Mau

TT-H

ue

Tien

Gia

ng

Ha T

ay

Long

An

Than

h H

oa

Bac

Gia

ng

Soc T

rang

Tay

Nin

h

Qua

ng N

gai

Med

ian

Prov

ince

27 25 21 21 21 17 13 13 13 12 0

Two-

Year

Impr

ovem

ent

2006

-200

8O

ne-Y

ear

Impr

ovem

ent

2007

-20

08O

ne-Y

ear

Impr

ovem

ent

2006

-20

07

Prov

ince

Prov

ince

Prov

ince

2008

PC

IRa

nk20

08 P

CI

Scor

eSu

b-In

dice

s M

ost

Impo

rtan

t fo

rIm

prov

emen

t

Impr

ovem

ent

inRa

nkin

gIm

prov

emen

tin

Rank

ing

Impr

ovem

ent

inRa

nkin

g


Finally, and most importantly, more than half of thegreatest leaps were experienced in provinces thatmade legal commitments to improve their PCIscores through official resolutions of the PartySecretary and People's Committee or throughaction plans sanctioned by the top leadership (TT-Hue,Tien Giang, Ca Mau, Long An, Quang Ngai).These documents go beyond noncommittalstatements that call for general improvements tospecifically identify key provincial weaknesses ineconomic governance, assign responsibility toindividual actors, provide clear targets for measuringsuccess, and identify local initiatives to help achievethose results.In essence, the provinces that have shownimprovement are the ones where local leaders openlycommitted themselves to the task, formally announcingcommitments to their subordinates and, in many cases,to the public at large.While the PCI research team hascollected and assembled a library of such documents,these reports do not have a direct impact on scoring.Our scoring is based only on the opinions ofindependent firms.Thus, the rise in scores in theprovinces that committed overtly to change is theresult of entrepreneurs in those areas havingexperienced the new environment first-hand andhaving recorded their improved perceptions in theirsurvey responses. Rises in scores are a credit to thework and openness of local officials in these locations.

1.3 Changes over TimeAs discussed above, there is little evidence ofgeneral improvement in economic governance thisyear. Most of the sub-indices show smallimprovements, but none that are as dramatic as theleaps taken between 2006 and 2007. Nevertheless,trends on a few critical indicators are worthy ofnote. Detailed chronological tables and descriptionsof indicators are available in Chapter Six.Thissection provides only a few highlights.

Positive SignsAs a result of one-stop shop (OSS) implementation,which we discuss in more detail below, waitingperiods for business registration and procedures for

start-up have declined significantly this year. Securityof property rights also continues to improve.Thisyear, 81 percent of PCI respondents have formalLand Use Rights Certificates, up from 75 percent in2007 and 55 percent in 2006.This solid increase issupported by official data from the Ministry ofNatural Resources and Environment. Importantly, asa result of a proliferation of Provincial LegalGazettes (see Section 2.2), access to legaldocuments also continues to improve.Thepercentage of firms negotiating their tax paymentswith local officials-another important indicator oftransparency-improved in 2008 as it has in previousyears. Only 36 percent of firms admit to this activityin 2008, down from 41 percent last year and 61percent in 2006. Finally, SOE bias is at a historicallow: only 39 percent report explicit bias on behalfof SOEs. More than 50 percent of firms believe thattheir province has a positive attitude toward privateentrepreneurs, and the number of SOEs operatingin provinces has declined by 60 percent since 2000.In fact, SOE Bias is now so low that we will likelydrop it in future iterations of the PCI. Althoughfirms still worry about central conglomerates, theyare much less concerned about favoritism towardlocal SOEs.

WorriesA disconcerting deterioration in performance isobservable in the Time Costs of RegulatoryCompliance sub-index.The decline in the medianscore from 6.2 to 5.8 is not nearly as dramatic asthe precipitous falls in PSD and Labor, nor is it astatistically significant change. Nevertheless,individual indicators within the sub-indexdemonstrate clear declines.The percentage of timespent on bureaucratic procedures continues toincrease annually. Currently, 23 percent of firmsspend more than 10 percent of their time onbureaucratic procedures, with few firms detectingany improvements in the burden caused byadministrative barriers.While entry costs haveimproved dramatically, it appears there is still muchwork to do on addressing the cumbersome post-registration costs of regulatory compliance. Mostpolicy efforts on post-registration regulation have


focused on inspections. Consequently, these havenow been reduced to minimal levels and do notappear to be an important obstacle faced byentrepreneurs. On the other hand, firms stillcomplain of cumbersome paperwork and otherprocedures in day-to-day interactions withprovincial bureaucrats. Clear guidelines involving thedistribution of value-added tax (red) receipt booksand renewals of environment, health, and safetycertificates top the lists of firms' priorities for policymakers. As always in Vietnam, interpretation andimplementation of administrative procedures variesheavily by province.

The Prime Minister's Master Plan on AdministrativeProcedures Simplification (Project 30) implementedby the Office of the Government and supported byVNCI go a long way toward resolving these issuesby inventorying and reviewing all of theadministrative procedures affecting firms at both thenational and provincial levels in a transparent,accessible, and comprehensive database. First theinventory phase will allow for a quick and thoroughassessment of the current post-registrationregulatory burden faced by firms in each economicsector and province. Firms will have betterunderstanding of their rights and obligations.Theinventory will allow for standardization of allnationally issued administrative procedures andready comparison of provincial regulatory initiatives,so local leaders can follow national initiatives andidentify the standards that best suit their needs. Thesecond phase of Project 30 to review, simplify, orabolish unnecessary, illegal, and cumbersomeprocedures will reduce the huge burden ofpaperwork and compliance costs affecting firms intheir daily operations, easing the interpretation andimplementation of administrative proceduresnationwide.

Other indicators do not demonstrate significantyear-on-year changes.Though not declines, thesetrends are worrying because they have beenidentified as key policy initiatives by the leadershipand, therefore, stagnation is disquieting. Indicators ofInformal Charges, for instance, have shown nodiscernable improvements. Frequency of, size of, andobstacles posed by bribery have remained the same

for the past three years, despite the government'shighly public efforts to reign in such activity 7. Oneof the cornerstones of public administration reformwas to increase the salaries and improve the qualityof civil servants so that bribes would become a lessimportant part of public service delivery.Theseambitious goals do not appear to have beenachieved at the local level.

Currently, two new initiatives are underway that areplanned to help address the deficiencies of previouscorruption efforts.The National Strategy onCombating and Preventing Corruption by 2020 andits Action Plan, which is being drafted, is planning anintegrated strategy across a number of lineministries and local agencies. In addition, the newlypassed Law on Public Officials and Civil Servants isexpected to upgrade and create a clean, motivated,and effective civil service through merit-basedperformance incentives, including performanceevaluations, competitive recruitment for contractpositions, and clearly defined job descriptions. PCIdata will certainly help gauge the effectiveness ofthese efforts as well.

In sum, tremendous improvements have beenrecorded, but more work remains.The continuingburden of cumbersome regulatory proceduresunnecessarily raises firms' costs and cuts into theirbottom lines, while the lack of real progress oninformal charges substantially raises the risks ofentrepreneurial activity.

1.4 Impact of GovernanceImprovements on Private SectorGrowth and Welfare Each year, the PCI research team providesregression results that establish a strong associationbetween the unweighted PCI results and keyprivate sector performance outcomes.

We consistently show that provinces with bettergovernance make more productive use of initialendowments. Initial conditions certainly matter ;

7.Vasavakul,Thaveeporn. 2008. "Recrafting State Identity: Corruptionand Anti-Corruption in Doi Moi Vietnam from a ComparativePerspective." Presented at the Conference on Re-Making theVietnamese State, Hong Kong, August 21.


more well-endowed provinces have much higherincomes (measured by gross domestic product(GDP) per capita) than their less well-endowedpeers. However, at every level of initialendowments, better-governed provinces (i.e., thosewith PCI scores above the 2006 median) out-perform less well-governed localities.This is truewhether we use overall level of GDP per capita(see Figure 1.6, Panel A) or focus solely on theaverage change in GDP per capita over the past

two years (see Panel B).The gap between the solidblue line (the predicted values for high PCIprovinces) and the dashed red line (predictions forlow PCI provinces) is what we term the "economicgovernance premium." It indicates that at each levelof initial conditions, better-governed provinces areable to use their endowments more efficiently andachieve higher levels of economic welfare. (SeeAppendix 1 [especially Appendix 1B] for detailedregression results.)

Figure 1.6: Governance Premium(The Impact of Good Governance on Provincial Wealth and GDP Growth)

Note for display purpose, this graph drops national-level cities and Ba Ria - Vung Tau, because they are outliers that are difficult to display in asingle figure.

While these regressions control for a range offactors-notably infrastructure, human capital, andproximity to major markets-we were never whollysatisfied with them. Because we did not havesufficient data on economic governance over time,we were only able to establish that there was anassociation between good governance andoutcomes among provinces within a given year.Thisis the relationship demonstrated in Panel A.That is,we could show that private sector performance inwell-governed provinces was better than that inpoorer provinces. In this case, the governancepremium in the figure is the relative difference inGDP per capita, within a single year, between well-and poorly governed provinces. Although this cross-sectional (across geographic space) correlation isimportant, it is static and limited because it cannoteffectively track the influence of changes over time.

From a policy perspective, it is more useful to allowfor more dynamic analysis of whether governanceimprovements within a single province influencethat specific province's business performance andincome in future years. In other words, it would behelpful to show that over-time improvements onPCI indicators lead to improvements in economicoutcomes. After all, this is the calculation that aprovincial leader must make, asking: "Will thisinitiative that I am considering today improve theeconomy in my province tomorrow?"

Now that comparable data on both the PCI scoresand economic outcomes going back three years areavailable, time-series analysis can be performedwithin each province (sometimes referred to aslongitudinal or panel analysis). Here is what welearned. All else being equal, a one-pointimprovement in the 100-point PCI score in a givenyear within a particular province is associated with a6.9 percent increase in active private companies, a2.6 percent increase in average investment percapita, and a 1.6 percent increase in GDP per capitain the subsequent year.This is the relationshipillustrated in Figure 1.6, Panel B.These effects arestrongly statistically significant and the findings arerobust relative to changes in control variables anddifferent regression specifications.To put these

figures in context, an equivalent shift in oneindicator of infrastructure quality in a province (anadditional telephone for every 10 citizens) isassociated with a 4 percent growth in privatecompanies, a 1.1 percent growth in averageinvestment, and a 2.1 percent change in GDP percapita.These results show that while governance isnot the only driver of improvements in economicconditions in a province, it does matter a great deal.

There is one important qualification for this finding.There is no statistically significant relationshipbetween improvements in governance and changesin profitability. Although there is a strong cross-sectional association (well-governed provinces havehigher profits), there is no clear evidence that asingle-year improvement in PCI scores influencesfirms' bottom lines in subsequent years.Improvements in governance seem to be moreclearly associated with the willingness of firms totake risks through entry or expanded investments.The bottom line is that firms in better governedprovinces have higher profitability, but profit marginshave not expanded along with governanceimprovements. Other factors are more influential inexplaining changes in profitability over time.

As in previous reports, we studied the differentialassociation of individual sub-index scores with keyoutcome variables. Once again, transparency ofbusiness information, proactivity of the leadership,and labor policies have the most consistent androbust impact on private sector development.Simply enhancing access to legal and planningdocuments through a one-point improvement intransparency is associated with a 3.7 percent rise inthe number of firms per 10,000 citizens in theprovince. Some sub-indices that have been onlymarginally important in past years proved to havesignificant effects on individual-outcome variablesthis year. Key among these was the Entry Costs sub-index, which, because of the differential impact inOSS implementation across the country, wassignificantly related to new enterprises established ina province. Interestingly, lowering entry costs wasalso negatively associated with profits per firms.Thiscan be explained because lower barriers to entry



increase competition for existing entrepreneurs,cutting into their profit margins. On the other hand,Land Access and Security of Tenure had a largeimpact on firm profitability. A single-pointimprovement on this proxy for property rights iscorrelated with almost VND 9 million per firm inannual profits.

The formal details of the regression specificationsare in Appendices 1C, D, and F, but there are threetechnical details that general readers shouldunderstand.

First, we use the unweighted PCI score because theweighted PCI is calibrated to private sector outcomes.

Second, the panel analysis includes provincial-levelfixed effects; that is, we only compare each provinceto itself over time and not to other provinces in thecountry. As a result, we do not need to worry

about historical or cultural factors that are specificto a locality.The fixed effects absorb this variation,allowing us to isolate the impact ofpolicy/institutional change independent of aprovince's initial conditions.

Third, there are a number of factors that affectprovincial economies but are out of control of theprovincial authorities in the short run, includinginfrastructure, human capital, population growth, andnational economic trends such as macroeconomicconditions, foreign investment, and commodity priceincreases. Our model controls for these factors sothat we can calculate the marginal improvement ofchanges in governance. Clearly these factors areimportant, but the purpose of our analysis is toassess the singular impact of governance on privatebusiness activity, holding constant the other driversof economic performance.

CHAPTER TWO

INDIVIDUAL POLICYEVALUATION USING THE

PCI DATASET


As provincial officials often remind the researchteam, it makes little sense to focus on the impact ofimprovements on the aggregate sub-indices. Afterall, these are composite measures that includemultiple factors and, therefore, the directconnection to the levers of policy makers is unclear.What exactly does a one-point improvement onthe Entry Costs sub-index mean anyway, and if Iwant to lower entry costs in my province this year,what should I do? What really matters for policymakers and stakeholders are the individualindicators that comprise the PCI data because thesereflect actual policy choices.What they need toknow are the potential benefits of specific practicesthat they can weigh against the costs ofimplementing them.The new time-series dataprovide a powerful tool to investigate thesequestions. Specifically, we can use the PCI data toevaluate policy changes and generate directmeasurement of their impact.Three changes arerelevant for this year's analysis: 1) one-stop shop(OSS) business registration; 2) the publication ofProvincial Gazettes; and 3) the increasing use ofcourts as a means of dispute resolution. We findthat OSS has significantly increased the speed ofbusiness registration; that publication of ProvincialGazettes, particularly online versions, hasdramatically increased transparency of businessdocumentation; and that increasing usage ofprovincial courts appears to be associated withfirms' stated confidence in legal institutions.

2.1 One-Stop Shop BusinessRegistrationAlthough the idea has been around for some timein Vietnam, and a number of provinces have

experimented with pilot versions, one-stop shopsfor business registration were not officiallyrecognized as a fundamental obligation of localofficials until October 23, 2006, under the PrimeMinister's Decision No. 236/2006/QD-TTg Thedecision formally approved the Small and Medium-Sized Enterprise (SME) Development Plan (2006-2010), which mentioned OSSs, while a subsequentinterministerial circular (02/2007/BKH-BTC-BCA)assigned responsibility for implementing them toprovincial Departments of Planning and Investment(DPIs) and articulated the specific processes.Circular 02 was entirely voluntary. Provinces werenot forced to implement it; rather, the legislationwas meant to provide guidance for DPIs interestedin moving forward with OSSs.

According to Circular 02, OSSs were meant toinvolve three separate business start-up procedures:the business registration certificate issued by DPI;the tax code issued by the Tax Authority; and thechop (or red seal) for stamping documents that isgranted by the local police department. Accordingto the decision, private companies need only visit asingle office and an individual within that office ischarged with collecting all three pieces ofdocumentation on behalf of the company. Circular02 also identified a maximum waiting period of 15days for all three procedures.

There was a wide variance in the interpretation ofthese policies among localities. Some provincesgranted official approval to have the chop made, butstill required entrepreneurs to visit provincialDepartments of Public Security to register the chopafter it was manufactured. Other provinces requiredthat an official from the Department of PublicSecurity work directly in the OSS, so the chop

INDIVIDUAL POLICYEVALUATION USING THEPCI DATASET

registration could be granted immediately. Someprovinces required separate applications for eachdocument, while others found ways to consolidatethe applications. Eventually, in December 2007, thedifferent interpretations of the procedures for chopissuance led central officials to take action.The choplicense was abolished by the Ministry of PublicSecurity entirely. Now, a firm must only have itschop made at a local establishment and take it tothe police for registration, a formality that cannottake longer than two days.8 Because the choplicense no longer exists, a revised Circular 05 wasissued in July 2008 that dropped the chop licenserequirement from the OSS procedures. It furtherreduced the maximum waiting period forregistration and tax certificates to five total days.

Have these national-level policy changes reducedthe time that firms wait to become fully legalestablishments? Thus far, most evaluations of OSSshave relied on official data supplied by DPIs.Thisevidence has been confirmatory, but has also beenviewed with some suspicion because DPI officialshave strong career self-incentives for improving theirnumbers.They cannot be considered independentevaluators of their own efforts. More directly, officialnumbers tend to underestimate the true amount oftime required for business entry becauseregistration officials only record a registrationprocess when a complete form is received, but donot account for days lost when applications arereturned because of minor errors in thedocumentation.9

PCI's firm-level data can provide an independenttest of the impact of changes in policy on businessentry.Table 2.1 takes a national-level perspective,dividing the entire PCI sample by the year in whichfirms registered.Three broad groups of dates arevisible by the shading: 1) firms that registeredbetween the formal recognition of private firms in

1991 and the inauguration of the Enterprise Law atthe end of 1999; 2) firms that registered betweenthe Enterprise Law and the initiation of OSSs; and3) firms that registered after Circular 02 onimplementation of OSSs.We do not consider theUnified Enterprise Law of 2005 separately becauseit did not change registration procedures for privatefirms dramatically, but, instead, united foreign andprivate enterprises under the same legal regime.

We should be a bit cautious about the recollectionsof firms that registered in the early 1990s.These arethe firms that survived and, consequently, may beviewing the past with slightly rose-colored glasses.Thirty days for business registration seemsexceedingly fast given the anecdotal evidence fromthat period. Even with the survivor bias of olderfirms, however, the evidence of improvement inregistration periods appears dramatic.The 1999Enterprise Law cut median registration times in half,to about 15 days, and the OSS procedures havefurther pushed down waiting periods to 10 days in2007 and to single-digits by 2008, although the2008 numbers only involve firms that registered inJanuary before the survey. Furthermore, in line withexpectations for OSSs, the total time before a firmis fully legal has declined dramatically-only 5 percentof firms in 2007 waited more than three months forall of their documentation. None has crossed thatthreshold thus far in 2008.

By way of comparison, we also include the data ontotal registered and active firms, as reported by theMinistry of Planning and Investment (MPI), theGeneral Statistics Office, and the Tax Authority.Focusing on the most accurate indicator-thenumber of registered firms listed on taxpayer rollsin 2008-we can see a dramatic post-OSS spike.More than 50,000 new firms registered theirbusinesses after the onset of OSSs, compared with26,000 the year before. Clearly, something significanthas occurred that is not captured in the WorldBank's Doing Business measure of the officialwaiting periods for business registration, which havebarely budged since 2004 despite the new OSSprocedures. Circulars 02 and 05 may have occurredtoo recently for Doing Business researchers toincorporate the changes into their analysis.


8. Le Nguyen Minh. 2008. "Boã giêëy pheáp khùæc dêëu" [Removal of the

Stamp Licenses] An ninh thuã àö [Capital Security], January 9.9. Association for Small and Medium Size Enterprise Development for

Circular 05 Drafting Committee: 2008. Report on Implementationof the One Stop Shop for Business Registration,Tax Registration, andStamps and Licenses for business establishment in accordance withthe Enterprise Law. Ha Noi,Viet Nam.


1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

Pre-Enterprise Law Post-Enterprise Law OSS

37 99 137

108

114

105

146

134

151

418

552

576

714

844

1059

1026

1139 67

30 30 30 30 30 30 30 30 22.5

15 15 15 15 15 15 14 10 8.5

4 3 5 3 4 4.5 4 5 3 3 3 3 3 3 3 2 2 2

32%

35%

30%

23%

32%

26%

32%

30%

25%

22%

24%

24%

25%

24%

21%

21%

19%

18%

13.0

4%

7.46

%

9.24

%

4.30

%

8.79

%

10.8

7%

8.33

%

12.3

9%

8.81

%

6.83

%

8.41

%

7.49

%

7.04

%

7.03

%

6.41

%

6.21

%

5.08

%

0.00

%

Prov

inci

al C

ompe

titiv

enes

s In

dex

Dat

a 20

08

PCI 2

008

Resp

onde

nts

Esta

blish

edD

urin

g Ye

ar (#

)

Med

ian D

ays

for

Regis

trat

ion

Firm

s w

aitin

gov

er th

ree

mon

ths

to b

efu

lly le

gal

com

pani

es (%

)

New

lyRe

giste

red

Firm

s (M

PI)

Tota

l Act

ivePr

ivate

Firm

s(E

nter

prise

Cen

sus)

Nat

iona

l Tax

Aut

horit

y Li

stof

Tax

-Pay

ing

Ope

ratio

ns

Regis

trat

ion

(Num

ber

ofPr

oced

ures

)

Regis

trat

ion

(Num

ber

ofD

ays)

Firm

s w

aitin

g ov

eron

e m

onth

to b

efu

lly le

gal

com

pani

es (%

)

Med

ianN

umbe

rof

Lice

nses

Year

61 3500

4239

2843

4222

2764

2764

3630

3601

8785

1966

2

1799

2

2598

2

4833

6

3991

8

3998

4

3410

0

518

924

1476

2

1486

2502

7

2596

5

2544

0

3500

4

4431

4

5523

6

6452

6

8400

3

1051

69

1233

92

1443

69*

1265

69

1512

49

1778

15

2282

55

12 11 11 11 11 11

63 56 50 50 50 50

Offi

cial

Dat

a on

Fir

m A

ctiv

ities

Wor

ld B

ank'

s D

oing

Bus

ines

s

Tabl

e 2.

1.H

istor

y of

Priv

ate

Busin

ess E

ntry

in V

ietn

am

Estim

ated

bas

ed o

n re

latio

nshi

p be

twee

n ac

tive

firm

s re

gist

ratio

n da

ta.

Currently, the Doing Business data still count thechop license as an official procedure.10 Anotherdiscrepancy between the PCI and Doing Business isthat the PCI counts some recurring activities, suchas obtaining value-added tax (VAT) receipts andregistrations for health and social insurance, underour measure of Time Costs, while Doing Businesscounts these as entry costs. As noted above, thePCI also finds that there has been little change inthese post-registration activities.

Although the aggregate annual measures show aclear change in performance of provincialregistration processes, it is not clear that we canattribute the improvement directly to the initiationof OSSs. It may simply be that local officials aredoing their jobs more efficiently as a result ofaccumulated experience and perhaps the greaterattention they are receiving from higherauthorities.The exact impact of OSSs on businessregistration can be estimated by exploiting cross-sectional variation in how OSSs are implementedamong different provinces.The needed data forthis test are available from the Agency for SMEDevelopment (ASMED) in the Ministry of Planningand Investment.The data were assembled forinternal use only by the Circular 05 draftingcommittee and were provided to the PCI team,but had not been made publicly available. ASMEDrecords the date that an OSS was initiated in aprovince. According to ASMED's analysis, HaiPhong established the first OSS on July 25, 2006.Lai Chau was the last province to establish anOSS, not until January 11, 2008. Only 60 provincessubmitted data to ASMED and, of those, only 50provided comprehensive data for analysis.

Consequently, our analysis drops 14 provinces,including some that first started experimentingwith OSSs in the mid-1990s, such as Binh Duongand Dong Nai.The missing data should cause us tounder-estimate the impact of OSSs by neglectingthese well-governed entities.

A second useful measure is the percentage offirms that registered through OSSs and those thatentered through older procedures. As provincialadministrations were phasing-in OSSs, they invitedfirms to use the new registration window butallowed others to follow the older methods. Insome cases, firms actually wanted to use the oldersystem because they knew an official in one of thecomponent offices or because they wanted tokeep better track of their file as it passed throughthe bureaucracy.Therefore, it is possible tocalculate a measure of the percentage of firms thatused an OSS in each province as a test of thesystem itself.This metric ranged from 100 percentin 21 provinces to zero in 15 provinces.

Controlling for structural conditions, such asproximity to markets and the capacity of localgovernment (proxied by the number and salary oflocal officials), we find that increasing thepercentage of firms using OSSs from zero to 50percent decreases the median registration periodin a province by 1.3 days.The number may seemsmall, but it is substantively important given thatthe median waiting period before OSSs was 14days. One day, therefore, represents a 9 percentimprovement.This finding is statistically significantand robust to alternative specifications (seeAppendix 2A for the regression results).

Other tests are not significant. According to PCIdata, usage of OSSs is not correlated with thedeclines in the number of documents required,the total time to complete all proceduresnecessary for a business to be fully legal, or thesatisfaction that firms have with the process. Sincethe mandate for OSSs is to reduce the total timerequired for a firm to achieve a legal status, thisfinding is critical. Registration times have beenreduced, but firms still claim dissatisfaction with anumber of other procedures that occurimmediately after registration. Officials may want


10. World Bank. 2008. Doing Business 2009: Comparing Regulation in181 Economies.World Bank:Washington, D.C.The 11 listed DoingBusiness procedures are:1. Obtain a business registration certificate.2. Obtain a seal-making license.3. Make a company seal.4. Pick up company seal.5. Open a bank account.6. Publish a notice in a business paper.7. Register for taxes.8. Buy pre-printed VAT invoices.9. Register with the local labor office.10. Register for health and social insurance.11. Register for trade union with Vietnam General Confederation

of Labour.

to re-orient their attention away from the OSSand on to bottlenecks facing a firm after businessregistration. In terms of the number ofprocedures, the unification of registrationcertificates and tax codes that is currently beingdevised by the MPI with the assistance of theUnited Nations Industrial DevelopmentOrganization is a helpful step in the rightdirection.The efforts of Project 30 to simplifyadministrative procedures and forms within OSSsand after business registration will further reducetime costs of regulatory compliance.

In sum, it would appear that business entryprocedures are improving dramatically. This canbe partially attributed to wide-scaleimplementation of one-stop shops, but it isunclear that OSSs are wholly responsible for thesuccess. General changes in attitude and efficiencyimprovements at the local level may have alsoinfluenced the process.

2.2. Publication of Provincial LegalGazettesFor the past four years, the PCI report has arguedthat the most critical factor in improving the localbusiness environment is the transparency ofbusiness information. Only with access to legal andplanning documents can entrepreneurs properlyplan for the future and assess their business risks.Consistently, firms tend to expand investment mostrapidly when operating in provinces whereregulatory information, master plans, andinfrastructure maps are most widely available.

National policy makers in Vietnam have recognizedthe need for greater access to information. Majorlegislation has been approved to improvetransparency, first with the revision to the Law onthe Promulgation of Legal Normative Documents(LNDs), better known as the Law on Laws, in 2002to require publication of all LNDs for 15 days in the

Official Gazette before coming into legal effect; nextwith the passage of the Law on Local Laws in 2005,which required the establishment of provincialOfficial Gazettes; and most recently with the furtherrevisions of the Law on Laws in 2008 to require 60-day public comment periods for LNDs.Theseinitiatives served not only to meet requirements inthe U.S.-Vietnam Bilateral Trade Agreement and foraccession to the World Trade Organization, but alsoto improve transparency and participation forVietnamese businesses and citizens.

Of particular interest for the PCI, given its focus onprovincial transparency, is the requirement specifiedin Decree 136/2005/ND-CP that mandated thatprovinces publish all LNDs passed at the provinciallevel in a provincial gazette (Công Báo)11. Publicationof the gazettes proceeded slowly at first. On April27, 2006, when the Public Administration ReformSteering Committee of the Government issued itsreport on implementation, it listed only nineprovinces and national-level cities that hadestablished provincial gazettes.12 By June 2008,however, a subsequent review found that all butone province (Ca Mau) published a hard copy of aprovincial official gazette and that an additional 20provinces published searchable, online gazettes.13

Has this growth in the publication of legaldocuments been a source of transparency atprovincial levels? The PCI is well-positioned toanalyze this question using a panel dataset of 2,500firms that have answered the PCI survey every year.Because these firms answer the PCI annually, theyare familiar both with the PCI survey instrumentand with the history of policy changes over time.Therefore, their responses are quite robust, becauseassessing their understanding of improvements overtime is less prone to sampling error and thus morereliable than a new cross-section of firms.

Observing the responses of these 2,500 firms aboutaccess to legal documents over time (Table 2.2)reveals a clear pattern. Far more firms believe they


11. This Circular filled in the details of the Law on Promulgation ofLegal Documents of People's Councils and People's Committees,December 3, 2004.

12. Ho Chi Minh, Da Nang, Ha Noi,Vinh Phuc, Nam Dinh, Binh Duong,Ben Tre, and Can Tho. See Report No.01/BC-BC?CCHC dated April27, 2006 by PAR Steering Committee of the Government.

13. Other important recent transparency initiatives that have not beenrequired by law include the establishment of an ElectronicGovernment Portal by the Office of the Government where manynational laws and regulations can be found. In 2004, the VietnamChamber of Commerce and Industry established the VIBONLINEwebpage (www.vibonline.com.vn), where it posted for publiccomment many draft laws and regulations.

Table 2.2:Access to Business Documentation(Percentage of the 2,500 PCI Panel Respondents who believe access to documentation is possible)


Document

Legal Normative Document

Central Laws and Decrees

Implementing Documents of Ministries

People's Committee Decisions

Incentive Policies

Changes in Tax Law

Official Legal Gazette

Planning Document

Provincial Budget

10-Year Master Plan

Annual Plan

Private Sector Development Plan

New Infrastructure Plans

Central Investment Plan

Access to Land Use Plan and Maps

Business Registration and Land Use Applications

Address and Phone of Local Agencies

Data on Registered Enterprises

Relationship Necessary to Get Documents (% Agree)

Central Changes in Laws are Predictable (% Usually)

Provicial Implementation is Predictable (% Usually)

2008

65.46

65.19

65.19

67.49

48.34

81.11

63.15

47.59

35.15

54.47

57.79

53.11

42.81

30.30

38.64

62.47

52.60

48.60

49.28

13.12

6.46

2007

61.19

60.31

58.26

62.61

47.70

77.09

43.85

31.09

50.77

53.07

49.74

40.54

27.68

36.12

61.75

56.3

13.99

8.42

2006

62.86

60.56

60.90

67.38

49.74

75.72

48.66

35.09

57.84

59.97

54.68

46.93

31.18

41.82

61.75

61.88

16.62

11.38

have access to legal documents than planningdocuments (65 percent vs. 48 percent), and therate of improvement between 2007 and 2008 inaccess is faster for legal documents.This is a signthat the promulgation of laws in provincial gazettesis working, because there is no corresponding lawregarding promulgation of planning documents. As isto be expected, given the timeline ofimplementation, the number of firms reporting

access to documentation in 2006 and 2007 arestatistically indistinguishable (within a 3 percentsampling error range).The major growth in accessto legal documentation occurred between 2007and 2008. Importantly, the percentage of firms thatbelieve relationships are necessary to access thedocuments has declined substantially since themandate that gazettes be published-down from 62percent in 2006 to 49 percent today.

provincial wealth. Several quite rural and poorprovinces, such as Thai Nguyen,Tay Ninh, Ninh Binh,and Ben Tre, have gazettes linked to their websites,while rich provinces like Binh Duong do not. In fact,there is no correlation between an internet version ofthe gazette and provincial wealth.Thirty-two percentof provinces above the median gross domesticproduct per capita of VND 4.5 million have gazettesonline, compared with 26 percent of provinces belowmedian wealth.A t-test shows that this difference isnot statistically significant.

Controlling for wealth and other structural factors,firms in provinces with gazettes online are significantlymore likely to believe they have access to legaldocumentation, especially provincial resolutions anddecisions. Figure 2.1 illustrates this point: it shows theannual change in accessibility of a sample ofimportant documents according to firms in the PCIpanel survey, separated by whether provinces postgazettes online or not. In every case, provinces withgazettes online have demonstrated stronger annualimprovements in accessibility. Confidence intervalsaround the mean values indicate that only three ofthese differences are statistically significant: 1) centrallaws; 2) implementing documents; and 3) provincialresolutions.The biggest substantive difference is inaccess to provincial documentation. Provinces withgazettes online have increased access by over 10percent, while provinces with only paper versions ofthe gazettes are actually rated by firms as havingdiminished accessibility.This makes sense becausethese are the documents most likely to be includedin gazette. Other documents may also be availableonline, but the positive change in the accessibility ofdocuments is highly related to those required by law.Central documents need not be included in agazette legally, but provinces with useful websitestend to post them online anyway, often with links.See Dong Nai's webpage for an excellent exampleof this cross-listing that accounts for the significantsigns on central documentation.

On the downside, predictability of new laws andthe implementation of those laws at the provinciallevel continue to decline despite the proliferation ofgazettes. Few firms seem to have any understandingof what legal changes are in the pipeline; they onlyhave a sense of the legal documents that havealready been issued.This represents a majordeficiency in the process.The few actors who dohave knowledge of potential changes in legislationand local implementation of that legislation have acritical business advantage over their competitors.But acknowledgement of this deficiency should betempered by the fact that the legal environment hasbecome noticeably more complex in recent years,resulting directly from Vietnam's development andexpanded international commitments. Firms'answers about predictability could be influenced bythese macro-developments.

Central officials are aware of the lack of legalpredictability and have already taken steps toaddress the issue. Recent reforms of the Law onLaws 2008, including the requirement for 60-daypublication of all draft legal normative documentsfor public comments and rigorous policy analysisthrough Regulatory Impact Assessments (RIA), willbe implemented by both central ministries andprovincial authorities shortly, which will hopefullyengage affected parties in the process of law andpolicy making. If successful, these innovations willenhance firms' perceptions of predictability.

Beyond the availability of a provincial officialgazette, a key aspect of information access from apolicy perspective is whether the gazette isavailable on provincial websites. As of July 2008,20 provinces have provincial gazettes linked totheir provincial websites. How does an internetversion of the gazette affect firm perceptions ofdocument accessibility?

To answer this question, first keep in mind thatinternet access to a gazette is not a function of



Figure 2.1:Access to Business Documentation, (By Online Availability)

A key conclusion is that the national policy ofpublishing all legal normative documents hasincreased transparency at provincial levels, butprovinces that have gone beyond the nationalregulations and posted their gazettes online havefacilitated much greater accessibility. Posting gazettesonline is an easy step that every province in thecountry can implement immediately.

2.3 Legal CasesAn important development in this year's PCI isthe large increase in the usage of provincialeconomic courts this year, as reported by thePeople's Supreme Court of Vietnam.The numberof total cases filed more than doubled between

2006 and 2007, from 2,445 to 5,198.The numberof cases filed by private entrepreneurs grew by130 percent in 2007 compared with 2006 (seeFigure 2.2). Higher usage of courts by privateactors is generally considered to be a positivedevelopment because it indicates thatentrepreneurs have greater faith in legalinstitutions and are more willing to move beyondsocial relations as a mechanism for enforcingcontracts. It also may reflect the considerableimprovement in contract and other commerciallaws in Vietnam over the past six years since theBilateral Trade Agreement with the US, which hasmade the law clearer in many regards and, thus,provides firms with a greater confidence in theirbusiness rights and responsibilities.


Figure 2.2: Changes in Usage of Court System Over Time

As Vietnam's private firms expand and start to dobusiness with actors in other provinces andcountries-outside of their immediate social circle-they must rely more on formal, arms-length legalprocesses, which must be perceived to be fair andindependent. Further analysis of the disaggregatedPeople's Supreme Court data by case type revealsthat the vast majority of legal cases filed (roughly 70percent) involve some form of contractenforcement on the purchase or sale of goods.Much smaller percentages have to do with financeand banking (8 percent) and problems related toconstruction (6 percent).Thus, on its face, thegreater usage of courts in Vietnam by private firmsappears to be an excellent indicator of successfullegal reform.

On the other hand, respondents to the PCI surveythroughout the country do not report significantly

higher usage of the courts or confidence in locallegal institutions. In fact, the number of firmsresponding that courts are among their preferredthree forms of dispute resolutions actually droppedfrom 18 percent to 12 percent.

Probing deeper into the data, we find that twocorrelates of court usage are the amount ofeconomic activity in the province and the presenceof foreign investors.There is a separate line item inthe Supreme Court data for foreign firms, butprivate firms are also more likely to use courtswhere there is a large foreign investor presence; thiscould be because of greater contractual interactionswith foreigners or simply because foreigners helpshape a more legally oriented business environment.Urban centers and southern locales also tend tohave more private cases filed.


Figure 2.3: Increasing Court Usage and Confidence in the System

Controlling for these factors and the increase ofprivate sector business generally, however, we dofind that provinces that experienced the greatestincreases in court usage from the Supreme Courtdata are also the ones with the greatest increase inconfidence in the court system according to thePCI survey (See Appendix 2B for details).

Figure 2.3 illustrates this relationship and

demonstrates that increasing court usage isassociated with growing confidence in the legalsystem. Provinces experiencing the greatest changein confidence are also the places where court usageappears to be spiking.We should be careful aboutasserting a causal story, because reverse causality isa possibility. Confidence may grow as firms seemore of their peers successfully filing cases.

CHAPTER THREE

INFLUENCE OF PCI ANDPOLICY IMPLICATIONS

The exciting and multifaceted ways the PCI hasbeen embraced and employed since its inaugurationfour years ago have far surpassed our originalexpectations. It has been rewarding to observe themany different individuals and organizations whohave found the PCI useful in their day-to-dayactivities.

3.1 Catalyst for reformSince its launch, the PCI has been widely used byprovincial governments to design economic reforminterventions and to benchmark their own effortsand progress against "star performer" provinces. Inthe past two years, four provinces (Ha Tay, KhanhHoa, Phu Yen, and Binh Phuoc) have issued Party orPeople's Committee Resolutions that formallycommit authorities to improving PCI scores, oftenindicating specific indicators in need ofimprovement and initiatives designed to achievethose efforts.14 Binh Phuoc's recent People's

Committee decision is particularly ambitiousbecause it singles out individual departments andoffers targets for their improved governance.Another 10 provinces have not issued legaldocuments but instead have released non-bindingaction plans for improving PCI scores.15 Despitetheir lower status, these documents show a keenunderstanding of the PCI indicators and publiclydemonstrate a commitment to improvement andaccountability to provincial constituents. A host ofother provincial officials have made public speechesabout the PCI.The speech of the People'sCommittee Chairman of Ha Noi in early 2007 wasmemorable for the detail of his research and theboldness of his suggestions.16

Speading Best PracticesIn addition to the formal commitments to reform,we also have observed provincial officials visitingand learning from one another. In an interview withthe People's Committee of Binh Duong, we weretold that officials from nearly every province havetraveled to Thu Dau Mot, the province's capital, inthe past few years to learn about best practicesfrom the perennial number one in the PCI survey.Intraregional visits are even more common. In a visitto the Mekong Delta, we observed a number ofcases where provincial initiatives had been borrowed


INFLUENCE OF PCI ANDPOLICY IMPLICATIONS

14. Tónh UÃy Haâ Têy. 2005.“Vïì töí chûác àúåt sinh hoaåt kiïím àiïím nhùçmnêng cao nùng lûåc caånh tranh vïì möi trûúâng àêìu tû cuãa tónh.”Nghõ quyïët 14/NQ-TU, June 4; [Party Secretariat of Ha Tay,Resolution 14,“On the organization of review activities in orderimprove the competitiveness and investment environment of theprovince.”].Tónh UÃy Khaánh Hoâa. 2008.“Vïì tiïëp tuåc caão thiïån möitrûúâng kinh doanh vaâ àêìu tû, nêng cao chó söë nùng lûåc caånh tranhcuãa tónh.” Chó Thõ 19/CT-TU, May 15. [Party Secretariat of KhanhHoa, Circular 19,“On continuing the improvement of theenvironment for business and investment, raising the ranking thePCI ranking of the province.”]. UÃy Ban Nhên Dên Tónh BònhPhûúác. Forthcoming.“Vïì viïåc Ban haânh chûúng trònh haânh àöångnêng cao chó söë nùng lûåc caånh tranh (PCI) tónh Bònh Phûúác.Quyïët Àõnh XX/ QD_UBND. [People’s Committee of BinhPhuoc, Decision XX,“On the promulgation of an administrativeprogram to raise the PCI scores of Binh Phuoc Province.”] HöåiÀöìng Nhên Dên Tónh Phuá Yïn Khoáa V, Kyâ Hoåp Thûá 10. Nghõquyïët vïì nhiïåm vuå kinh tïë—xaä höåi, an ninh—quöëc phoâng nùm2008. [People’s Council of Phu Yen, 10th Meeting of the 5thCouncil, “Resolution on the responsibilities for economics,society, security, and defense in the year 2008.]

15. These provinces include Binh Duong, Hai Phong, Ha Noi, DaNang, Ho Chi Minh City,Thai Nguyen,Tien Giang, Kien Giang, CaMau,Thua-Thien Hue, and Lao Cai.

16. Nguyïîn Thïë Thaão. 2008.“Phaát biïíu cuãa Chuã tõch UBND TP - taåikyâ hoåp thûá 11 HÀND khoáa XIII:Nùng àöång saáng taåo trong töí chûác thûåc hiïån, phêën àêëu hoaânthaânh vûúåt mûác kïë hoaåch KT-XH.” [Presentation of the People’sCommittee Chairman of Ha Noi at the 11th Meeting of the 9thHa Noi People’s Council: Proactivity and Creativity withinOrganization and Implementation; Striving for Exceeding theGoals of the Socio-Economic Plan.]

from neighboring provinces.Tra Vinh and Long Anprovinces were particularly proactive about learningfrom their compatriots, citing registration procedures,land policy, and investment promotion as specificsources of inspiration and cooperation. Oneexplanation for the startling improvement andconvergence among provinces in the delta has beentheir willingness to share reform initiatives.

3.2 Uses of the PCIPrivate Sector - Investment Decisions andAdvocacy

Private firms have used the index to better informtheir investment decisions, and to advocate for localinitiatives intended to make their locale moreconducive for business. The PCI is increasingly usedby Provincial People's Councils to question theperformance of the administrative departmentsduring annual query sessions, which includesrepresentatives from the business community.As aresult, local officials have requested meetings with thePCI research team to discuss the details of thepolicies cited by the People's Council, so that they canbetter understand the methodology and design of thePCI.Another useful case of local business advocacyoccurred in Binh Duong province, when anentrepreneur objected to a particular local regulationand went on a radio show to declare that he wouldgive Binh Duong lower PCI scores if the regulationwas not altered.The PCI research team has witnessedsuch advocacy in the course of our provincialdiagnostics in a number of localities.17 Firms often usethe PCI indicators as a frame of reference for specificlocal initiatives that they would like to see changed.

Donors and Academic WorldMembers of the international donor community-both multilateral and bilateral agencies-have usedspecific PCI indicators for monitoring and evaluatingtheir own local economic development projects.The U.S. Ambassador and Consul General in HoChi Minh City are frequent users of the data in

their interactions with provincial leaders, but thereach of the PCI extends beyond just the Americancommunity.The Danish International DevelopmentAgency program specifically monitors success in itsfour target provinces by using PCI benchmarks.18

The German Technical Cooperation Agency,International Labour Organization, and InternationalFinance Corporation also make frequent use of thedataset and findings.19

PCI data have been employed in a range of academicstudies. Key PCI results have been used in researchregarding private sector performance,20 determinants offoreign direct investment,21 access to bank credit,22 statesector crowding out of private business,23 industrialpolicy,24 poverty alleviation,25 and even the role of theNational Assembly in policy making.26


17. Provincial diagnostics are tailored presentations of the PCI results toprovincial leaders.To date, PCI teams have travelled to more than 40provinces, some more than once, to discuss results with leaders.

18. Danish Embassy. 2008. “ Danida’s Support to Ha Tay ProvinceImproves Business Climate Significantly.” April 4. http://www.ambHaNoi.um.dk/en/menu/AboutUs/News/DanidasSupportToTheProvinceOfHaTayImprovesBusinessClimateSignificantly.htm

19. International Finance Corporation. 2008. "Provincial EconomicGovernance and Its Impact on Local Competitiveness." BusinessIssues Bulletin. 23(26), March.

20. Tran,Thi Bich, R. Quentin Grafton, and Tom Kompas (2008)."Institutions Matter :The Case of Vietnam." Journal ofSocioeconomics. (forthcoming).

21. Vu Thanh Tu Anh, Le Viet Thai,Vo Tat Thang. 2007. "ProvincialExtralegal Investment Incentives in the Context of Decentralisationin Viet Nam: Mutually Beneficial or a Race to the Bottom?" UnitedNations Development Program Vietnam Working Paper Series. HaNoi:Vietnam; Malesky, Edmund. 2008. Provincial Governance andForeign Direct Investment in Vietnam." in Twenty Years of ForeignInvestment in Vietnam." Ho Chi Minh City: Saigon Economic Times;Reihl, Evan (2008). "Regional Regulations and Foreign Investment inVietnam." Unpublished Manuscript. University of Washington: St.Louis

22. Malesky, Edmund and Markus Taussig. 2008. "Where is Credit Due?Legal Institutions, Connections, and the Efficiency of Bank Lending inVietnam." Journal of Law, Economics, and Organization. Availableonline June 20(http://jleo.oxfordjournals.org/cgi/content/abstract/ewn011v1?andorexacttitleabs=and&HITS=2&hits=2&andorexacttitle=and&maxtoshow=&andorexactfulltext=and&resourcetype=HWCIT&titleabstract=vietnam&searchid=1&RESULTFORMAT=)

23. Nguyen Van Thang. 2005. "Is The Development of State-OwnedEnterprises (SOEs) Crowding Out The Private Sector? Evidencefrom PCI Survey," Ha Noi,Vietnam: US-AID, August.

24. Perkins, Dwight and Vu Thanh Tu Anh. 2008. "Vietnam's IndustrialPolicy: Designing Policies for Sustainable Development. UNDP-Harvard Policy Dialogue Papers.

25. Nguyen, Hoang-Phuong, 2008. "What is in it for the poor? Evidencefrom fiscal decentralization in Vietnam," Journal of Public andInternational Affairs, Chapter 4.http://www.hbs.edu/research/pdf/08-099WP.pdf.

26. Malesky, Edmund and Paul Schuler. 2009. "Paint-by-NumbersDemocracy:The Stakes, Structure, Results, and Implications of the 2007Vietnamese National Assembly Elections," (with Paul Schuler). Journalof Vietnamese Studies. Forthcoming (January 2009).

3.3 Influence of the PCIAccording to our records,Vietnamese journalistshave cited the index in more than 500 investigativeand news articles on provincial businessenvironments. Each year, the Saigon EconomicTimes,Vietnam's most important economic weeklyand the local equivalent of the Economist, has aspecial issue with the PCI results displayed on itsfront cover.27 But information from the PCI has alsoextended beyond coverage of the launch and hasbegun to inform normal journalistic research. Manyjournalists use PCI data in the leads of their story asa way to quickly introduce the businessenvironment in a province or group of provinces.These articles in turn have had an impact on policymakers. A sampling includes a story on differing taxpolicies in Ben Tre and Tien Giang provinces, a seriesdiscussing planned governance improvements in HaTay, and the challenges facing a new People'sCommittee Chairman in Hau Giang.28

Indeed, because of the PCI's impact, VTV 1 ranked itamong the 10 most important economic events tooccur in Vietnam in 2005. Reflecting the extent towhich the PCI has entered into wider socialawareness, it even found its way into the final questionfor university students competing in the well-knowntelevision game show Ring the Golden Bell and Face Offon VTV3.29

National policy makers have also taken note ofthe PCI. Both Prime Ministers Phan Van Khai and

Nguyen Tan Dung quoted PCI statistics indialogues with the private sector and uponintroducing new legislation.30 Truong Tan Sang,Director of the Standing Committee of the PartySecretariat, recently discussed the utility of thePCI for evaluating the performance of localofficials and creating incentives for policy makers.31

Sang even suggested that a ranking of national lineministries be created.

In short, the PCI has become an importantcomponent in Vietnam's policy dialogue relating toeconomic and administrative reform, as well asbusiness liberalization, both at the provincial and thenational levels. However, with the increased visibilityand application of the PCI index, there have alsocome increasing responsibilities for the researchteam.We are obligated to ensure that policymakers and other readers of the PCI report aregiven the most accurate, carefully analyzed, andunbiased information available. It is a responsibilitythat we take very seriously.

3.4 The Way forwardThe multiple uses of the PCI have been quiteremarkable and inspiring, but there are ampleopportunities for new applications. Surveying theeconomic policy horizon, here are somesuggestions of how to build PCI indicators intopolicy evaluation programs.


27. Hoang Lang. 2005.“Caånh tranh cêëp tónh” [Competiteveness ofthe Provinces].Thúâi baáo Kinh tïë Saâigoân [Saigon Economic Times],May 26: 14. Phuong Quynh. 2006.“Nhûäng Bêët Ngúâ Cuãa PCI2006” [Many Surprises: PCI 2006].Thúâi baáo Kinh tïë Saâigoân[Saigon Economic Times]. June 1: 12. Quynh Nhu. 2007.“PCI2007: Àùçng Sau Nhûäng Con Söë” [PCI 2007: Behind theNumbers].Thúâi baáo Kinh tïë Saâigoân [Saigon Economic Times],November 8, 2007: 10.

28. Van Truong. 2005.“Tiïìn Giang: Dên kïu vò thuïë” [Tien Giang:Citizens Complain about Taxes].Tuöíi Treã Online [Youth Online],August 6. Le Dang Khoat. 2006.“Haâ Têy caãi thiïån möi trûúâng àêìutû.” [Ha Tay improves investment environment].Thúâi baáo Kinh tïëViïåt Nam [Vietnam Economic Times].Vol. 42; Nguyen Anh. 2008.“Chuã tõch UBND tónh Hêåu Giang Huyânh Minh Chùæc: Döìn sûác bûátphaá trûúác vêån höåi múái [People’s Committee Chairman of HauGiang, Huynh Minh Chac:Tearing up Everything Before a NewCulture.” Kinh tïë Nöng Thön [Agricultural Economics] January 22.http://www.kinhtenongthon.com.vn/Story/xuanmauty/gapgodauxuan/2008/1/8921.html

29. http://www.vtv.vn/VN/TrangChu/Media/#.

30. Phan Vùn Khaãi. 2006. "Doanh nhên laâ nhûäng chiïën sô xung kñch thûåchiïån nhiïåm vuå phaát triïín kinh tïë - xaäc höåi 2006 - 2010 vaâ höåi nhêåpquöëc tïë thaânh cöng" [Entrepreneurs are the soldiers, aggressivelypushing forward in the implementation of economic development(2006-201) and successful international integration]. Phaát biïíu cuãaThuã tûúáng Phan Vùn Khaãi kïët thuác Höåi nghõ vúái doanh nghiïåp, HaNoi, January 9, 2006 [Speech of Prime Minister Phan Van Khai toconclude the Enterprise Conference];Thöng Baáo Kïët luêån cuãa Thuãtûúáng Chñnh phuã Nguyïîn Têën Duäng taåi buöíi laâm viïåc vúái laänh àaåothaânh phöë Haâ Nöåi [Communication of Prime Minister Nguyen TanDung to conclude the working meeting with the leaders of Ha NoiCity], S?: 262/TB-VPCP December 18, 2007.

31. Tuöíi Treã [Youth]. 2008. "Thûúâng trûåc ban bñ thû Trûúng Têën Sang:Cêìnào chó söë haâi loâng vïì caác böå ngaânh." [Head of the Party SecretariatTruong Tan Sang: We need a ranking of ministries], July 22.http://www.tuoitre.com.vn/Tianyon/Index.aspx?ArticleID=269871&ChannelID=3

Year-on-year progress of the Government's path-breaking Project 30 to reform administrativeprocedures across all ministries and provinces canbe tracked directly by the PCI sub-indices of entrycosts, land access, transparency of information,time costs of regulatory compliance, and informalcharges. Provinces that are diligent in carrying outthe inventory and review of administrativeprocedures in order to cut and simplify thecumbersome paperwork for firms will most likelysee these efforts reflected in PCI indicators, asfirms come to appreciate how the efforts haveimproved transparency and reduced informationgathering costs.

Moreover, the PCI's detailed questions aboutinformal charges and commissions paid by firms togovernment officials can be employed to evaluatethe efforts of the government's new NationalStrategy for Combating and Preventing Corruptionthrough the Government Inspectorate. Futurescores can be compared to the historical indicatorscollected by the PCI research team. Findings from

this comparative analysis can be used to monitorprogress and devise new procedures if initialendeavors fail to bear fruit.

Appropriate evaluation of the impact ofdecentralization is a major area for concern forGovernment leaders, because no systematicmechanism for evaluation currently exists.The PCIprovides both yearly snapshots of local authorityperformance in economic governance and paneldata on annual changes that can be easily integratedinto these efforts. Initial conversations with theMinistry of Home Affairs and the CentralCommittee of the Party indicate they are interestedin using PCI in their daily oversight of thedecentralization policies in Vietnam.

The National Assembly of Vietnam is working hard tocreate the rule-of-law for the emerging marketeconomy in Vietnam. PCI equips deputies with a toolto oversee the effectiveness and efficiency of theGovernment in carrying out a wide range of reforms.As noted in Chapter 2, judicial reform can also beassessed by the rich dataset PCI obtains in court casesas well as the confidence of firms in the court system.


CHAPTER FOUR

ANALYSIS OF CHANGINGEXPECTATIONS AND THEDECLINE IN PCI SCORES

As noted in Chapter One, two sub-index scoresshow considerable decline-Labor and Private SectorDevelopment (PSD) policies.These fundamentaldrops dramatically affected the overall PCI scores,causing a consistent slide across the rankings. Inaddition, a battery of questions about infrastructurethat were employed in the brand new infrastructureindex (described in Chapter Five), also showedsignificant year-on-year decline.What caused this fallin scores? It is important to understand the roots ofthe decline because it will have critical implicationsfor Vietnam's future growth and directions in policymaking. After exploring a range of possible causes,we were able to rule out any PCI methodologicaldecisions as reasons for the fall in scores. Decreasesare due primarily to the increasing expectations ofprivate entrepreneurs and the inability ofVietnamese policy makers to keep pace with theirneeds. In fact, as we document below, there haveactually been significant declines in public servicedelivery over the past year that have contributed tofalling perceptions.

As always, this year's PCI survey was mailed torespondents after Tet.Tet is an auspicious time;employees of firms are more likely to be home and

have the free time to spend a few moments fillingout the Vietnam Chamber of Commerce andIndustry queries.We usually accept the slightlyupward bias in scores caused by the celebratoryatmosphere because of the higher response ratewe receive during this period.This year, however, thebad macroeconomic news during the season mayhave influenced the answers firms gave to many ofthe survey questions.

It is critical to note that not all indices receiveddeclining scores. In a number of areas, particularlythose highlighted in Chapter One (Entry Costs,Property Rights,Transparency, and Confidence inCourts), improvements in the policy environmentare clearly reflected in more positive firmperceptions.Table 4.1 (also see Chapter 6)demonstrates, in fact, that most sub-index andindicator scores registered marginal improvementsthis year.The fact that declines were onlyexperienced in a particular set of questions isinteresting; it indicates that we are not capturinggeneral negativity in perceptions. Firms are able todiscriminate among issues in the businessenvironment that have improved and others thatfrustrate them.


ANALYSIS OF CHANGINGEXPECTATIONS AND THEDECLINE IN PCI SCORES


Fina

l Wei

ghte

dPr

ovin

cial

Com

petit

iven

ess

Inde

x

Unw

eigh

ted

Tota

lIn

dex

Entr

y C

osts

Land

Acc

ess

&Se

curi

ty o

f Ten

ure

Min

Med

ian

Max

Cor

rela

tion

w/

Prev

ious

Year

Min

Med

ian

Max

Cor

rela

tion

w/

Prev

ious

Year

Min

Med

ian

Max

Cor

rela

tion

w/

Prev

ious

Year

Min

Med

ian

Max

Cor

rela

tion

w/

Prev

ious

Year

36.0

7

52.4

1

77.6

1

NA

42.5

1

55.2

3

74.8

7

NA

4.96 7.4

9.17 NA

3.84 6 7.98 NA

Lai C

hau

Lam

Don

g/Th

ai N

guye

n

Binh

Duo

ng

Lai C

hau

Hoa

Bin

h/La

m D

ong

Binh

Duo

ng

Binh

Phu

oc

Ha T

inh/

BRVT

Da

Nan

g

Lai C

hau

Qua

ng N

gai/B

ac K

an

Soc T

rang

37.9

6

55.5

6

77.2

0.85

*

43.9

3

58.4

9

76.0

2

0.82

*

6.23

7.87

9.49

0.33

*

4.32

6.27

7.71

0.68

*

Dak

Non

g

Bac

Gia

ng/P

hu T

ho

Binh

Duo

ng

Dak

Non

g

Thai

Bin

h/H

a G

iang

Binh

Duo

ng

Hau

Gia

ng

Hai

Duo

ng/L

am D

ong

Qua

ng T

ri

Ha

Noi

Lao

Cai

/Tha

i Ngu

yen

Long

An

36.3

9

53.1

7

72.1

8

0.90

*

45.2

9

58.1

7

72.8

7

0.88

*

6.31

8.25

9.36

0.25

*

4.73

6.68

8.05

0.73

*

Die

n Bi

en

Dak

Lak

/Bin

h Ph

uoc

Da

Nan

g

Die

n Bi

en

Phu

Tho/

Kien

Gia

ng

Binh

Duo

ng

Bac

Gia

ng

Nam

Din

h/Kh

anh

Hoa

Da

Nan

g

Ha

Noi

Than

h H

oa/H

oa B

inh

Don

g Tha

p

Inde

xM

easu

re20

0620

0720

08

Scor

ePr

ovin

ceSc

ore

Prov

ince

Scor

ePr

ovin

ce

Tabl

e 4.

1:C

ompa

rison

of O

vera

ll Sc

ores

& S

ub-In

dice

s (20

06-2

008)


Tran

spar

ency

Tim

e C

osts

of

Regu

lato

ryC

ompl

ianc

e

Info

rmal

Cha

rges

Bias

tow

ard

Stat

eO

wne

d En

terp

rise

Min

Med

ian

Max

Cor

rela

tion

w/

Prev

ious

Year

Min

Med

ian

Max

Cor

rela

tion

w/

Prev

ious

Year

Min

Med

ian

Max

Cor

rela

tion

w/P

revi

ous Y

ear

Min

Med

ian

Max

Cor

rela

tion

w/P

revi

ous Y

ear

2.15

5.43 8.5

NA

2.64

4.42

7.12 NA

5.05

6.33

8.35 4.7

6.49 8.4

NA

Dak

Non

g

TT-H

ue/B

RVT

Binh

Duo

ng

Phu

Yen

Qua

ng N

gai/K

ien G

iang

Binh

Duo

ng

Ha T

inh

Bac

Kan/

Bac

Lieu

Ben

Tre

Ha

Noi

Da

Nan

g/La

ng S

on

Lao

Cai

2.24

5.83

8.56

0.6*

2.99

6.21

8.18

0.36

*

5.35

6.58

7.71

0.33

*

4.79

6.72

8.29

0.56

*

Dak

Non

g

Ha T

inh/

Phu

Tho

Lao

Cai

Lai C

hau

Vinh

Lon

g/Ph

u Th

o

Ha T

ay

Ha

Noi

An

Gia

ng/H

a G

iang

Hun

g Yen

Dak

Non

g

Qua

ng B

inh/Q

uang

Nam

Binh

Duo

ng

2.99

6.32

7.92

0.74

*

2.85

5.38

6.52

0.62

*

5.7

6.65 8.3

0.50

*

5.99

7.53

8.77

0.54

*

Dak

Non

g

Hai

Pho

ng/D

ak L

ak

Da

Nan

g

Die

n Bi

en

Tuye

n Q

uang

/TT-

Hue

Binh

Phu

oc

Bac

Kan

Nin

h Th

uan/

An

Gia

ng

Hun

g Yen

Bac

Lieu

Hai

Duo

ng/T

ien

Gia

ng

Ha

Gia

ng

Inde

xM

easu

re20

0620

0720

08

Scor

ePr

ovin

ceSc

ore

Prov

ince

Scor

ePr

ovin

ce


* Si

gnific

ant

at 5

% L

evel

;NA

= N

ot A

pplic

able

All

valu

es a

re a

t th

e pr

ovin

cial

-leve

l.D

ata

incl

udes

onl

y fir

ms

regi

ster

ed w

ithin

tw

o ca

lend

ar y

ears

pre

cedi

ng t

he s

urve

y.20

05 d

ata

only

incl

ude

42 p

rovi

nces

and

do

not

incl

ude

the

full

set

of in

dica

tors

use

d in

sub

sequ

ent

year

s,re

flect

ing

chan

ges

in s

urve

y qu

estio

ns a

nd o

rder

ing

in 2

006.

Proa

ctiv

ity

Priv

ate

Sect

orD

evel

opm

ent

Polic

ies

Labo

r Po

licie

s

Lega

l Ins

titut

ions

Min

Med

ian

Max

Cor

rela

tion

w/

Prev

ious

Year

Min

Med

ian

Max

Cor

rela

tion

w/

Prev

ious

Year

Min

Med

ian

Max

Cor

rela

tion

w/P

revi

ous Y

ear

Min

Med

ian

Max

Cor

rela

tion

w/P

revi

ous Y

ear

1.54

4.83 10 NA

2.4

4.88

9.62 NA

1.99 5.1

9.6

NA

2.13

3.63

6.55

Qua

ng N

gai

Tuye

n Q

uang

/Tha

i Bin

h

Binh

Duo

ng

Dak

Non

g

Ha

Gia

ng/K

ien

Gia

ng

Da

Nan

g

Lai C

hau

Lang

Son

/Qua

ng N

am

Da

Nan

g

Qua

ng N

gai

Son

La/N

inh

Binh

Bac

Gia

ng

2.3

4.95 9.2

0.79

*

2.26

4.71

8.73

0.79

*

1.92

5.02

8.34

0.81

*

2.24

4.33

6.56

0.37

*

Cao

Ban

g

Phu

Tho/

Thai

Bin

h

Binh

Duo

ng

Bac

Lieu

Ben

Tre/

Tuye

n Q

uang

HC

MC

Lai C

hau

Qua

ng N

am/Q

uang

Nga

i

Da

Nan

g

Ha T

inh

Phu

Tho/

Vinh

Phu

c

Bac

Kan

2.32

5.56

8.45

0.78

*

1.4

3.35

6.35

0.86

*

1.84

4.25 8.4

0.85

*

2.5

4.66 6.7

0.33

*

Bac

Kan

Lai C

hau/

Phu

Tho

Binh

Duo

ng

Bac

Lieu

Nam

Dinh

/Tuy

en Q

uang

HC

MC

Die

n Bi

en

Thai

Bin

h/Ba

c Ka

n

Da

Nan

g

Cao

Ban

g

Son

La/H

ai D

uong

Gia

Lai

Inde

xM

easu

re20

0620

0720

08

Scor

ePr

ovin

ceSc

ore

Prov

ince

Scor

ePr

ovin

ce


Could minor changes in the sample of firmsresponding have caused the drop in the PSD andLabor sub-indices? We can say definitively that thiswas not the case. It is true that the absolutenumber of firms responding, the response rate, andthe proportion of firms covered in each provincewere all significantly higher this year. Perhaps the well-intentioned efforts to improve response rates andreach a larger portion of the private sectorpopulation caused the survey to tap into a morenegative group of firms that had not bothered torespond in earlier iterations.This possibility can easilybe checked by using the panel data of 2,500 firmswho fill out the survey every year. Because these arethe same firms every year, we need not worry abouta new population of angry enterprises.The 2,500panel firms reflect the same negativity as the largersample, and, in some cases, the decline among panelrespondents was even larger. Furthermore, otherquestions showed significant upticks.Thus, includingnew firms with a bias toward negative responses wasnot to blame for lower scores.

A second methodological concern is that a marginalchange in survey design could have introduced a“framing effect” into the survey. A framing effectoccurs when the choice of language in one part ofthe survey inadvertently or intentionally influencessubsequent responses by manipulating the mindset

of the respondent (for example, asking respondentsto reflect on the number of survivors of a planecrash as opposed to the number of casualties). Intrying to improve the flow of the survey, wechanged the title of Section E from “Policies andPlanning” to “Infrastructure and Other Elements ofBusiness Activity.” Not a single question changed; theonly alteration was the bolded heading at the topof the page. It was a minor adjustment, to be sure,but perhaps it inadvertently introduced a morenegative frame of mind. Because most questionsrelated to PSD and Labor were in this section, thisissue deserves further analysis.

As shown in Table 4.2, the decline across items isnot proportional. Firms record a drop of almost 40percent in their thoughts about the quality ofeducation policy and infrastructure, but only a 16percent decline related to telecommunications. And,for the most part, the majority of firms were quitepositive about telecommunications.While framingmay have played a minor role in focusing firms onthe actual quality of these items and away fromprovincial plans to improve them, it alone cannotaccount for the steep fall—not seen since 2005—on education, PSD, and indicators of infrastructure,which are not included in the PCI ranking, but areused subsequently in Chapter Five to create a newinfrastructure index..


FFuullll SSaammppllee ((SSeelleecctteedd QQuueessttiioonnss))

Education

General Infrastructure

Roads

Telecom

PPaanneell

Education

General Infrastructure

Roads

Telecom

Market Information

Consulting on Regulatory Information

Industrial Zones

Equitization

Business Partner Match-Making

Attraction of FDI

Vocational Training

Labor Exchange Services

Export Promotion and Trade Fairs

Item

36.56

21.78

29.34

66.93

28.68

16.77

24.17

61.02

16.17

17.11

20.68

19.06

8.09

13.28

17.02

13.87

18.38

2008 2007 2006 2005

73.6

61.02

65.96

82.41

62.18

52.47

61.93

77.26

37.99

41.99

43.10

47.44

25.55

26.83

47.27

42.16

47.19

42.6

34.79

30.91

59.79

72.65

58.73

68.41

82.33

55.79

46.17

60.56

74.96

43.10

39.86

33.73

38.07

22.83

20.10

45.06

36.88

40.80

Table 4.2.Assessment of Provincial Services and Infrastructure(% Believing Quality is Good or Very Good)

A third potential factor in the lower scores was thegeneral negativity among respondents caused by thebad news in the macroeconomy. Newspaperheadlines in February and March announced theonset of hyperinflation, bank failure, currencydepreciation, and a sharp slowdown in economicgrowth. Could these factors have enhanced thepessimism of firms in the country? We explored thispossibility by studying the average tenor of theresponses by the month the PCI survey was

received. If the downturn in the economy,accompanied by the desperate news stories,influenced the responses of entrepreneurs, weshould be able to see it based on the date ofreturn. Firms responding in April and May should bemore negative than those responding immediatelyafter Tet in February and March-the worsteconomic news appeared at the end of March andanxiety among economic indicators increased untilthe end of July.Table 4.3 chronicles the results.


Tabl

e 4.

3:Fi

rms'

Evalu

atio

ns o

f Inf

rast

ruct

ure

and

Publ

ic Se

rvice

,By

Mon

th o

f PCI

Res

pons

e an

d Typ

e of

Firm

Own

ersh

ip

Full

Sam

ple

All

Firm

s

- Fe

brua

ry 0

8

- M

arch

08

- Apr

il/M

ay 0

8

Sole

Pro

prie

rtors

hips

All

Firm

s

- Fe

brua

ry 0

8

- M

arch

08

- Apr

il/M

ay 0

8

Limite

d Lia

bilit

y

Com

pani

es

All

Firm

s

- Fe

brua

ry 0

8

- M

arch

08

- Apr

il/M

ay 0

8

Resp

onse

Dat

e

26.5

0%

19.8

6%

41.7

8%

28.6

4%

28.7

9%

27.4

1%

27.6

3%

29.5

8%

27.5

7%

25.4

6%

25.5

2%

31.8

2%

Road

s

71.0

4%

77.5

0%

68.5

3%

66.5

5%

58.5

9%

66.8

9%

56.8

4%

50.1

4%

77.8

3%

80.6

6%

76.2

7%

77.3

2%

Firm

s Pl

anni

ngto

Exp

and

Ope

ratio

ns

(%)

(% G

ood

or V

ery

good

)

Tele

com

Educ

atio

nM

arke

tIn

form

atio

n

52.6

0%

48.3

6%

50.5

3%

58.4

2%

66.0

6%

63.1

7%

67.2

8%

67.3

2%

61.9

4%

60.0

7%

62.2

4%

65.9

1%

17.9

3%

9.24

%

15.2

3%

18.6

0%

18.1

1%

16.4

9%

18.8

0%

18.3

1%

18.3

4%

16.4

4%

17.6

9%

18.7

5%

Con

sulti

ng o

nRe

gula

tory

Info

rmat

ion

18.5

0%

8.70

%

18.1

5%

20.2

0%

19.4

1%

18.6

3%

20.1

7%

20.0

0%

18.8

2%

18.4

0%

19.8

5%

18.1

8%

Gen

eral

Infra

stru

ctur

e

20.9

2%

15.4

5%

15.8

5%

22.3

9%

18.4

4%

15.4

2%

19.9

4%

18.8

7%

20.4

4%

19.2

1%

18.8

8%

22.4

4%

27.7

6%

24.0

9%

25.3

9%

26.7

9%

32.4

6%

31.6

9%

33.4

9%

28.7

3%

30.9

5%

30.6

7%

29.2

5%

33.5

2%

In response to whether firms plan to expand theirbusinesses in the coming years, there is a 5 percentdownturn for the population as a whole. Small soleproprietorships reported an 8 percent decline andwere more greatly affected than larger limitedliability companies. Nevertheless, this cautiousnessabout business prospects did not appear totranslate into negativity regarding infrastructure andother public services. No pattern of decline isevident among any population, and some indicatorseven demonstrate a small uptick.

In addition to the impact of macroeconomicinstability, firm expectations provide anotherexplanation for the decline in PCI scores. Firmevaluations of services have declined either becausethe firms have witnessed a decline in the quality ofservices offered or because their expectations havegrown to the point that they expect continualimprovements but instead have seen stagnation orrelative decline. During the boom times of the pasttwo years, with opportunities expanding likewildfire, these issues were an afterthought. Now,however, with the price of inputs rising and growthslowing, such concerns have become more pressing.To illustrate this point, Figure 4.1 charts the PCIperception of education quality alongside the actualpercentage of secondary graduates in the country,measured in the General Statistics Office handbook.This shows that there was a substantial real drop inthe percentage of secondary graduates between2006 and 2007, from 93 percent to 82 percentnationally, with a freefall from 72 percent to 38percent occurring in the most poorly educatedprovinces. Note that this decline was recorded oneyear before the Ministry of Education recalibratedthe standard for secondary school graduationrates.32 These declines have probably beenexacerbated in 2008 as inflation and tighteningpurse strings have made it more difficult for poorer

students to afford school fees.Vocational schoolshave experienced similar declines in enrollmentrates.Thus, when the PCI survey asked firms to ratethe quality of labor in their provinces, only 18.5percent believed that labor quality was sufficient forthe needs of local businesses.

Indeed, a group of Harvard Researchers havesingled-out education and training as the singlelargest deficiency in the Vietnamese economy, callingit a crisis at all levels.33 It is not just aboutinsufficient spending; a great deal of waste exists inthe education system. A recent report highlightedextensive corruption in the educational bureaucracythat impedes proper training and skill-setdevelopment.34


32. Tiïën Duäng. 2008. “Gêìn 250.000 thñ sinh THPT trûúåt töët nghiïåp”[Nearly 250,000 secondary school examinees fail to graduate.”VNEXPRESS, June 20. http://vnexpress.net/GL/Xa-hoi/Giao-duc/2008/06/3BA0391A/. Kiïìu Oanh. 2007. “Khöng duâng tyã lïåtöët nghiïåp THPT àïí àaánh giaá thi àua” [The proportion ofsecondary school graduates is not correct enough in order toevaluate competitiveness] VietnamNet, May 3.

33. Harvard Fulbright Program. 2008. Choosing Success:The Lessons ofEast and Southeast Asia and Vietnam’s Future. Ha Noi: UnitedNational Development Programme: 4

34. Nguyen Dinh Cu. 2007. Corruption in General Education inVietnam. Ha Noi: Institute of Population and Social Studies –National Economic University.


Figure 4.1:Average Firm Evaluation of Education versus Actual Educational Quality

We cannot attribute the entire drop in the LaborIndex to declining educational attainment; otherfactors certainly played a role. Expansion of privatefirms over the past few years has tightened thelabor market.The most talented individuals are notavailable anymore, so firms may be tapping intosecond-tier employees. Similarly, migration to high-growth areas has declined in the past year for avariety of socioeconomic reasons, which furtherreduces the pool of talent in the areas that mostneed high-quality labor. All of these elements pointto the fact that firm complaints about labor qualityare reflecting real trends in the businessenvironment that have taken on greater saliency ina declining economy.

A similar trend can be seen in infrastructure.Though road quality has not seen a decline inabsolute quality, it also has not improved

dramatically enough for the needs of businesses. Asa result, firms have expressed their displeasure inthe survey. Since 2002, the coverage of paved roadsin the country grew from 40,000 kilometers (about34 percent of all roads) to 64,000 kilometers(about 42 percent of all roads).This is an annualgrowth rate of 15 percent, compared with a 26percent annual growth in private companies overthe same period. Infrastructure development issimply not keeping pace with the demands on itsusage and firms are beginning to become frustratedwith the constraints it imposes on theirdevelopment. According to the 2008 data, privatefirms lose an average of 7.5 days per year ofbusiness because roads to the provincial capital areblocked by rainy conditions; 71 percent ofmanufacturing firms experience some loss in goodsdue to breakage on poor roads, which in turn coststhe average firm about VND 43 million (US$2,530).


35. Harvard Fulbright Program. 2008.36. Support for Trade Acceleration Project (STAR). 2008. Supporting

Vietnam's Legal and Governance Transformation. Ha Noi: USAID,February.

The downturn is also evident in survey responsesregarding business services provided by provincialauthorities. Firms did not express severe difficultieswith the availability of business services whenexpansion was easy, but in the current environment-when market information and trade fairs are ofcritical importance-local authorities have not metthe demands of the private firms in their provinces.Provincial budgets indicate that budget trimming isresponsible for cuts in some programs. More likely,however, is that the increasing growth of privatefirms has augmented the demands for theseservices at the same time the growth rate inprovincial bureaucracies has been slasheddramatically. Between 2002 and 2003, the numberof provincial government employees grew by 10percent, compared with 1 percent growth in 2004and an actual 0.19 percent decline in 2005.Whilethere are important reasons for reducing thenumber of local government officials, it is importantto acknowledge that such an action does impactservice delivery in the short term. A wide array ofprivate providers of businesses services-about 3,825law and consulting firms - have begun to fill thevacuum by providing many of the same services.However, these private providers are mostlyconcentrated in the large metropolitan areas andhave yet to reach out to more distant locales.

In conclusion, there is no evidence that the negativeresponses reported in this year's PCI report were

caused by data and methodological problems.Weare confident that the responses of private firmsreflect a growing perception that provincialgovernments were less effective in 2008 than 2007in providing public services-specifically labor training,private sector development policies, andinfrastructure. Firms have grown as fast as possible,given the existing conditions for development, butthey have begun to push up against the parametersset by those conditions.This finding confirmsresearch by the Harvard Fulbright School, whichfound that education and infrastructure rank amongthe most pressing developmental issues facingVietnam today.35 This has important ramifications,especially in the context of the major reformsimplemented at the national and provincial levelsover the past five years. Indeed, the Support forTrade Acceleration Project (STAR) reports haveidentified a similar problem as the implementationgap between better "letter of the law" compliancebut less improvement in the application of the newlaw in practice.36

CHAPTER FIVE

INFRASTRUCTURE ANDINFORMATION ANDCOMMUNICATION

TECHNOLOGY

Infrastructure is one of the most importantconstraints on economic growth and productivity inVietnam. Inadequate infrastructure hinders marketparticipation and trade facilitation, limitscompetition, and increases transportation costs forall enterprises. Indeed, the costs of shipping a 40-foot container are nearly twice as high in Vietnam asthey are in China.37 In previous years, the PCIanalysis has controlled infrastructure as a naturalendowment to more precisely estimate the impactof economic governance in provinces. This reportincludes a provincial analysis of infrastructure toincrease awareness of its role in provincialcompetitiveness and to identify possible areas forimprovement.

Demand for investment in infrastructure by localauthorities is growing very rapidly.The governmenthas been attempting to deal with this growingdemand, but the results so far have been marginal.This chapter presents baseline data on existinginfrastructure capabilities in each province to creategreater awareness of the opportunities fordeveloping and financing infrastructure.

Governance of infrastructure reaches beyond thejurisdiction of provincial leaders, as they aredependent on central ministries for the finaldeterminations about scope and quality. Recently,the Vietnamese government has come underincreased scrutiny regarding infrastructuredevelopment. In a recent National Assembly querysession, parliamentary delegates were sharply criticalof several aspects of economic governance relatedto infrastructure decisions, including: the haphazard,irrational, and inconsistent master planning linking

road networks, sea and air ports and relatedfacilities into a coherent transport hub; inadequateadministrative and technical capacity for transparentprocurement of private investment in public-purpose infrastructure; poor prioritization andallocation of limited budget resources; limitedfinancing through state budget and governmentbonds; and inadequate legal provisions to mobilizefinancing and expertise to develop infrastructure.38

In a high-profile report to the Prime Minister, ateam of researchers highlighted infrastructure asone of the two most important constraints (alongwith education) to Vietnam's development.Thereport stresses that the infrastructure deficienciesresult not from insufficient spending, but insteadfrom poor project selection, time and cost overruns,and a politically infused process that does notallocate infrastructure investments to the fast-growing regions that need it most.39 As a result,road, water, rail, and air transport networks all lagsignificantly behind Vietnam's major competitors;and the problem is most severe in Vietnam'sindustrial hub in the North Southeast.40

These are important arguments to which


INFRASTRUCTURE ANDINFORMATION ANDCOMMUNICATIONTECHNOLOGY

37. Harvard Fulbright Program. 2008.38. Xuan Toan. 2008. "Legislators urge tighter monitoring of public

projects." Thanh Nien News, November 6.www.thanhniennews.com/politics/?catid=1&newsid=43521

39. Abrami, Regina, Edmund Malesky, and Yu Zheng. 2008."Accountability and Inequality in Single Party Regimes: AComparative Analysis of Vietnam and China." Harvard BusinessSchool Working Papers 08-099 (http://www.eldis.org/go/country-profiles&region=22)

40. Frost and Sullivan. 2008. "Vietnam Transportation and Logistics -Challenges and Opportunities." APL Logistics, Singapore. AsianDevelopment Bank, 2008. "Vietnam," Asian Development Outlook2008 Update. ADB: Manila: 233-8.

policymakers should pay careful attention, but thekey decisions about infrastructure developmentchoices and financing must begin with a thoroughassessment of available data on actual quality andthe perception of the infrastructure's end-users.Thus far, data on infrastructure at the provinciallevel has not been organized or easy-to-understand.

FFiinnaanncciinngg llooccaall iinnffrraassttrruuccttuurree.. Correcting thisdeficiency is particularly important becauseVietnam's fiscal decentralization efforts, especiallyunder the 2002 Budget Law, have given provincialleaders increasingly more power over provincialexpenditures. Provinces, especially net-revenueproducers, are less dependent on centralgovernment permission to finance infrastructureimprovements than they have been in the past.Provinces can now finance infrastructuredevelopment directly in four ways. First, they canuse surplus revenue above their biannuallydetermined targets. Alternatively, they can mobilizecentral transfers that are not earmarked for specificexpenditures (although some provinces have limitedamounts of non-marked transfers).Third, they canuse pre-determined project-based expenditures,which account for between 80 and 90 percent ofprovincial allotments.41 Finally, they can use debtfinancing, providing they can find interestedinvestors.

The 2002 Budget Law allows provincialgovernments to finance up to 30 percent of theirinfrastructure investments using bonds. Further, thelaw applies a 100% infrastructure investment cap tothe amount of debt that Ha Noi and Ho Chi MinhCity can raise. Decree 78/2007/ND-CP givesprovinces the right to grant licenses directly forpublic-private partnership vehicles-such as build-operate-transfer, build-transfer-operate, and build-transfer-that lure private investors into theinfrastructure sector.42 Such projects no longer needthe permission of the Ministry of Planning andInvestment (MPI) or the Prime Minister, though MPIretains a right to approve and appraise and has setinvestment thresholds.

Despite these advances, financing for infrastructureby local authorities remains limited because debtmarkets are insufficiently developed.Vietnam'scurrent laws and regulations do not adequatelyallow for the use of modern project financingtechniques including project revenue bonds andsecuritization, or create sufficient confidence forprivate investment in infrastructure. As a result,provinces have limited capacity to finance large-scaleinfrastructure, such as ports, roads, railroads,airports, and other assets that affect thetransportation costs and investment returns forboth small and large enterprises.

Along with increased responsibilities, provinces needa better understanding of their needs relative to theirpeers so they can make appropriate cost-benefitanalyses. Such an understanding should improve theoperational capacity of local governments toefficiently plan and execute large-scale infrastructureenhancements. This chapter attempts to push suchanalyses forward by bringing data from the PCIsurvey to bear on the question of infrastructuredevelopment.The data presented below should beseen as a first step. Before we can document theimpact of infrastructure, we first need to measure it.This analysis presents a snapshot of the stock ofinfrastructure as it exists in 2008.We look at twotypes of infrastructure: bricks and mortarinfrastructure, as this serves as the backbone forVietnam's current manufacturing and export-ledgrowth economy, and information andcommunication technology (ICT) to gauge Vietnam'spreparation for the information age. More importantthan this year's assessment, however, will be theaccumulated data on this index that allow forassessing the effect of infrastructure improvementson business growth and development.

5.1 Core Elements of the PCIInfrastructure IndexThe PCI Infrastructure Index is divided into foursub-indices:

! Industrial zones and small and medium-sizedenterprise (SME) concentrations, measuring thecapacity and quality of local industrial zones;

! Road and transport, gauging the coverage ofroads in Vietnam and the indirect and directcosts of transport that result from them;


41. Kim N.B. Ninh and Vu Thanh Tu Anh. 2008. Decentralization inVietnam: Changes and Policy Implications for Sustainable EconomicGrowth. Ha Noi:Vietnam Competitiveness Initiative.

42. Decree 78/2007/N?-CP dated 11 May 2007.http://www.vir.com.vn/Client/VIR/index.asp?url=content.asp&doc=13545

! Utilities, measuring the costs and reliability oftelecommunications and energy delivery in theprovince; and

! Major infrastructure, which simply detailswhether a province has an airport orseaport within its borders, which wouldlower transport costs for businesses in thatprovince.

As in the PCI, each sub-index is a combination ofhard data from published sources and perceptionsdata gleaned from the 7,820 PCI respondents.Figure 5.1 details the final scores on theinfrastructure index, where the first panel includesthe first three sub-indices, and the second paneladds the measures of major infrastructure to themix.Table 5.1 provides source data and summarystatistics on the indicators used in each sub-index.


Figure 5.1. PCI Infrastructure Index


Table 5.1 Indicators Used In Infrastructure Index

1) Industrial ZoneQuality andCoverage

2) Road Quality andTransport Costs

3) Utilities (EnergyandTelecommunications)

Number of Industrial Zonesand Concentrations in Province

Percentage of total IZ surfacearea that currently has occupants.

Firm Rating of ProvinicalIndustrial Zone Quality (% VeryGood or Good).

Number of days annually thatroads are impassable due torainfall.*

Transport costs of a 40-footcontainer from provincial capitalto nearest major ports (HP,HCMC, DN) in Millions of VND*

Monetary loss annually fromspoiled and damaged products inthe past year (Millions of VND)*

Percentage of roads in province(national, provincial, or district)that are paved with asphalt

Hours of Telecommunicationsoutages in the province permonth*

Ministry of Planning andInvestment (August 2007)

Ministry of Planning andInvestment (August 2007)

PCI Survey Question E2

PCI Survey Question: E3

General Statistics Office


Average estimates bythree local transport

companies

PCI Survey Question: E1.14

Min

Median

Max

Min

Median

Max

Min

Median

Max

Min

Median

Max

Min

Median

Max

Min

Median

Max

Min

Median

Max

Min

Median

Max

Min

Median

Max

Min

Median

Max

0

1

23

0.00

30.13

93.55

3.07

23.87

72.89

3

7

19

14.6

31.9

83.1

2.1

6.5

16.0

1.98

51.28

100.00

7

13

50

75.00

94.18

100.00

1.2

11.7

113.1

Sub-Index Indicator Source 2008 Measures

Assessment oftelecommunications quality (%Good or Very Good)

Telephones (Land and Cellular)per 100,000 Citizens in 2007

PCI Survey Question E1.2

Ministry of Informationand Communications

Unlike the original PCI, the Infrastructure Index isnot a definitive evaluation of the quality of localleadership. Many of the indicators measured are outof the control of provincial authorities.

Much of the current infrastructure was completedlong before the tenure of today's current provincialleaders: some was built in the early central planningyears after independence and unification and someeven dates back to the 19th century underEmperor Minh Mang. Provincial officials cannot beheld responsible for the infrastructure stock theyinherited from a bygone era.

Second, firms in distant areas have higher transportcosts that are inherently difficult to overcome.Themountainous regions offer harsher and moreexpensive terrain for building major thoroughfaresthan provinces in the Red River Delta. ManyMekong Delta provinces use waterways as theirprimary means of transport, but because many ofthese waterways are quite small, transportcompanies cannot realize economies of scale frompacking products in large containers. Often productsmust be reloaded after passage on the MekongRiver before being sent on to the nearest port.

Third, many infrastructure decisions are made by

central government planners. Provinces cansupplement infrastructure spending out of theirown budgets as discussed above, but poorerprovinces do not have this option and must dependon central transfers.Winning some nationalinfrastructure monies is occasionally dependent onwho has better access to central officials or is ableto plead their case for central supplements morepersuasively. It makes little sense to reward orpunish provincial officials based on the success oftheir lobbying efforts.

Fourth linkages of infrastructure across provincialborders affect firm perceptions in ways that aredifficult to disentangle using provincial-level surveydata. It would be unfair to rank province A below itsneighbor, province B, simply because centralauthorities selected B as the sight of the nationalhighway. Alternatively, a province that has done agood job of marshalling local resources for newroads and maintenance may be downgraded byfirms that are forced to ship products outside thoseprovincial borders over the roads of neighbors thatmay not have been so diligent about infrastructuredevelopment and upkeep. Another example ofspillover is a province that invests in a port thatbenefits firms in an adjacent province more than itdoes some of its own entrepreneurs, who might be


Electricity Vietnam (EVN)


APL Logistics

APL Logistics

APL Logistics

APL Logistics

4) MajorInfrastructure(Ports/Airports)

Min

Median

Max

Min

Median

Max

Provinces with

Provinces w/o

Provinces with

Provinces w/o

Provinces with

Provinces w/o

Provinces with

Provinces w/o

595.51

776.17

27.00

44.00

101.00

5

59

25

39

4

60

15

49

Sub-Index Indicator Source 2008 Measures

Average cost per kilowat ofenergy in province (VND)

Hours of electricity outages inthe last month*

National Seaport (ContainerCargo > 34,000 TEU)

Local Seaport (ContainerCargo > 2,000 TEU)

International Airport

Domestic Airport

too far away to make use of it. It is not clear thatentrepreneurs have a perfect understanding ofwhere provincial borders are situated, and whichauthorities deserve praise or complaint for poorinfrastructure.We had hoped to use provincialbudgets on infrastructure expenditures tounderstand local-level commitment to improvingthe quality of its infrastructure, but such provincialbudgets are of such disparate quality that it isdifficult to draw firm conclusions.

Consequently, the Infrastructure Index is simply anassessment of total infrastructure quality.We makeno assumptions about credit or culpability, andpresent it simply as a tool to inform the investmentdecisions of local entrepreneurs and the policypriorities of central and local officials.

5.2 Industrial Zone QualitySub-index 1 measures the scale, usage, and overallquality of industrial zones (IZs) in each province.Tocalculate this, we first count the number of IZs,export processing zones, and industrialconcentrations (cuåm cöng nghiïåp) within eachprovince's borders. Dong Nai with 23 zones, HoChi Minh City (HCMC) and Binh Duong with 15,

and Long An with 8 have the most, while 14provinces have no zones at all. Because a provincecan build IZs that nobody uses, we supplement thenumber of zones with a measure of the percentageof IZ surface area that is currently occupied by afirm.This offers a useful indicator of whether theprovinces are building IZs that meet the needs ofbusiness, or creating zones for other reasons.Utilization rates are highest in Phu Yen at 93.5percent and Thanh Hoa at 89 percent.The third andfinal indicator is the assessment by individual firms inthe PCI on the quality of zones in their province.Firms were asked to assess not only large IZs, whichare primarily targeted at foreign-investedenterprises, but also the smaller industrialconcentrations that were built specifically to attractSMEs. Binh Duong at 72 percent and Vinh Phuc at63 percent had the highest number of firms rankingIZ quality as good or very good, while less than 8percent of firms in the mountainous provinces ofHa Giang, Cao Bang, Dien Bien, and Lang Sonreported that zones in their province were of highquality. Figure 5.2 provides a composite graphic withall three IZ indicators in a single view. Note thestrong correlations between the three measures,which are all significant at the .05 level.


Figure 5.2: Industrial Zone Quality in Province

Overall, Binh Duong, Dong Nai,Vinh Phuc, andHCMC ranked as the provinces with the best policyregarding IZs, while rural Cao Bang, Bac Lieu, andHa Giang has the weakest results.

5.3 Road Quality and Transport CostsSub-index 2 combines the quality of roadinfrastructure in each province with the direct andindirect costs of shipping goods on provincial roads.The first indicator is road coverage, the percentageof total roads (measured in kilometers squared) inthe province covered with asphalt.Total roadsincluded thoroughfares managed by three levels ofgovernment-national, provincial, and districtauthorities.The quality of district and provincialroads is probably the best measure of an individualprovince's contribution to infrastructure, but thismeasure does not fully capture how roads in aprovince affect business costs. Most entrepreneurscannot distinguish between the different types ofroad (district, provincial, national). All they careabout is the overall quality of roads, regardless ofwho manages them. It would make little differenceto them that national roads are excellent, if their

goods are damaged on provincial roadways.Moreover, the quality of local (provincial anddistrict) roads and national roads are interrelated. Aprovince that is fortunate to be chosen as thelocation of a major thoroughfare may not need tospend as much of its own resources.Thus, it is notuseful to consider the types of roads separately; wehave included all three road types together in asingle measure.

A 0.9 correlation between provincial and total roadquality can be seen in the top row of Figure 5.3, ascatterplot matrix, where there is a tight relationshipbetween district and provincial coverage with totalcoverage, but little relationship between thelocation of paved national roads and total roadcoverage.This is because in many provinces the onlypaved road within their border is the nationalhighway that crosses through the province.This isparticularly true in the Northwest and CentralHighlands, where over 40 percent of total pavedroad is accounted for by national highways. Lai Chauis a particular outlier even within this region with 65percent of its paved road within its borders comingfrom national roads.


Figure 5.3. Percentage of Road Covered with Asphalt, By Responsible Authority

Hai Phong and Nam Dinh province have roadcoverage of 100 percent, followed by Thai Binh(99.7), Ha Noi (95.7 percent), and Vinh Phuc (95.6percent). Noticeably, all of these provinces are inthe Red River Delta Region. In fact, 8 of the top 12provinces in road quality are Red River provinces.The region as a whole has 83 percent coverage.The lowest road quality is found in Ca Mau (2percent) and the mountainous provinces of TuyenQuang (13 percent),Thai Nguyen (13.6 percent),Son La (18.5 percent), and Phu Tho (19.1 percent).

The second hard indicator of infrastructure in theprovince was created by asking three logistics andshipping companies to provide the PCI researchteam with estimates for how much it would cost toship a 40-foot container from the provincial capitalto the major port in each region of the country. Allthree companies were careful to acknowledge thatmost of their shipping is from IZs and that theyrarely are asked to ship from the most ruralprovinces. If they ever do have to ship from deep inthe Mekong Delta or the Northwest, they generallypick up the product at specific loading zones. Howthe cargo gets to the container loading zones ishighly variable, often involving small boats in theMekong and even horses in the Northwest. As aresult, the estimates for provinces in these regionsare much less precise than for those closer to themain cities. Nevertheless, the estimates by theshipping companies provide an interesting measureof the transportation barriers faced by smallentrepreneurs in each province to sell goods tomajor domestic consuming areas or to export.

About 30 percent of the variance among provincesin reported transport costs is accounted for by thedistance it is from one of the three major ports.After all, most shipping companies charge by thekilometer. Our regression analysis reveals that eachadditional kilometer between a major port and theprovince of origin adds VND 10,000 in additionalshipping charges to the total bill. Controlling fordistance, each kilometer of paved road reducesoverall transport costs by just over VND 2 million.The remaining variance in transport costs isaccounted for by region-specific effects, such asmountainous terrain or dense waterways, as well asmeasurement error in the estimates of logisticsfirms (See Appendix 3).

Higher transport cost is only one type of expensecaused by poor infrastructure.The PCI survey data

capture two other important costs to businesses.First, firms were asked how many days in the yearthat provincial roads were rendered impassable byheavy rainfall-meaning that firms were unable toship.Waiting periods can be especially costly forfirms in food processing industries, because it canoften lead to spoilage. Firms in the average provincelost about seven days per year due to inadequateinfrastructure for bad weather; on the end of thespectrum, firms in Bac Lieu province lost almost 19full days in the past year.

A final cost for firms was the monetary cost ofbreakage or spoilage caused by poor infrastructurequality. Firms in Khanh Hoa suffered the leastdamage from breakage and spoilage, losing onlyabout VND 15 million or US$880 per firm indamaged product, while firms in mountainous SonLa had VND 83 million (or $4,900) in broken orspoiled shipments.

Considering all these issues, Ba Ria-Vung Tau(BRVT), Binh Duong, Da Nang, and Hung Yen areranked as the provinces with the lowest transportcosts in the country. Mountainous Lai Chau, DienBien, and Son La, due to the difficult terrain anddistance from urban centers, suffered from thehighest transport costs.

5.4 Utilities (Telecommunicationsand Energy)As Vietnam has transitioned away from anagriculture-based economy toward manufacturingand services, affordable telecommunications havebecome more vital to economic development.Telecommunications infrastructure connectsentrepreneurs to the wider world, allowing them toreach a larger spectrum of suppliers and potentialcustomers, while simultaneously expediting thespeed of their transactions. Electricity, as theHarvard researchers put, is "the oxygen of amodern economy."43 They move on to explain thatVietnam is on the verge of suffocation due to overinvestment in hydroelectricity, leading to costlyoutages and unpredictable supplies.

Although they represent separate dimensions ofinfrastructure, telecommunications and energy arecombined into the same sub-index because they


43. Harvard Fulbright School. 2008: 3.

are both utilities.The two sectors are dominated bylarge, state conglomerates-Vietnam Post andTelecommunications (VNPT) and Electricity ofVietnam (EVN). Compared to energy,telecommunications is a far more competitivesector, as a number of providers are available, evenin the most remote provinces, especially in terms ofproviding internet and cell-phone services. In fact,EVN was recently granted a license for cell-phoneservice.VNPT continues to have a monopoly onfixed-line telephony and in access to foreigncommunication markets, especially high-volumeinternational satellite transmission.The moreextensive competition in telecommunications mayhelp account for the higher satisfaction ratings itreceives from PCI respondents, compared toenergy.Vietnam's World Trade Organizationcommitment will allow a phase-in of foreign energyand telecom suppliers, which should drive downprices further. Nevertheless, foreign competition isstill a few years off.

Our hard data measure of telecommunicationservice is the total number of landline and cellulartelephones per 100,000 citizens in each province.Wesupplement this with a measure from the PCI surveythat captures the number of outages in phoneservice experienced by firms in the month beforethe survey. On average, firms experienced about 13outages in the month before the release of the PCIsurvey. Firms in Ben Tre experienced the least phoneoutages, with seven in a month, but this is still high byinternational standards. Binh Phuoc firms experiencedan astounding 50 outages in January 2008, about 1.6events per day. Finally, firms ranked the quality oftelecommunications in their province on a five-pointscale. In general, firms were quite pleased. Strikingly,only Phu Tho province had an approval rating fortelecommunications that was below 80 percent.

Energy reliability and cost have always beenimportant for producers, but have become moresalient in the past two years due to rising oil andgas prices. Local newspapers have been filled withstories of businesses closing their doors because therising cost of energy had put them in the red.Weuse two measures to capture the provision ofenergy-one a measure of reliable energy provisionand the other a measure of the cost of energy.

As with telecommunications, firms were asked tolist the number of power outages they experiencedin the month before the survey. Answers ranged

from a best of 27 outages in Da Nang, which isconsiderable in itself, to an exceedingly high 101outages in Ha Tay province.

To supplement this measure of reliability, theaverage cost of energy was calculated using datafrom EVN for each province. Prices are quitecomplicated, differing by the end user. For instance,households pay less than businesses. In addition,there are discounts built in for development andpolicy purposes. Consequently, we used the averageenergy price across end-users within a province.44

Prices were highest in the three major cities:HCMC, at VND 1,068 per kW; Ha Noi, at VND996 per kW; and Da Nang, at VND 912 per kW.The lowest energy costs were all in rural regions,indicating that energy in rural, poor provinces aresubsidized relative to urban areas. Energy costswere VND 595 per kW in Bac Giang and VND 616per kW in Ha Tinh. According to EVN data, energyprices have increased on average by about 45percent in the past year, and will continue to evolveas global energy markets fluctuate.

5.5 Major InfrastructureThe final measure of infrastructure represents theexistence of major sea and air ports in a province.Provinces receive a point if they have aninternational airport, a domestic airport, a majorseaport (capable of handling more than 34,000 TEUor 2 million tons of cargo), or a minor seaport usedprimarily for local shipping.

The median province in Vietnam receives a zero onmajor infrastructure, because it does not have an airor sea port.Three national-level cities (HCMC, HaiPhong, and Da Nang) receive scores of 4, followedclosely by Binh Dinh, which receives a 3 due to itsactive, deep-water port and domestic airport.

We include this measure with some trepidation. First,the location of major infrastructure is primarily theresult of central-level planning. Not many provinceshave the resources to self-finance their own port, norshould they. It makes little sense for neighboringprovinces to each have their own seaport, or domesticairport for that matter. In these cases, rational regionalplanners should allocate infrastructure in the mannermost efficient to the region and country as a whole.


44. http://www.hanoipc.evn.com.vn/EVNShow/tintuc1.asp?InforID=11847&CategoryID=881&Pos=893&rCount=2

Second, central infrastructure decisions may respond toeconomic growth in some localities rather than causeit. For instance, it may make sense to open a newairport in a province that has attracted a great deal ofinvestment to meet the needs of existing investors.

Nevertheless, firms do benefit from the lower costsand new opportunities that result from proximity toports or airports, so in the spirit of exhaustivenesswe have included it. Readers should feel free to usethe dimensions of infrastructure that best fit theirneeds and should not feel limited to our method ofaggregation.

5.6. The PCI Infrastructure Indexand Economic GovernanceAs Figure 5.4 illustrates, infrastructure (including thefour sub-indices) and economic governance areclosely correlated, making it difficult to disentangletheir effects.The bivariate correlation between theunweighted PCI and the final infrastructure is about0.6, which is significant at the .01 level. A number offactors can explain this relationship:

! Proactive provinces that are interested inprivate-sector development are likely to invest

resources in both providing betterinfrastructure and improving the regulatoryenvironment.There is no reason why theymust necessarily choose one over another.

! Good governance can engender betterinfrastructure. Fast-growing provinces havemore resources to spend on infrastructure. Ifgood governance helps a province like BinhDinh develop a robust private sector thatgenerates healthy provincial revenue, it willhave more money to spend on infrastructure.Alternatively, the central government mayallocate money for roads or ports toaccommodate the needs of fast-growingclusters of firms.

! Infrastructure attracts foreign investors whointeract with policy makers, providing theminsights into regulatory reform that benefit notonly foreign firms, but also private firms.

! Poorer provinces are more dependent oncentral transfers for infrastructure and aretherefore less able to develop independentgovernance institutions of their own.They aremore likely to invest their scarce resources inappealing for greater central transfers.


Figure 5.4: Relationship Between Infrastructure and Governance

Even given this correlation between economicgovernance and infrastructure, Figure 5.4 separatesout provinces with four combinations of governanceand infrastructure attributes.The dotted lines on thescatterplot indicate the median scores for theinfrastructure and unweighted PCI index.Thenortheast quadrant identifies provinces with above-average governance and infrastructure provision,which represent the best locations for investors.Provinces in the southeast quadrant offer above-average infrastructure but less-than-friendly businessenvironments.Those in the northwest quadrant arehandicapped by poor infrastructure, but offset thisimpediment with above-average governance. Finally,provinces in the southeast quadrant are the mostproblematic, with poor infrastructure and pooreconomic governance. As infrastructure is a keylong-term development issue, the provinces in thislast quadrant could benefit the most from initiatinga virtuous circle of economic development byimproving their regulatory environment, whichcould lead to higher revenue or investors in bondsor build-operate-transfer, and consequentlyadditional infrastructure spending.

5.7 Infrastructure as a Constrainton Fast Growing RegionsThe finding that infrastructure is highly correlatedwith gross domestic product (GDP) is useful, butuninteresting. As we mentioned above, the legacy ofinfrastructure stocks goes back several generations.We demonstrate in every PCI report that thetelecommunications capacity in 1995 is stronglycorrelated with economic outcomes today. Causalityis difficult to ascertain, because infrastructure andeconomic performance are mutually reinforcing.Initial infrastructure was among the primarydeterminants of investment in the early years ofVietnam reform era.45 High investment at that timeexpanded provincial revenue and allowed betterprovinces to invest more in infrastructure, therebyincreasing the infrastructure gap.

Furthermore, equivalent infrastructure acrossprovinces is unwanted. Infrastructure decisions needto be national-level policy decisions. Projects shouldlead to coordination across provincial boundaries.Local authorities should not be encouraged toinvest in duplicative projects. Most importantly,major infrastructure investments should go wherethey are most needed from a national developmentperspective.The most common criticism fromeconomic analysts and foreign investors interestedin Vietnam is not that the infrastructure in thecountry is unequal; it is quite the opposite: criticsassert that Vietnam's policy apparatus directsinfrastructure on a political basis, using it as aredistributive tool. As a result, the high growth areasof Vietnam are disproportionately neglected incentral infrastructure schemes.They may havehigher levels of infrastructure on an absolute basis,but improvements in infrastructure are not keepingpace with the demands on them. As the authors ofChoosing Success put it, "Provinces with slowpopulation growth receive "gold plated"infrastructure-broad high speed roads and bridges-while infrastructure in the Southeast focal economiczone is stretched to the breaking point."46

Indeed, using PCI data we can confirm that thisappears to be true. Road coverage is fairly high in BaRia-Vung Tau (83 percent), but the four otherprovinces that comprise Vietnam's growth hub-BinhDuong (64 percent), HCMC (56 percent), Dong Nai(52 percent), and Long An (29 percent)-rank in themiddle of the pack or worse. Moreover, within theseprovinces, most of the paved roads are locallymanaged and do not benefit from spillovers fromnational highways as much as other localities. Nationalhighways account for 22 percent of total paved roadcoverage in the average province. In the five growthcenters cited above, they account for only 12 percent.

Has growth in road coverage kept pace withdemands on its usage? The key issue in ChoosingSuccess is not that the roads are generally bad;rather, it is that funding for infrastructure has beenallocated to the wrong places. As noted above, thenumber of paved kilometers of road throughoutVietnam has grown at an annualized rate of 15


45. Malesky, Edmund. 2008. "Straight Ahead on Red: How ForeignDirect Investment Empowers Provincial Leaders." Journal of Politics70(1): 97-119(http://journals.cambridge.org/download.php?file=%2FJOP%2FJOP70_01%2FS0022381607080085a.pdf&code=ff21ae7c9a9eb3d901f7ea49a2a202a6) 46. Harvard Fulbright School. 2008: 25

percent, compared to a 26 percent rate in thegrowth of private firms. In the North Southeast,which is the engine of the Vietnamese economy,growth in asphalted roads has averaged only 9.6percent, despite the fact that the region accountsfor 36 percent of GDP and 40 percent of privateeconomic activity.

Table 5.2 demonstrates the issue succinctly by usingpopulation growth as a proxy for the change indemands on infrastructure due to increasedurbanization, economic activity, and labor migration.This is the indicator highlighted by the ChoosingSuccess authors when speaking of the NorthSoutheast, because of the rapid growth inpopulation relative to other areas, despite relatively

constant fertility rates in the country.47 The tableshows that the population of the North Southeasthas been growing much faster than the rest of thecountry, but the change in asphalted roads lagsbehind every region except the Red River Deltaeven though the Red River Delta already has thehighest coverage in the country by quite somemargin. Average coverage in the North Southeastalso lags behind the both the North and SouthCentral Coast. More interesting still, Ba Ria-Vung Tauand Dong Nai have experienced road coveragedeclines, despite above-average population growth.At least on the measure of road coverage, it doesseem that high-growth provinces are more heavilyconstrained by infrastructure.


Entire Country

Red River Delta

Northern Uplands

North Central Coast

South Central Coast

Central Higlands

North Southeast

Mekong Delta

Ba Ria - Vung Tau

Binh Duong

Dong Nai

HCMC

Long An

1.4%

1.1%

0.1%

0.8%

1.1%

1.6%

2.3%

0.2%

2.0%

6.0%

2.0%

3.0%

1.0%

100%

20.9%

8.3%

6.1%

7.0%

4.2%

36.4%

17.2%

7.7%

1.7%

4.0%

18.3%

1.5%

100%

27.5%

6.0%

5.0%

8.5%

2.9%

39.0%

11.0%

1.1%

2.3%

2.4%

29.7%

1.3%

15.2%

8.7%

17.1%

52.7%

18.7%

10.3%

9.7%

19.9%

-5.6%

10.0%

-3.0%

1.9%

18.7%

42.0%

83.5%

38.1%

55.8%

60.6%

49.4%

52.0%

40.7%

82.5%

64.5%

52.2%

55.9%

28.5%

777.0

734.0

766.3

687.6

789.9

779.8

865.8

791.7

860.7

871.9

865.1

1068.1

813.9

PopulationGrowth

(2002-2006)

Share ofGDP

(2007)

Share ofActivePrivateFirms(2007)

Growth inAsphaltedKilometers

of Road(2002-2006)

Percentage ofTotal

Kilometers ofRoad Covered

by Asphalt(2006)

AveragePrice ofEnergy

(VND/KW)

Table 5.2: Infrastructure Growth Versus Demand for Usage

47. Harvard Fulbright School. 2008: 3


It is not just roads.The same provinces are alsopaying more for electricity than other regions inthe country. While the average provinces paysabout VND 777 per kW, the fast-growingprovinces of the North Southeast are all payingwell over VND 800 per kW.This indirect transfersystem could potentially burden development inhigh growth areas.

Vietnamese national leaders must work out theirinfrastructure priorities and set out a national planthat takes into account redistributive needs, whilenot handcuffing the high-growth centers. In addition,provinces should be allowed to take moreadvantage of the privileges granted to them underdecentralization, so they can use debt markets orpublic-private partnerships to finance infrastructureon their own accord.

5.8. Information andCommunication TechnologyIn addition to the traditional types of infrastructurediscussed above, more and more firms in Vietnamrecognize the utility of information technology forfacilitating interactions with business partners andacquiring market and technology information. In fact,27 percent of respondents recorded active e-mailaddresses on their PCI survey forms. In the two majorcities of Ha Noi and HCMC, more than 60 percentof firms used e-mail addresses for their contactinformation.Annual growth in ICT penetration inVietnam is estimated at 8 percent per year. 48

The 2008 PCI attempted to incorporate an ICTsub-index into the infrastructure index. It becameclear that the Office of the Steering Committee forNational Information Technology Readiness of theMinistry of Information and Communications hadalready developed all the information that would beneeded for such an index.49 Not only did this officehave the best repository of data available on ICT inthe country, but also, together with the Associationfor the Study of ICT in Vietnam, they created anannual index of provincial readiness for ICT.The ICTReport, however, is available only in Vietnamese andhas not been widely circulated among researchers

and the public.The government created it primarilyas an internal policy document. Nevertheless, bothits methodology and data are of excellent quality.Rather than attempting to create a rival index, webase our analysis on ICT capabilities on this strong,existing report.

The ICT Index is subdivided into five sub-indices,which are described along with their indicators inTable 5.3. Sub-index 1 is a measure of the level ofinfrastructure necessary to promote ICT. It includesbasic infrastructure such as telecommunicationscapacity, but also provides the number of householdand business computer users and internetsubscribers in each province. Most importantly, itcalculates overall capacity, including domestic andinternational bandwidth, as well as broadbandservices. As with the measure of infrastructureabove, sub-index 1 is endogenous to economicdevelopment. Richer provinces and citizens are likelyto have more resources to spend on ICTequipment. A correlation with development,therefore, does not necessarily indicate that ICTgenerates growth, since the opposite relationship ispossible as well. Consequently, it is not surprisingthat the four highest-ranked provinces are Ha Noi,HCMC, Da Nang, and BRVT.The five localities withthe worse ICT scores are all in the poor andmountainous Northern Uplands (Cao Bang, Son La,Bac Giang, Hoa Binh, and Ha Giang).

48 Le Truong Tung. 2007.Vietnam ICT Outlook 2007. Presented at theVietnam Computer Electronics World Expo, July.

49. Le Truong Tung. 2007. Vùn Phoâng Ban Chó Àaåo Quöëc Gia Vïì CNTT[Office of the Steering Committee for National InformationTechnology Readiness] and Höåi Tin Hoåc Viïåt Nam [Association ofInformatik Studies]. 2006 ."Baáo Caáo Chó Söë Sùén Saâng ChoPhaátTriïín Vaâ ûáng Duång CNTT-TT Viïåt Nam Nùm 2006 [Reporton the Index for the Develpment and Usage of Information andCommincation Technology in Vietnam 2006].


1) ICTInfrastructure(Level ofhardware andphysicalequipment fortechnologyusage)

2) ICT HumanResources(Schooling,training, andpersonnelrequired fortechnology skillacquisition)

a. Land line telephones/100 peopleb. Mobile telephones/100 peoplec. Dial-up internet accounts/100 peopled. Broadband internet subscriptions/1000 peoplee. Domestic bandwidth (traffic on Vietnamese

websites)/per capitaf. International bandwidth (traffic on international

websites)/per capitag.g. Households with a computerh. Households with broadband interneti. Computers/per capita in state officesj. Computers with internet in state officesk. Computers with broadband internet in state officesl. Businesses with broadband internet

a. Primary schools with information technology (IT) coursesb. Middle schools with IT coursesc. High schools with IT coursesd. State workers with computer skillse. Educational institutes (higher than high school) with IT

courses/100000 peoplef. Graduates from above institutes/1000 peopleg. Official training centers with IT courses/100000 peopleh. Graduates from above centers/1000 people

Min 0.04

Min 0.14

Median 0.21

Median 0.31

Max 0.7

Max 0.65

Sub-Index Sub-Index 2008 Measures

Table 5.3: Indicators Used in ICT Index


a. Total time of domestic phone calls/per capitab. Total time of international and long distance calls/per

capitac. State budget for ICT/per capitad. Small administrations and districts with websitese.e. Businesses with websitesf. Provinces with websites or information portal and

basic information available (including: a generalintroduction, administration structure, news andevents, activities of the local administration,instructions for individuals and businesses to completeadministrative procedures, legal documents,development plans and policies, business information,questions and answers on ICT policies, searchfunction, visitor comments, frequently asked questions(FAQ), and discussion).

g. Public services offered on the internet (including:business registration, investment licensing, constructionpermission, payment methods for utility bills, privateregistration for pharmaceuticals, job advertisements,public transport services, queries and status ofadministration procedures).

a. Number of ICT businesses/1000 peopleb. Total revenue of ICT businesses/per capita

a. Number of ICT steering committees and number ofmeetings of these committees

b. ICT policies for development: strategy and plans,encouragement, human resources development

c. Interest of local state leader/officals in ICTdevelopment and application: no interest, littleinterest, some interest, interested, very interested

3) ICTApplications(Scope, scale,and efficiencyof uses ofinformationtechnology onexistinginfrastructure).

4) Resources ofPrivate ICTProviders

5) Organizationand Regulationof ICT (Localpolicies,administrativestructure, andofficialregulationsgoverninginternet usage)

Min 0

Min 0

Median 0.17

Median 0.09

Max 0.67

Max

Min

Median

Max

0.11

0.705

1

0.89

Sub-Index Sub-Index 2008 Measures


The second ICT measure looks beyond thehardware of telephone lines and computer usage,and gauges social and human resource capacity-thereadiness of the local populace to take advantage ofinternet expansion.This indicator analyzes ICTtraining at every level of education, from primaryschool through university education, as well as thenumber of training centers and graduates fromthose centers in each province. Ultimately, the indexhelps us understand where ICT infrastructure islikely to be used most efficiently. Ha Noi tops thehuman resources rankings, followed, surprisingly, byThai Nguyen province. HCMC, Binh Duong, andVinh Phuc are the third to fifth strongest provincesin terms of ICT human resources. Son La in the farnorth of Vietnam and Ca Mau in the far south, onthe other hand, have the lowest ranked humanresources capacity for internet usage.

Sub-index 3 examines the scope and efficiency ofICT applications that are already in operation in theprovince.These include measures oftelecommunication usage, provincial expenditureson ICT, local businesses with websites, the quality ofinformation available on administrative websites, andthe quantity of e-governance services available onprovincial websites.The final two indicators arequite similar to the indicators of webpage opennessfrom the PCI Transparency sub-index, except thatthey also include information and services that areuseful to the public at large, such as options forpaying utility bills, filing driver's-license applications,and registering for pharmaceuticals. In fact, the twovariables are correlated at a 0.6 level (See Figure5.6). HCMC, Ha Noi,TT-Hue, Da Nang, and BinhDuong have the strongest internet applications,while Dak Nong and Yen Bai have the lowestpenetration of internet applications.

Figure 5.5: Relationship Between ICT Infrastructure and Email Usage in PCI

As a test of the external validity of the ICT index,Figure 5.5 shows the bivariate correlation betweenICT infrastructure and the percentage of PCIrespondents with e-mail addresses in each province.

The two variables are almost perfectly correlated,giving us a clear indication that the ICT index is auseful gauge of internet readiness.


Sub-index 4 tracks the number and total sales ofICT-related businesses in the province, as a way ofgauging both the receptiveness of local officialstoward internet oriented-businesses and the privateservices available to citizens and entrepreneurs.Once again, this figure is affected by economicperformance, as ICT-related businesses are morelikely to operate in locations where there is demandfor their services. HCMC, predictably, has thestrongest ICT business capabilities, but there aresome surprises at the top. Kon Tum and Bac Ninhhave a high proportion of ICT-related businesses,given the small size of their businesses communities.The two provinces rank second and third, above HaNoi and BRVT. Similar surprises are evident at thebottom.Tien Giang and Long An, two highlybusiness-friendly provinces, have no ICT businessesregistered at all.

Finally, sub-index 5 measures the commitment ofprovincial authorities to develop the ICTinfrastructure. It tracks the number of ICT steeringcommittees within local administrative units and the

number of meetings held by those units in order toassess their commitment to the task. It analyzesprovincial planning documents to see whether theyspecifically indentify strategies for developing localICT infrastructure and human resources. Finally, ituses the results of a survey of citizens to analyzetheir perspectives on the interest of local officials indeveloping ICT in their province. Eleven provinces,including all of the provinces ranked 'excellent' inthe PCI index, received perfect scores on thismeasure.

Figure 5.7 shows the final ICT index ranking. HCMCand Ha Noi, with their deep and sophisticatedmarkets, are rated the provinces with the best ICTcapabilities, followed by a familiar group ofprovinces, including Binh Duong, Da Nang,TT-Hue,and Vinh Phuc.While there are some notablesurprises, strong correlations exist between the PCI,the Infrastructure index, and the ICT index.Table 5.4shows these correlations and establishes bivariatecorrelations between these indices and severalimportant outcome variables.

Figure 5.6: Relationship between ICT Application and the PCI Webpage Score


Figure 5.7: Final ICT Readiness Index

5.9 ConclusionsAs we indicated in last year's report, there isevidence that increasing inequality is beginning toappear across the country.Top performing provincesexcel at all aspects of economic development, asshown in this report in terms of economicgovernance, infrastructure, and ICT capability.Thisgroup is pulling away from the rest of the country.At the same time, another group that must strugglewith weak initial conditions and poor infrastructure

has not been able to develop the good governancepractices to compensate for their handicap.Theseare steadily falling behind the performance of theirpeers.The extent of this divergence can be seen inFigure 5.8, a total investment environment index,created by standardizing the PCI, infrastructure, andICT indices to a common 100-point scale andadding them up. More than 100-points separate theurban and industrial centers at the top of the indexfrom the rural mountainous regions at the bottom.


Final Weighted PCI

Unweighted PCI

Standard Infrastructure

Full Infrastructure Index

ICT Index

Enterprises per Capita (ln)

Investment per Capita (ln)

Profit per Enterprise

GDP per Capita (ln)

* Significant at the .05 level

ln: Natural Log

1

0.9785*

0.6186*

0.5657*

0.3715*

0.3943*

0.4219*

0.3191*

0.3527*

1

0.5983*

0.5186*

0.2794*

0.3335*

0.3851*

0.3273*

0.3197*

1

0.8533*

0.4923*

0.4372*

0.4946*

0.1859

0.5102*

1

0.6237*

0.6757*

0.6534*

0.1662

0.5806*

1

0.7794*

0.7223*

0.0455

0.4763*

1

0.8534*

0.1141

0.6784*

1

0.2121*

0.6166*

1

0.2916* 1

Table 5.4: Bivariate Correlations of Final Indices and Outcome Variables

Figure 5.8:Total Investment Environment


CHAPTER SIX

DETAILED DESCRIPTION OFINDIVIDUAL SUB-INDICES

6.1 Entry CostsThe intellectual origin for this section of the reportarose from the World Bank's surveys of start-upcosts for entrepreneurs in developing50 andtransition51 countries.The goal of this sub-index is toassess the differences in entry costs for new firmsacross provinces. According to the two versions ofthe Enterprise Law (2000 and 2005), as well as thePrime Minister's Decision 236 regarding SMEdevelopment, these procedures should now beuniform across all provinces, but studies of businessregistration indicate they are not.52 The PCImeasures the extent of the variance in entry costswith eight variables (see Table 6.1 and Figure 6.1).

! Length of business registration in days. Accordingto the new one-stop shot procedures cited inJoint-Ministerial Circular 05-08, provincialdepartments of planning and investment arerequired to complete registration proceduresin five days or less.

! Length of business re-registration in days. For avariety of reasons, registered business mayneed to re-register their operations. Suchmotivations may include changing the legalform of the company, increasing the registeredcapital amount, or changing the product line inwhich a firm operates. Such a re-registrationshould take a maximum of five days accordingto Decree 88/2008/ND-CP.

! Number of business licensing requirementsrequired to operate. Article 6 of the 2000Enterprise Law provided the legal foundationfor removing hundreds of irrelevant businesslicensing requirements.This was followed by aseries of Prime Ministerial decisions anddecrees that removed these licenses by name,beginning with the 84 licenses articulated in

the well-known Decision No. 19/2000/QD-TTg.53 Unfortunately, many line ministries andProvincial People's Committees have creatednew license requirements (more more than300) since the Enterprise Law was enacted.Hidden licenses in the form of written"approvals" have also emerged, contributing tolengthier and more expensive entry costs forprivate firms.54 The PCI attempts to capturethe variance in licenses across the country byasking: "How many registrations, licenses(environmental, labor, natural resourceexploitation, etc.), and permits does your firmpresently need to operate (Please count allthe licenses, permits, stamps, etc. issued bydifferent agencies, even if they deal with thesame type of activity.)?"

! Percentage of firms having difficulty obtaining all thenecessary documentation to start a business.Simply asking about the number of requireddocuments is limiting, as it does not adequatelycapture the fact that only one extra license canbe a significant burden if it takes an inordinately


DETAILED DESCRIPTION OFINDIVIDUAL SUB-INDICES

50. See Doing Business in 2009 at http://www.doingbusiness.org/ formore information on the survey and data set.

51. See The Business Environment and Enterprise Performance Survey(BEEPs) at http://info.worldbank.org/governance/beeps/

52. Le Dang Doanh.Tinh Hinh Thuc Hien Luat Doanh Nghiep(Implementation Situation of the Enterprise Law), paper presentedat the Vietnam Consultative Group-Private Sector Forum, Ha Noi,2000. Central Institute for Economic Management (CIEM), 2001."One Year Enforcement of the Enterprise Law: Results andRemaining Problems," unpublished mimeo. Central Institute forEconomic Management (CIEM),Task Force for Enterprise LawEndorsement, 2003. "Assessment Report on Three Years of theImplementation of the Law on Enterprises," Vietnam BusinessForum Mid-Year Consultative Group Meeting, Ha Noi:World Bank,IFC, and Ministry of Planning and Investment, June.

53. CIEM and German Technical Cooperation Agency (GTZ). 2006. "6Years of Implementing the Enterprise Law: Issues and LessonsLearnt," Business Issues 05, Ha Noi, p. 26.

54. Ibid, p. 27. Prime Ministers Research Commission (PMRC). GTZ,and Asian Development Bank. 2006. "Business Licensing: CurrentStatus and the Ways Forwards." Business Issues 04, Ha Noi.

long time to obtain or requires several additionalunexpected return visits.

! Percentage of firms waiting more than a monthto complete all steps necessary to start theirbusiness. We gauge bureaucratic entry barriersby including an indicator of the percentage offirms in each province waiting over a month toreceive all the remaining licenses and completeall the steps necessary to begin businessactivities since the promulgation of theEnterprise Law.

! Percentage of firms waiting more than threemonths to complete all steps necessary to starttheir business. This indicator captures provinceswith particularly slow environments, wherefirms' investment plans were held up for anentire fiscal quarter.

! Length of wait for land to begin businessactivities. A number of research projects haveidentified obtaining land for business start upas one of the most critical obstacles faced byfirms throughout Vietnam, even under theLand Law of 2004.55 A recent InternationalFinance Corporation study determined that toset up a factory in Bac Ninh and TT-Hueprovince, it would require 234 procedures, 48documents, and a 399-day wait for land.56

Unfortunately, calculating the total amount oftime, beginning with the initial communicationwith local authorities to the date the land wasfinally accessed leads to unfair comparisonsacross provinces. Some provinces take a longertime to introduce land to firms because thereis simply too little to go around for the firms

that need it. Under these conditions, earnestefforts to locate land for a firm would becounted as a longer land wait.To compensatefor this issue, our indicator simply focuses onthe sum of two baskets of bureaucraticprocedures that are unrelated to landconstraints.57

- From the start of negotiations with the originalowners until completion, how long did it takethe business to complete the purchase (indays)? This measure gauges the arbitrationefforts provided by the province to ease landnegotiation and compensation.We includethis measure because the 2003 Land Lawspecifically states that enterprises mustcooperate with Provincial People'sCommittees in negotiating compensationwith local residents.The People's Committeecan often complicate the negotiations whenmembers allow its provincial socio-economicgoals, such as employment generation, toinfluence the discussions.

- From the day the business submitted itsapplication, how long did it take to receive theLand Use Rights Certificate (LURC) (in days)?Thisis a simple measure of the ability of theprovincial apparatus to formalize the land userights of a firm after the land has beenobtained. Often this involves coordinationbetween multiple agencies at both the districtand province level including the People'sCommittee, the Land Authority, and theDepartment of Construction to sign off on theprocedure, thereby adding to waiting periods.


57. In the 2005 PCI, the available land constraint issue was resolved bydividing the land-use wait by the number of private firms per 1,000.This quantity was then multiplied by the ratio of land in theprovince currently zoned for business purposes. In essence,effective land use wait = (land wait/enterprises per1,000)*(percentage of business land/100).This year, the survey wasredesigned to capture the different time costs available, so such aformula was unnecessary.

55. CIEM and GTZ. 2005. "From Business Idea to Reality: Still a Longand Costly Journey." Business Issues 04, Ha Noi, 15. Carlier, Amanda,and Son Thanh Tra. 2004. "Firm Dynamism: Beyond Registration:How are Vietnam's new domestic firms faring?" Vietnam PrivateSector Development Policy Note, Ha Noi:World Bank.

56. Foreign Investment Advisory Service and International FinanceCorporation. 2008. Improving Land Access in Vietnam: Study of BacNinh and Thue Thien Hue. Ha Noi:World Bank Group.VietnamInvestment Review. 2008. Land Use Ruling Closes Gates on Firms,August 29. CIEM and GTZ, 2005, 21.

* Significant at 5% Level; NA = Not ApplicableAll values are at the provincial-level.Data includes only firms registered within two calendar years preceding the survey.2005 data only include 42 provinces and do not include the full set of indicators used in subsequent years, reflecting changes in survey questions andordering in 2006.


PCI Survey Question: C4






Min

Median

Max

Correlationw/Previous Year

Min

Median

Max


Min

Median

MaxCorrelationw/Previous Year

Min

Median

Max


Min

Median

Max


Min

Median

Max


Min

Median

Max


9.76

33.33

63.41

NA

0

5.9

21.95

3.23

25.81

44

0.24

0

5.78

25.64

0.02

0

12.4

27.27

NA

12

20

58

NA

6

10

35

NA

2

4

7.5

NA

65

231

1318

NA

5.18

27.21

53.8

0.26*

0

6.78

27.27

0.15

0

11.1

36.8

0.30*

7

15

22.5

0.27*

3

7

15

0.24*

1

2.5

5

0.15

45

90

450

0.05

6.67

21.91

39.13

0.15

0

5.72

16

0.18

2.08

10.05

26.92

0.1

5

12.25

15

0.46*

3

7

10

0.53*

1

2

4

0.31

42.5

81

195

0.45*

Indicator Source Measure 2005 2006 2007 2008

Percentage of firms waitingover a month to completeall steps necessary to startoperations.

Percentage of firms waitingover three months tocomplete all stepsnecessary to startoperations.

Percentage of firms havingdifficulty obtaining alllicenses and permitsnecessary to do business.

Length of businessregistration in days(Median).

Length of business re-registration in days(Median).

Number of licenses andpermits necessary to startoperations (Median).

Table 6.1: Comparison of Entry Costs Sub-Index (2005-2008)

PCI Survey Questions:B4.1.2 (Wait for Land

Use Rights Certificate) +B4.1.3 (Negotiations with

Holders beforeConversion)

Effective Wait for BusinessPremises (Median).


Figure 6.1: Entry Costs Sub-Index

Original Indicators are those used in the original 2005 survey, while New Indicators are those first used in 2006. Both dimensions account for 50% ofthe final index.

6.2 Land Access and Security ofTenureLand issues can be divided into two dimensions.First, the PCI considers access to business premisesbecause the inability to locate and obtain productiveland not only reduces opportunities for investment ina new business, but also limits access to capitalbecause firms cannot use land-use rights as collateralfor bank loans.58 A second dimension of land policy,however, is the security of tenure on a particularproperty: do firms feel that their land rights arecertain over the long term? The more secure thetenure, the more firms will be emboldened to investin the long-term productivity of their premises.59 Ifexpropriation or materially damaging changes in leasecontracts are a possibility, firms will take a short-termoutlook with their investment and business plans.Employing the latter approach tends to undermineoverall provincial welfare in terms of income andemployment creation.60

6.2.1. Land AccessFour indicators in determining land access follow:

! The percentage of firms with LURCs or in theprocess of receiving them. While technically allVietnamese land belongs to the state, therights to its use have been assigned toindividuals and firms through LURCs since1993.These certificates legalize their owners'rights to the long-term use of the allocated

land (for as little as 20, but up to 70 years)and to transfer, exchange, lease, inherit, andmortgage the land-use right. Particularlyimportant is the ability to use a formal LURCas collateral in accessing bank loans.Thepercentage of private firms that do havesecure land use rights varies considerablyacross provinces. Many provinces have myriadfirms with informal land rights inherited fromprevious generations or purchased throughinformal exchange.Thus, we measure the totalnumber of respondents with LURCs.

! Percentage of firms that feel land availabilityconstrains their business expansion. The secondindicator are whether firms feel investment hasbeen sacrificed due to the lack of availableland. Firms are asked whether they havelimited their expansion plans due to theunavailability of land.

! Provincial land conversion policies. Provincial landconversion policies must be thought of in twoperiods.The first is provincial zoning policiesbefore a firm takes possession of the plot. Anumber of provinces have worked hard toconvert agricultural land for manufacturing usein order to increase the supply of land, therebylowering waiting periods and prices.61 Thesecond type of land conversion policy takesplace after a firm has assumed possession ofagricultural land and then wants to use it forcommercial purposes. In these cases, provincesare obligated by law to respond to a firmpetition for land conversion within 20 days. Ifapproved and the firm pays all the appropriatefees, the Provincial People's Committee mustissue an amended LURC within five days.ThePCI uses the percentage of firms in eachprovince that rated land conversion policies asgood or very good.

! Percentage of total land in province with anLURC according to the Ministry of Natural


58. De Soto, Hernando. 2000.The Mystery of Capital. New York: BasicBooks.

59. De Soto, Hernando. 2000; Field, Erica, 2007. "Entitled to Work:Urban Tenure Security and Labor Supply in Peru." The QuarterlyJournal of Economics,Vol. 122, No. 4 (November), 1561-1602; DiTella, Rafael, Sebastian Galiani, and Ernesto Schargrodsky, 2007. "TheFormation of Beliefs: Evidence from th Allocation of Land Titles toSquatters." The Quarterly Journal of Economics,Vol. 122, No. 1(February).

60. Knack, Stephen, and Phillip Keefer. 1995. "Institutions and EconomicPerformance: Cross-Country Testing Using Alternative InstitutionalMeasures," Economics and Politics 7(3), 207-228. 61. Cung et al.

Resources and the Environment. This data, basedon the proportion of all land in the provincethat has been formally granted an LURC isused as hard data to anchor firm perceptionsof land access.The measure includesagriculture, household, manufacturing, urban,military, and religious land.

6.2.2. Security of Land TenureFour indicators for determining the security of landtenure follow:

! Risk of expropriation. As noted above, once aninvestor has land, the fundamental determinantof the effort and expense they put intodeveloping their investment is directly relatedto their fears that the land on which their firmsits could be taken from them. Indeed, justover half of all LURC holders in the sample offirms believe that the risk of expropriation ishigh or very high. For this indicator, we takethe mean score on a five-point scale, wherefive represents the very low risk ofexpropriation.

! Fair compensation for expropriated land. Fear ofexpropriation must be tempered by the factthat some declarations of eminent domain bylocal government authorities may beunavoidable. After all, firms themselves demandbetter infrastructure, which generally entailswider roads and highways; land clearance andconversion for development, which ofteninvolves the repossession of large swaths ofland for industrial zones; and better publicservices, which may entail government usageof land. Even the U.S. Supreme Court hasrecently ruled, in the hotly debated 2004 Keloversus New London Case, that landrepossession of private houses and businessescould proceed to promote economicdevelopment. In these cases, the relevantquestion is whether the individual or firmforced to surrender their property receives

fair compensation for the value of that land.This is no easy task.While the 2003 Land Lawattempts to bring state compensation pricescloser to market prices, there is always a delaybetween the bureaucratic process of revaluingland and the daily changes in market value. Inaddition, entrepreneurs are often concernedabout the future value of land as opposed tothe present market value. New infrastructureand re-zoning often enhance the value of theland, but the market rate is based on theland's current status.

To capture whether firms felt that compensation intheir province was just, we posed the followingquestion, "Based on your observations of othercases in your province, do you believefirms/individuals receive fair compensation forexpropriated land?" We used the percentage offirms who ticked usually or always as an indicator.

! Risk of change in lease contract. For firms thatlease property as opposed to purchasing it,the analogous risk to expropriation is achange in the lease contract that doesmaterial harm to their business. Such changesare common occurrences in Vietnam, wheretenants often find after a few months ofproductive activity that their space has beenpromised to another investor, who offered ahigher rent even before the end of the lease.Obviously, this can do major harm to aninvestor, who has put sufficient effort intodeveloping the property.To maintainequivalency with the expropriation measure,we used an analogous five-point scale.

! Perception of fair process of disputing changes inlease contracts. Just as some eminent domain isunavoidable, disputes between landlords andtenants are also bound to arise. In these cases,the important issue is whether firms feel thereis a fair system for disputing these changes.Once again, this question is scored in anequivalent manner to the question regardingdisputes over expropriation.



Figure 6.2: Land Access and Security of Tenure Sub-Index

Access to Land is a measure of how long and how costly it is to obtain business premises. Security Tenure is a measure of the risk of expropriationor loss of business property. Both dimensions account for 50% of the final index.



Ministry of NaturalResources and the

Environment 2003 &2007 Datasets†

PCI Survey Question: B4.1.4

PCI Survey Question: B5.2



Percentage of firms ratingprovincial land conversionpolicies as good of verygood.

Total land in province withofficial Land Use RightsCertificates

Firm rating ofexpropriation risk (1:VeryHigh to 5:Very Low)

If land expropriated, firsreceive fair compensation(% Always or Usually).

Firm rating of changes inlease contracts (1:VeryHigh to 5:Very Low)

If changes in leasescontracts, is there a fairprocess for disputing them(% Always or Usually).

* Significant at 5% Level; NA = Not ApplicableAll values are at the provincial-level.2005 data only include 42 provinces and do not include the full set of indicators used in subsequent years, reflecting changes in survey questions andordering in 2006.† MONRE changed the calculation of LURCs between 2003 and 2007 in the 5 national-level cities, leading to major reductions. To address this theold calculation was applied to cities.


Percentage of firms thatfeel land availabilityconstrains their businessexpansion.

PCI Survey Question: B4.1Percentage of firms inpossession of Land UseRights Certificate (LURC).

MinMedianMaxCorrelationw/Previous YearMinMedianMaxCorrelationw/Previous YearMinMedianMaxCorrelationw/Previous YearMinMedianMaxCorrelationw/Previous YearMinMedianMaxCorrelationw/Previous YearMinMedianMaxCorrelationw/Previous YearMinMedianMaxCorrelationw/Previous YearMinMedianMaxCorrelationw/Previous Year

48.4871.3181.08NA

48.5764.2778.380.28

23.2955.2877.78

33.7352.4582.14

11.369.296.5

1.952.493.05

21.4340

58.33

2.553.09

4

044.4469.7

47.0664.7781.160.51*

51.3575.5792.450.76*

23.5356.8381.250.78*

13.2863.1397.460.85*

1.742.242.570.28*

22.2240.7657.140.37*

2.593.13.590.15

17.6540

60.710.24

49.5665.3777.060.52*

38.3681.1694.740.70*

1.1721.5159.40.48*

19.5277.5698.750.78*

1.632.042.490.95*

21.2538.8252.750.34*

2.633.123.540.25*

20.6939.09

600.27*


Table 6.2: Comparison of Land Access and Tenure Security Sub-Index (2005-2008)

62. Kaufman, Daniel, et al. 2002. Governance Matters,World BankPolicy Research Working Paper No. 2772, February, 5-7. Florini, AnnM., 1999. Does the Invisible Hand Need a Transparent Glove? ThePolitics of Transparency, paper prepared for the World BankConference on Development Economics,Washington, D.C., April.Tenev, Stoyan, Amanda Carlier, Omar Chaudry, and Quynh-TrangNguyen, 2003. Informality and the Playing Field in Vietnam'sBusiness Sector,Washington, D.C: International FinanceCorporation.

6.3 TransparencyTransparency is one of the most crucial factorshighlighted by academics and developmentpractitioners in identifying environments conduciveto private sector business.62 Tara Vishwanath andDaniel Kaufman define transparency as:

[T]he increased flow of timely and reliableeconomic, social and political information aboutgovernment service provision, monetary andfiscal policy.… Contrariwise, a lack oftransparency may be described as someone …

deliberately withholding access to, ormisrepresenting, information or failure to ensurethat the information provided is of adequaterelevance and quality.63

A working measure of transparency shouldtherefore encompass the following four dimensions:access, equity of information, predictability andopenness.We measure transparency according tothese attributes, using the nine indicators describedin the subsections that follow.64 Results of theseindicators are shown in Table 6.3 and Figure 6.3.


Table 6.3: Comparison of Transparency Sub-Index (2005-2008)

PCI Survey Question:F1.1-F1.13 (Factor 1)†

Transparency of planningdocuments

PCI Survey Question:F1.1-F1.13 (Factor 2)†

Transparency of legaldecisions and decrees

PCI Survey Question: F2

Relationship important orvery important to getaccess to provincialdocuments (% Importantor Very Important)

Min

Median

Max


Min

Median

Max


Min

Median

Max


-0.79

-0.04

1.49

NA

-0.49

0.04

0.48

NA

50

72.11

100

NA

-0.4

-0.02

0.68

0.40*

-0.45

0.01

0.35

0.46*

31.48

62.5

77.14

0.27

-0.39

-0.002

0.57

0.63*

-0.4

0.02

0.34

0.55*

38.4

56.6

73.4

0.38*

-0.41

0.008

0.48

0.62*

-0.35

0.02

0.4

0.59*

33.57

49.82

67.9

0.55*


63. Vishwanath,Tara, and Daniel Kaufmann. 1999. "TowardsTransparency and Finance and Governance," World Bank Mimeofound at www.worldbank.org/wbi/governance/pdf/tarawish.pdf.

64. Vishwanath and Kaufman concede, however, that transparency isnotoriously difficult to measure, because it deals with agents whoactively try to hide information. Measuring transparency in Vietnam canbe additionally troublesome, because the term (minh bach) is notobvious to many respondents. It can take on different meanings acrossgeographical boundaries and across firms with different relationships tothe provincial government. Misunderstanding of the term is notsystematic across provinces, but varies according to firms' interactionswith Western donors and investors.


* Significant at 5% Level; NA = Not ApplicableAll values are at the provincial-level.2005 data only include 42 provinces.† Indicators result from factor analysis of a battery of 13 documents.? In 2007 and 2008, 0.5 values were allowed to denote provinces that provided the relevant information, but not in a sufficient manner to be useful.


PCI Survey Question: G9.3



PCI Survey Question: E.1.5

Analysis by VNCI ResearchTeam

Min

Median

Max


Min

Median

Max


Min

Median

Max


Min

Median

Max


Min

Median

Max


Min

Median

Max


34.35

56.07

80

NA

52.17

75.22

96.15

NA

4.35

14.91

60.38

NA

0

12.16

61.54

NA

0

10

21

NA

37.74

57.21

82.35

0.22

47.17

61.05

86.96

-0.16

2.76

9.49

37.88

0.38*

0

8.84

20.9

-0.29

24.49

48.05

60.94

NA

0

9

18

0.36*

38.7

54.7

65.1

0.45*

24.1

44.7

73.2

0.52*

1.89

7.96

18.3

0.46*

0.9

7.57

21.62

0.45*

30.3

48.28

72.84

0.63*

0

13.75

20

0.51*

40

53.04

67.47

0.55*

17.39

36.71

54,25

0.73*

1.03

6.94

15.69

0.3*

1.21

8.57

18.6

0.52*

6.67

20.08

33.77

0.53*

0

14.25

20

0.70*


Friends important fornegotiating withgovernment (% Importantor Very Important)

Negotiations with taxauthority are an essentialpart of doing business (% Agree or Strongly Agree)

Predictability ofimplementation of Centrallaws at the provincial level(% Usually or Always)

Province discussed changesin laws with you (%Usually or Always)

Services Provided byProvincial Agencies:Consulting on National andProvincial Regulations (%Very Good or Good).

Openness of ProvincialWeb Page Score


Figure 6.3: Transparency Sub-Index

Access to Documents measures the availability of 13 different legal and planning documents. Equity and Consistency gauges whether thosedocuments and other services are universally available. Predictability is the ability of firms to predict laws and implementation of those laws. Each offirst three dimensions are drawn from survey data and are weighted at 20% of the final index. Openness of Web Page is a accounting of businessinginformation and services on provincial web pages. It accounts for 40% of the final index.


6.3.1. AccessThe first dimension of transparency is access,defined as the timely dissemination of information.While legally, information on land and provincialplanning may be available to all, accessing thatinformation can often be quite problematic. Accessincludes the availability of new laws, implementingdocuments, or provincial decisions.When changes inthe legal regime are not readily accessible, a firmmay operate for several years, only to find itself onthe wrong side of the law, simply out of ignorance.In most cases, such ignorance will cost the firm verylittle, but there is always the potential for anunscrupulous government official to exploit theasymmetric information about the legal code totheir advantage. Conversely, a firm may be eligiblefor savings, investment opportunities, or tax refunds,but never take advantage of these because they areunaware of the benefits available to them.

To measure access to information across thesurveyed provinces, a list of the 13 provincialdocuments most vital to business operations wasgiven to each firm.They were then asked to ratetheir access to these documents on a scale rangingfrom very easy to impossible. Factor analysis of theiranswers generates the same two baskets of

business documents discovered in every PCI goingback to 2005:

! Access to provincial planning documents (Factor 1),which include a firm's evaluation of theiraccess to the Provincial Budget, 10 - and 5-Year Master Plans, Annual Socio-EconomicPlans and Infrastructure Development Plans,Private Sector Action Plans, Central InvestmentPlans, and Land Use Allocation Maps.

! Access to laws and regulations (Factor 2), whichinclude a firm's evaluation of their access toCentral Laws and Decrees, MinisterialImplementing Documents, Decisions of theProvincial People's Committee, applications forregistration and land use, and changes in taxinformation.

Table 6.4 displays the factor loadings of the twovariables. High loadings represent a strongcorrelation with the underlying factors -- planningor legal documents.Together the two factorsexplain over 60 percent of the total variance amongall documentation. Figure 6.4 plots the averagescore of these two factors on a scatter plot.Provinces located in the northwest quadrant arelocalities that demonstrate high transparency onboth types of documentation.

Table 6.4: Factor Analysis of Planning and Legal Documents

10 and 5 Year Master PlansYearly Planning DocumentsProvincial BudgetCentral Investment PlansPrivate Sector Action PlansPlans for InfrastructureLand Use Allocation MapsProvincial Investment Incentive PoliciesImplementing DocumentsInformation on Changes in Tax LawsCentral Decisions and DecreesPCOM Decisions and CircularsApplications for Registration and Land UseEigen ValueCumulative Variance Explained

Document

0.80870.79930.73740.69850.69430.68170.61590.58250.30110.07670.34270.3950.30676.7934.26

Planning Legal

0.22780.26350.16830.33630.32520.40720.31720.47260.80670.76120.74910.74480.5715

1.261.42

Documents listed in Questions F1_1 to F1_13 of Survey; Extraction Method: Principal Component Analysis; Rotation Method:Varimax with KaiserNormalization.

6.3.2. Equity and Consistency of ApplicationWhile dimension captures access to information, itdoes not necessarily imply equitable access, whichcan lead to severe inefficiencies in the use ofprovincial resources-inefficiencies that representmore than a simple transfer of resources from oneparty to another.Take, for instance, provincialplanning.The impact of grand infrastructure andland conversion plans will be marginal if the detailsare available to only a select few insiders.65 If only a

few knowledgeable insiders know the location offuture infrastructure projects and industrial zones,these insiders can then make one-way bets bybuying up the land ahead of time. Other investorsin real estate must make large conjectures based onsmall bits of information. Similarly, opportunities forcorruption are rife when the provincial budget isnot available to firms, so firms are unable tocompare a province's actual expenditures to itsplanned outlays.Therefore, a follow-up questionmeasures different aspects of equitable access in thetransparency sub-index.Three variables are used tocapture equity:

! Access dependent on relationship with provincialgovernment is an indicator measuring the


Figure 6.4:Access to Planning and Legal Documents by Province

65. The arrests in Phu Quoc island of Le Minh Be and Do To, theDistrict People's Committee Chairman and People's Council,illustrate this point. According to Thanh Nien, these officials werearrested for taking bribes to hand over land documents that shouldhave been available legally (Thanh Nien, 8 September 2004).


percentage of firms that felt that having arelationship with someone in the provincialgovernment is important or very importantfor facilitating access to the above documents.

! Friends/family important for negotiations withprovincial officials. Many firms rely on anextended network of relations to receivepositive judgments from provincial officials thatmay not be possible for less connected firms.Researchers measured this question byincluding an indicator for the percentage offirms that agree or strongly agree with thesentence: "Friends and family are important fornegotiations with provincial officials".

! Negotiations over tax payments with the localtax authority are an essential part of doingbusiness is used to capture how consistentlytax rules are applied.While negotiations are anormal part of tax collection with householdbusinesses, private firms should have a taxcode and regularly submit payments based onvalue added tax receipts.

6.3.3. PredictabilityThe third attribute of transparency is predictability,or the notion that provincial laws and regulationsare implemented in a manner that allows firms toforecast and build changes into their business plans.Two indicators are used to capture the extent ofpredictability of implementation. In this context, it isimportant to capture information as to whether

firms understand how provincial decisions are madeand how they will be implemented, so that they cancorrectly understand the direction of long-termstrategies and increase their ability to makeinformed investment decisions. Predictability isassessed based on the percentage of firmsanswering always or usually to the following twoquestions:

! How predictable is the implementation ofcentral rules, laws, and regulations thatmaterially affect your business at the provinciallevel?

! How often do provincial leaders discusschanges in laws with your firm?

6.3.4 OpennessA measure of openness is created by assessing theprovincial webpages of every province to determinethe business information available to the firms. Aprovince receives one point for all of the types ofinformation listed in Table 6.5.Together theseelements total 15. Because the number of self-reported hits on a website is prone to error, weuse the Google Alexa ranking of the most clicked-on and linked-to websites, as our measure of howimportant the website is in the provincial businessenvironment.66

66. http://developers.evrsoft.com/find-traffic-rank.shtml.Thanks to AnGiang's People's Committee for suggesting this procedure.


Province has website

Website includes map of province

Information on actual infrastructure/project planning

Statistics on province's size/weather/human resources

Incentive Policy to attract Investment(domestic)

Incentive Policy to attract Investment(foreign)

Information on Industrial Zones/Industrial Concentration

Statistics on existing Investors

Business/Economic information of districts within provinces

Information on specialized provinces endowments/capacity

Reports on Provincial economic achievements

Application for Registration/ Incentive, land use

Contact Information for relevant authorities

Information on obtaining VAT receipts

Other (Something special that firms may need for business)

Sub-Total

Google Alexa Ranking

Possible Points

Document

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

15

0-200,000 = 5

200,001-400,000= 4

400,000-600,000 =3

600,000-800,000= 2

800,000+ = 1

20

Planning

Table 6.5: Scoring System for Provincial Websites


6.4 Time Costs of RegulatoryComplianceThe study of transaction costs in time has been animportant element of the economic transitionliterature.67 The old business maxim "time is money"is particularly relevant in the Vietnamese provinces.Firm managers are often torn away from theirbusiness operations in order to deal with andattend to mundane bureaucratic problems-time thatcould be more productively spent managing theoperations of the company is lost.The PCIconsiders two, equally weighted dimensions of timecosts: bureaucratic procedures and time lost toinspections (see Table 6.6 and Figure 6.5).

Indicators for bureaucratic procedures include:

! What percentage of management's time is spentdealing with bureaucratic procedures andpaperwork? This question is given to firms inthe form of a five-point Likert scale.Theindicator is taken from the number of firmsthat answered three or above, essentiallycapturing the percentage of firms who spendmore than 10 percent of their time dealingwith government required paperwork.

! Have the above "time taxes" been reduced overthe past three years? This indicator measuresthe progress provinces have made since theyear 2000.

To some extent, inspections policy is unavoidable,especially as Vietnam moves to a fully regulatorysystem.68 As more and more environmental, health,and safety licenses are removed at the onset ofbusiness activities, the responsibility shifts toregulatory agencies to ensure that these standardsare continually met by private firms.The trick is tomanage these regulatory responsibilities withoutexpensive and costly intervention.

! How many inspections a year must firms endure?According to present Vietnamese Law, no firmmay receive more than two visits from allagencies per year. In fact, the median numberof inspections does not exceed two for anyprovince-the vast majority of provincesaverage only one inspection. Because of thispositive finding, the number of inspections

receives a diminished weight in the final index.There is no longer any variation on thisindicator.

! Firms believing inspection policy has improvedover the past three years. The second indicatorfor inspection captures the percentage of firmsthat felt inspection policy had improved sincethe passage of the Enterprise Law.While newimplementing documents potentially loweredthe number of inspections, new burdensplaced on regulatory agencies create additionalincentives to ramp up inspections.

! Median length of tax inspections. Someprovinces make up for the lower absolutenumber of inspections by increasing theirduration.Therefore, the third indicator recordsthe number of hours it takes to complete taxinspections.There is some confusion about therole of the tax authority, as in some provinces,officers of the tax authority are sent to smallbusinesses to take a cursory look at theirbooks and offer assistance with new taxpolicies.These visits are not technically'inspections', but firms often had troubledistinguishing them from true inspections.ThePCI survey instrument is careful to distinguishbetween these voluntary visits and formalinspections.

67. European Bank of Reconstruction and Development. 1999. EBRDTransition Report 1999, London:, 120-128.World Bank, 2002. 2002Transition:The First Ten Years: Analysis for Eastern Europe and theFormer Soviet Union,Washington, D.C., 2002, 103-107. Hellman,Joel et al, 2002. Seize the State, Seize the Day: State Capture,Corruption, Influence in Transition,World Bank Policy ResearchWorking Paper No. 2444,World Bank Institute, September, 7-14.Hellman, Joel, et al. 2002. "Measuring Governance, Corruption, andState Capture: How Firms and Bureaucrats Shape the BusinessEnvironment in Transition Economies," World Bank Policy ResearchWorking Paper No. 2312,World Bank Institute, April.

68. Though Vietnamese authorities often distinguish between kiem tra(short-term control visits) and thanh tra (when local authorities arecalled in due to suspected problems), in practice, there is simplytoo much overlap to separate them.This survey considers themtogether.


PCI Survey Question: G1

Percentage of firmsspending over 10% of theirtime dealing withbureaucracy orbureaucratic regulations.

PCI Survey Question: G2Days spent on bureacracyreduced in past two years(%)†

PCI Survey Question: D1Median number ofinspections (all agencies)

PCI Survey Question: D2Inspections have decreasedin past two years (%)†

PCI Survey Question: D4Median Tax Inspectionhours

Min

Median

Max


Min

Median

Max


Min

Median

Max


Min

Median

Max


Min

Median

Max


3.64

13.67

30.43

NA

18.18

40

78.57

NA

1

1

3

NA

12.5

42.12

70

NA

1

7.5

24

NA

6.52

21.24

39.39

0.44*

23.94

41.72

60.87

0.18

0

1

2

0.35*

28.07

45.52

73.91

0.26

1

8

40

0.62*

10.94

21.87

43.75

0.62*

13.75

22.86

35.04

0.21

1

1

2

0.30*

11.9

24.36

36.92

0.30*

2

8

40

0.86*

13.83

22.99

42.55

0.67*

12.78

23.85

34.86

0.31*

1

1

2

0.46*

11.54

24.51

37.59

0.51*

1

8

32

0.88*


Table 6.6: Comparison of Time Costs of Regulatory Compliance (2005-2008)

* Significant at 5% Level; NA = Not ApplicableAll values are at the provincial-level.2005 data only include 42 provinces.† In 2005 and 2006, provinces were asked to compare changes to before Enterprise Law not the past two years. Pre-testing for 2007, however,revealed many firms could not remember the pre-Enterprise period sufficiently enough. This small wording change is the reason behind the lowerpercentages.


Figure 6.5:Time Costs of Regulatory Compliance

Regulatory Compliance is a measure of the wait and costliness of bureaucratic procedures. Inspections Policy is the length and frequency ofregulatory inspections in the province Both dimensions account for 50% of the final index.

6.5 Informal ChargesThis section analyses the amount that firms pay infees, fines, and extraordinary payments as a normalpart of doing business (see Table 6.7 and Figure 6.6below).This group of five indicators measures theextent of the problem by gauging the frequency,type, and amount of extra payments:

! The percentage of firms that believe that extrapayments are an obstacle is used as a simplemeasure of the scope of extra payments inthe sample.

! The percentage of firms that felt thatenterprises in their line of business weresubject to bribes from provincial authorities.Note that this question was phrased so thatfirms were answering about firms in generalas opposed to their own behavior withbribes-thereby increasing the response ratesubstantially. Sixty-six percent of firms in themedian province believed such paymentswere normal.

! The percentage of firms paying more than 10percent of their revenue in extra payments isused to measure the scale of extra payments.To counter the distortion from firms notanswering the question, we performed twodiagnostic checks. First, we ran a test to ensurethat the percentage of nonresponses in eachprovince was unrelated to firm scores(percentage missing correlates with theamount paid at -0.1 which is not significantlydifferent from 0). Second, missing data was

imputed using a program known as NORM.69

This procedure imputes an expected score fora firm, given its answers to all other questionsincluding descriptive characteristics as well asother measures of governance.The imputationprocedure raised average scores for allprovinces by about two percent, but noprovinces were affected dramatically by theoperations. As a result, researchers feltcomfortable using the non-imputed data in theanalysis.

! Officials use compliance with local regulationsto extract rents.The percentage of firmsagreeing with this statement provides addednuance on the business environment. Bribery isof course an issue, but when new barriers tobusiness are erected to generate rents forlocal officials, the effects can be pernicious.Thirty-seven percent of firms in the medianprovince agreed or strongly agreed with thestatement.

! Informal charges delivered expected results:Field testing the survey revealed that somefirms believed informal charges were beneficialif they expedited bureaucratic procedures,were predictable, and delivered the expectedresults. In fact, it is possible that firmsvoluntarily supplement provincial fees. Suchbehavior is premised on the notion that firmsget what they pay for. As a result, we ask firmsif their informal payments provided theexpected results.


69. Schafer, J.L. 1997. Analysis of Incomplete Multivariate Data. London:Chapman & Hall. A free version of this software can be obtained athttp://www.stat.psu.edu/~jls/misoftwa.html



Percentage of firms thatbelieve that extrapayments are a majorobstacle to doing business.


Percentage of firms that feltthat enterprises in their lineof business were subject tobribe requests fromprovincial authorities.

PCI Survey Question: G4Percentage of firms payingover 10% of their revenuein extra payments.


Government usescompliance with localregulations to extract rents(% Strongly Agree orAgree)

PCI Survey Question: G5Informal Charges DeliveredExpected Result (% Usuallyor Always)

Min

Median

Max


Min

Median

Max


Min

Median

Max


Min

Median

Max


Min

Median

Max


5

26.42

60.61

NA

6.67

26.57

48.28

NA

0

9.6

29.41

NA

22.73

42.59

65.09

-0.48*

53.57

70

84.62

0.05

4.35

12.99

34.38

0.21

22.86

39.76

76.74

NA

20.83

47.89

65.93

NA

13.95

26.03

44.4

0.47*

40

68.25

82.72

0.56*

1.39

11.54

26.19

0.45*

17.44

38.21

79.41

0.78*

29.03

48.28

59.8

0.2

18.75

27.71

55

0.46*

45.54

65.93

83.59

0.64*

2.13

9.89

22.08

0.55*

20

37.12

64.54

0.68*

27.94

48.99

62.91

0.50*


Table 6.7: Comparison of Informal Charges (2005-2008)

* Significant at 5% Level; NA = Not ApplicableAll values are at the provincial-level.2005 data only include 42 provinces and do not include the full set of indicators used in subsequent years.


Figure 6.6: Informal Charges Sub-Index

Original Indicators are those used in the original 2005 survey, while New Indicators are those first used in 2006. Both dimensions account for 50% ofthe final index.

6.6 Competition Environment andState-Owned Enterprise Bias Do private firms feel that they must they overcomea subtle bias toward the state in the form of specialadvantages or soft budget constraints? State sectorfavoritism is a measure of the bias of provincialgovernments toward SOEs in terms of contracting,land, incentives, and access to capital. State sectorbias does not necessarily include direct competitionin the same product or service; it also may includeassistance in competition for skilled labor, land, orcredit. Some provinces have stated explicitly thattheir primary goal is to promote large state-ownedchampions as the primary engine of growth,allowing the private sector to offer only secondaryservices and intermediate good to those provincialchampions.70 Others may not have such an explicitbias, but instead have an institutional incentive topromote SOEs, because of the high employment orrevenue they generate for the province. Results areshown below in Table 6.8 and Figure 6.7. Indeed, anearlier Vietnam National Competitiveness Initiative(2006) study by Nguyen Van Thang using PCI dataconcluded that, "the density of SOEs (in a province)has a negative impact on the private sector's accessto key resources (land and bank loans) and markets,and a negative influence on the private sector'sgrowth in terms of number of firms andemployment."71

6.6.1. Perceptions IndicatorsPerception indicators are based on five-point scalesasking how strongly firms agree with a statement.The percentage of firms answering agree or stronglyagree is used for each indicator.

! Favoritism toward SOEs offers direct evidenceof bias toward the state sector by accountingfor the percentage of firms that agree thatthere is a bias in decision making on the partof provincial officials.

! Attitude toward private sector captures thepercentage of firms that agree that the attitudeof their provincial government toward theprivate sector is positive.

! Attitude has improved over past three years is amore dynamic question, measuring whetherfirms agree that the attitude has improvedsince 2005.

! Attitude depends on contribution probeswhether firms agree that biases in provincialpolicies is determined, not by ownership, butresults from a bias toward a specific set ofrevenue or labor-producing firms, regardless oftheir ownership.This variable is intended tocapture whether administrations are biasedtoward provincial champions and raise entrybarriers to possible competitors.

! Equitization policy: Firms were asked to ranktheir perception of provincial equitization.Thismeasure demonstrates whether the leadershiphas made a concerted effort to reduce thenumber of local SOEs, thereby promotingprivate sector competition.

! Bias toward equitized firms takes the previousequitization question a step further by askingwhether, despite wide-ranging equitization, statefavoritism persists by provincial officials, allowingspecial privileges to continue for equitized firms.Recent evidence has suggested that connectionsbetween former general managers of local SOEsand provincial leaders led to equitizedcompanies being favored in access to capitalpublic procurement contracts.72

6.6.2. Hard IndicatorsHard indicators toward state sector favoritismfollow.

! Change in the number of local SOEs.TheStatistical Handbook of the General Statistical


70. Malesky, Edmund. 2004. "Entrepreneurs on the Periphery: A Studyof Private Sector Development in Beyond the High PerformingCities and Provinces of Vietnam," Mekong Private SectorDevelopment Facility Private Sector Discussion Series, Number 18,November;Vietnam.

71. Nguyen Van Thang. 2005. "Is The Development of State-OwnedEnterprises (SOEs) Crowding Out The Private Sector? Evidence fromPCI Survey," Ha Noi:Vietnam Competitiveness Initiative,August.

72. Malesky and Taussig. 2008. "Where is Credit Due? Companies,Banks, and Locally Differentiated Investment Growth in Vietnam,"Journal of Law, Economics, and Organization

Office (GSO) reveals that despite the earlierSOE reforms initiatives in 1995, equitizationdid not begin in earnest in all provinces until1997.To maintain a consistent source, theresearch team calculates the decline in locallymanaged SOEs between 2000 and 2007 fromGSO census data-2000 rather than 1997 isused because this was the first year of theGSO census.We chose this number, ratherthan the simple number of equitizations,because the decline also captured the numberof firms that were liquidated, closed due topoor performance, or merged with otherenterprises.

! The ratio of local SOE share of outstandingliabilities to their share of the provincialeconomy.This final SOE indicator is designedto capture disproportionate lending to thestate sector (provinces lending large amountsto inefficient and unproductive SOEs).Theformula described below will deliver a scoregreater than 1 if the state-sector share ofoutstanding liabilities exceeds the state-sector'sshare of the economy, and less than 1 in theopposite case:

Ratio =

Loans, in the formula above, are outstandingliabilities as measured by the most recent (2007)GSO Enterprise Census. PGDP is the provincialgross domestic profit as proxied by total enterpriserevenue in the same census. In each case, thesubscripts LSOE and Prov represent data for localstate-owned enterprises and the total provincerespectively. Given the higher productivity ofinvestment in the private sector (as measured byreturns on investment, employment creation, and soon), it would be desirable to see more provinceswith ratio scores less than 1.


LoansLSOE

LoansProv PGDPLSOE

PGDPProv


PCI Survey Question:H4.10

PCI Survey Question: H1

PCI Survey Question: H2


PCI Survey Question:H7.12


Min

Median

Max


Min

Median

Max


Min

Median

Max


Min

Median

Max


Min

Median

Max


Min

Median

Max


33.33

59.4

78.95

NA

20.59

47.83

78.26

NA

53.85

71.34

95.45

NA

29.09

52.17

78.57

NA

18.18

31.87

50

NA

30

44.35

63.59

0.25

30.21

48.28

71.56

0.63*

46.15

68.34

83.08

0.31*

18.28

36.56

50

0.23

11.29

29.45

41.67

0.14

25

57.28

74.55

NA

27.38

42.16

59.33

0.25*

24.5

44.97

67.37

0.67*

36.23

63.41

80.91

0.61*

22.22

35.1

50.75

0.17

12.7

29.68

42.72

0.35*

25.53

59.5

87.67

0.58*

23.81

38.38

50.29

0.35*

32.71

53.4

72.22

0.53*

44.86

68.78

81.06

0.60*

51.94

65.07

84.62

0.03

17.19

30.78

42.45

0.24

0

1.65

6.56

0.65*


Favoritism toward the statesector is an obstacle tofirm's business (% StronglyAgree or Agree).

Perceived attitude ofprovincial governmenttoward private sector (% Very Positive or Positive).

Attitude has improved inthe past two years (%).†

Attitude does not dependon contribution toprovincial revenue (%Strongly Agree or Agree).

Favoritsism towardequitzied companies anobstacle to firm's business(% Strongly Agree orAgree).

Firm rating of provincialequitization policies (%Very Good or Good).

Table 6.8: Comparison of SOE Bias and Competition Environment (2005-2008)


GSO Enterprise Census2000-2006

(Author's Calculation)

Percentage change innumber of locally-managedstate owned enterprisessince 2000.

GSO Enterprise Census2005 & 2006

(Author's Calculation)

Ratio of Local SOE shareof liabilities to their shareof revenue contribution inprevious year.

Directly from StateCommercial Banks

Average proportion ofState Commercial BankLoans to SOEs!

Min

Median

Max


Min

Median

Max


Min

Median

Max


-66.67

-30.72

60

NA

0.49

1.27

4.13

NA

1.5

16.2

71.1

-71.11

-41.65

24

0.89*

0.4

1.29

10.33

0.25*

-93.94

-60

25

0.80*

0.27

1.39

3.91

0.29*


* Significant at 5% Level; NA = Not ApplicableAll values are at the provincial-level.2005 data only include 42 provinces.† In 2005 and 2006, provinces were asked to compare changes to before Enterprise Law not the past two years. Pre-testing for 2007, however,revealed many firms could not remember the pre-Enterprise period sufficiently enough. This small wording change is the reason behind the lowerpercentages.! Due to propietary concerns related to State Commercial Bank Equitization, this data was not made available in 2007.


Figure 6.7: Bias Toward State-Owned Enterprises Sub-Index

Firm perceptions of SOE Bias account for 60% of the index, hard data of bias accounts for 40%.

6.7 Proactivity of ProvincialLeadershipAmbiguity is a standard part of doing business inVietnam and is often the result of unclear wording inlegal documents, long delays in implementingdocuments being promulgated for central laws ordecrees, and contradictions between implementingdocuments (circulars, directives, official letters, andPeople's Committee decisions) and even central lawsthemselves. For many industry segments new toVietnam, there is a lack of any clear legal regime.Whenbusiness projects are delayed because of legalambiguity, the choices of the provincial government canmake a huge difference in the success of the businessventure. Provincial officials can cost businessesconsiderable time and money by forcing them to waituntil the ambiguity is cleared up by subsequentimplementing documents or an appeal to centralauthorities. Unfortunately, a few provinces even usethese uncertainties as an entry barrier to firms thatmight offer competition to their local champions. Onthe other hand, provinces that are creative and cleverabout working within the confines of central law canbe of major assistance to private sector firms. Similarly,provinces with a knack for crafting proactive provincialinitiatives to solve the problems of private firms canhave a positive impact on private sector development.

Four indicators determine the extent of provincialdynamism, by recording the percentage of firms that

agree or strongly agree with the followingstatements (see Table 6.9 and Figure 6.8):

! Provincial officials are knowledgeable enoughabout national law to find opportunities withincurrent legal documents to solve firm problems.This indicator captures how well the provincialleadership operates within the existing legalframework.

! Provincial officials are creative and clever aboutworking within national law to solve the problemsof private sector firms. The measure recordswhether the provincial leadership is capable ofseeking opportunities or loopholes in thepresent central law that may aid firmperformance.

! All good initiatives come from the provincialgovernment, but central authorities frustratethem. This indicator seeks to gauge therelationship between provincial and centralauthorities, specifically whether the province ismore proactive about assisting the privatesector than central officials.

! There are no good initiatives at the provincial level;all important policy comes from the centralgovernment. The final indicator simply recordsthe percentage of firms that believe that theirprovincial leadership has nothing to do with newinitiatives that aid the private sector.



PCI Survey Question: H7.2

Provincial officials areknowledgeable enough aboutpresent national law to findopportunities within existinglaw to solve firm problems(% Strongly Agree or Agree)


Provincial officials arecreative and clever aboutworking within the nationallaw to solve the problemsof private sector firms (%Strongly Agree or Agree).


All good initiatives comefrom the provincialgovernment, but the centerfrustrates them (% StronglyAgree or Agree).


There are no goodinitiatives at the provinciallevel; all important policycomes from the centralgovernment (% StronglyAgree or Agree).

Min

Median

Max


Min

Median

Max


Min

Median

Max


Min

Median

Max


43.75

76.93

94.29

NA

31.25

63.27

85.71

NA

4.76

31.35

60

NA

7.89

33.33

60.42

NA

51.61

74.44

93.48

0.60*

40

61.88

88.64

0.69*

16.04

29.07

61.54

0.40*

14.63

32.88

48.84

0.59*

53.68

71.74

92.47

0.68*

40.22

58.12

87.91

0.76*

16.67

30.95

56.63

0.47*

12.2

33.33

58.33

0.55*

57.35

77.28

91.41

0.68*

40.9

61.5

85.05

0.75*

10.84

20.99

55.17

0.63*

17.95

32.99

66.25

0.53*


* Significant at 5% Level; NA = Not ApplicableAll values are at the provincial-level.2005 data only include 42 provinces.

Table 6.9: Comparison of Proactivity (2005-2008)


Figure 6.8: Proactivity Sub-Index

6.8 Private Sector DevelopmentPoliciesThe next sub-index goes a step beyond theproactivity sub-index by asking how specificprovincial initiatives promote private sectordevelopment. In a range of surveys of the privatesector in Vietnam, firms listed among their chiefobstacles difficulties in obtaining information onoverseas and domestic markets, difficulties inunderstanding new changes in regulatoryinformation and problems finding enough skilledemployees to conduct their operations.The finalsub-index is comprised of four questions thatattempt to measure how well provincial officials aredoing in resolving these problems on behalf of firms(see Table 6.10 and Figure 6.9). Firms were askedto rank their provinces on a five-point scale, gauginghow effective they thought their province's policieswere in the following four areas:

! Provision of market information through thepublication of local pamphlets listing majordomestic and overseas buyers for keyprovincial products.

! Matchmaking for business partners throughactivities such as the introduction ofinternational exporters to local firms, theidentification of suppliers for intermediategoods, and the availability of business servicesfor firms unable to locate business partners ontheir own.

! Business information on changes in regulation.High-scoring provinces offer assistance to firmsdrowning in documentation from changes inthe legal code.

! Industrial zones in their province. How firmsfeel about their provinces' attempts to assistdevelopment by creating industrial zones andsmaller industrial concentrations, therebyallowing forms to access business premises andtop-notch infrastructure.

! Technology and technology-related services.This includes informing firms about theefficiency improvements of new technologiesin their business areas, as well as basiccomputer training, which will help them accessonline services provided by the province.

! The number of trade fairs held by theprovince in 2006 and planned for 2007 is ahard data indicator inserted to account forpotential perception biases held by firms.

! The number of private providers of businessservices is calculated from the National TaxAuthority's list of registered firms. Allowingprivate firms to provide private sectordevelopment services for profit savesprovincial governments money and ensurestheir sustainability. In addition, private providersare often more efficient and can better tailortheir products to the needs of end users.



Table 6.10: Comparison of Private Sector Development Policies (2005-2008)

* Significant at 5% Level; NA = Not ApplicableAll values are at the provincial-level.2005 data only include 42 provinces and do not include the full set of indicators used in subsequent years.Because maximum scored by HCMC is an outlier on both of these variables. Lower values of 10 and 100, the number scored by the second highestprovince were used to calculate the sub-index.


Services Provided byProvincial Agencies: Provisionof market information (% Very Good or Good)

PCI Survey Question:E1.11



PCI Survey Question:E.16

PCI Survey Question: G9.3(Data provided by VietTrade of the Ministry of

Trade)

Tax Authority 2008(Author's Calculation)

Min

Median

Max


Min

Median

Max


Min

Median

Max


Min

Median

MaxCorrelationw/Previous Year

Min

Median

Max


Min

Median

Max


Min

Median

Max


0

21.99

41.54

0

13.36

26.19

30.43

49.72

64.89

0.17

24.49

48.05

60.94

0.18

27.03

50.68

79.03

NA

11.91

45.8

81.36

NA

18.92

41.73

72.34

NA

0

0

6

NA

23.52

44.19

66.93

0.18

15

31.52

62.96

0.46*

29.17

56.22

79.55

0.76*

6.67

50.84

83.48

0.84*

14.29

43.88

79.55

0.32*

0

0

12

0.18

8.16

20

34.86

0.67*

1.4

11.59

30

0.69*

1.37

20.69

48.84

3.07

23.87

72.89

0.84*

4.28

15.87

48.76

0.82*

0

2.25

80

0.62*

0

1

3529

NA


Services Provided byProvincial Agencies:Matchmaking for businesspartners (% Very Good orGood).

Services Provided byProvincial Agencies: ExportPromotion and Trade Fairs(% Very Good or Good).

Services Provided byProvincial Agencies:Industrial Zones and SMEConcentrations (% VeryGood or Good).

Services Provided byProvincial Agencies:Technology and TechnologyRelated Services (% VeryGood or Good).

Trade fairs held byprovince in previous yearand registered for presentyear.

Number of private publicservice providers inprovince


Figure 6.9: Private Sector Development Policies Sub-Index

Firm perceptions of PSD delivery account for 60% of the index, hard data of bias accounts for 40%.

6.9 Labor Training Sub-IndexA major challenge confronting Vietnam is to createnew jobs for the 1.4 million new job seekersentering the labor force each year.73 At the sametime, one of the most consistent complaints fromfirms over the past few years has been the lowcapacity of workforce and firms' inability to find andrecruit skilled and semi-skilled workers.74 Thejuxtaposition of the two commonly citedstatements is telling.Vietnam is relying on its privatesector to absorb the new labor entrants, but theprivate sector is not willing to employ just anyone;they want employees with the skill to add value totheir businesses. As a result, provincial efforts toimprove the skills sets of their local labor forces area critical determinant of a successful businessenvironment. Neoclassical economists might arguethat firms should invest in their own labor training,and many have. If the firms do train workers,however, they are often poached by othercompanies, which pay higher wages but do not bearthe upfront investment in training.This collectiveaction problem can be solved by provinces offeringgeneral labor training.

The following indicators identify the state of labortraining in the provinces (see Table 6.11 and Figure 6.10):

! The percentage of firms who rate education intheir province as good or very good. This indictoris a baseline measure of firms' perceptions ofeducation in the province. After all, many newhires will be coming straight from public schoolprograms rather than from vocational centers.

! The percentage of firms that view vocationaltraining in their province as good or very good.This is a more specific indicator, narrowing thescope on exactly the way provinces cansupport labor skill upgrading.

! The percentage of firms that rate provincialexchange services as good or very good.Exchange services are fairs or agencies, whichhelp match employees with firms that makeuse of their talents.These are crucial forlowering the transaction costs faced by firmssearching for employees on their own.

! The number of vocational schools per 100,000provincial citizens.This hard indicator anchors firmperceptions. This data, supplied by the GeneralDepartment of Vocational Training, includescenters and agencies funded from localbudgets or privately funded ventures withlicenses to operate in the province.

! The actual number of labor exchange bureaus isadded as an additional hard indicator tomodify firm perceptions.


73. -United Nations, 2003. "Tap the Energies of Youth," UN Message onInternational Youth Day, August 12,http://209.85.173.104/search?q=cache:7K47xNuxn6AJ:www.un.org/esa/socdev/unyin/workshops/curtain.pdf+United+Nations,+2003.+%E2%80%9CTap+the+Energies+of+Youth,%E2%80%9D+UN+Message+on+International+Youth+Day,+August+12&hl=en&ct=clnk&cd=2&gl=us

74. Ministry of Labor, Invalids, and Social Affairs, 2006. "Labor and SocialIssues Emerging from Vietnam's Accession to the WTO."http://209.85.173.104/search?q=cache:hxM4imcQSZgJ:www.vnep.org.vn/Modules/CMS/Upload/6/tac%2520dong%2520cua%2520WTO_Mr.Tuyen.doc+%E2%80%9CLabor+and+Social+Issues+Emerging+from+Vietnam%E2%80%99s+Accession+to+the+WTO&hl=en&ct=clnk&cd=1&gl=us



Services Provided byProvincial Agencies:General Education (% VeryGood or Good)


Services Provided byProvincial Agencies: LaborVocational Training (% VeryGood or Good)


Services Provided byProvincial Agencies: LaborExchange Services(% VeryGood or Good)

General Department ofVocational Training

Number of locallymanaged vocationalschools per 100,000citizens.

Ministry of Labor, Invalidsand Socal Affairs: General

Labor Department

Number of labor exchangebureaus per 100,000citizens.

Min

Median

Max


Min

Median

Max


Min

Median

Max


Min

Median

Max


Min

Median

Max


50.00

72.72

88.52

NA

31.25

55.43

73.17

NA

16.67

48.62

74.68

NA

0.103

0.705

2.09

NA

NA

51.51

73.29

87.34

0.21

24

55.9

79.49

0.66*

21.88

49.43

78.05

0.76*

0.14

0.81

2.19

0.92*

0

0.136

0.683

NA

17.71

35.20

56.90

0.61*

6.25

19.81

46.28

0.78*

4.84

16.56

41.79

0.78*

0.09

0.79

2.05

0.87*

0.00

0.17

0.62

0.39*

Indicator Source Measure 2006 2007 2008

Table 6.11: Comparison of Labor Policies (2006-2008)

* Significant at 5% Level; NA = Not ApplicableAll values are at the provincial-level.Labor Sub-Index did not exist in 2005


Figure 6.10: Labor Policies Sub-Index

Firm perceptions of labor policy account for 60% of the index, hard data of bias accounts for 40%.

6.10 Legal InstitutionsFor years, scholars and practitioners have stressedlegal development and formal modes of disputeresolution as the weak link in Vietnam'stransformation.75 Most individuals and private firmsstill opt for informal mechanisms of disputeresolution (see Table 6.12 and Figure 6.11).Tomeasure this, our indicators included:

! The percentage firms that agreed or stronglyagreed that they are confident that theprovincial legal system will uphold contract andproperty rights in business disputes.This is astraightforward measure of firm perceptions ofthe legal environment in their provinces. It isuseful, but not a perfect indicator, becausemost firms have never used the court systemand are only replying based on hearsay. In fact,firms who have not used the system may bethe most likely to respond positively, becausethey have never tried and failed.

! The percentage of firms that believe they canalways or usually go to a higher authority forredress in dealing with overzealous inspectionsor unofficial payments.This indicator is ameasure of whether the local legal system andbureaucracy provides a mechanism for firms toappeal corrupt behavior on the part ofofficials. Once again, this is an interestingindicator, but not perfect, because it relies onthe fact that firms must have alreadyexperienced some type of overzealousbehavior.

! Use of legal institutions. Firms were asked torank their top three modes of disputeresolution.76 As noted above, most firms do

not use formal institutions at all, but wedeveloped an intricate ranking system toreward provinces in which firms do feelcomfortable enough to use these processes.Firms have two primary legal modes to resolvedisputes: Provincial People's Courts andarbitration through the local administrativeapparatus. As the administrative channel opensup possibilities for rent-seeking and selectivebehavior, it receives only half the weight of thecourts channel in the measure.The finalformula is:

Use of Legal Mechanisms =

6* (percent of firms using courts as primarydispute resolution mechanism out of 10listed)

+ 4* (percent of firms using courts assecondary dispute resolution mechanism)

+ 2* (percent of firms using courts as tertiarydispute resolution mechanism)

+ 3* (percent of firms using provincialgovernment as primary dispute resolutionmechanism)

+2* (percent of firms using provincialgovernment as secondary disputeresolution mechanism)

+ 1* (percent of firms using provincialgovernment as tertiary dispute resolutionmechanism)

! Actual usage of Provincial People's Courts is ahard data measure that captures how muchconfidence firms have in the court system by themost direct measure possible: how frequentlythey use it. Filing cases to the provincial courts isnot without costs, so businesses behavingrationally would not submit their cases if theythought it pointless. Using data from theNational People's Court, we calculate thenumber of cases (where claimant was adomestic private firm and not an SOE or foreignfirm) per 100 active domestic private firms andthe percentage of total cases in the provincefiled by private actors


75. Lan Cao and Spencer Weber. 1997. "Law Reform in Vietnam:TheUneven Legacy of Doi Moi," New York University Journal ofInternational Law & Politics Vol. 29, 557-576. AusAid, 2000."Vietnam: Legal and Judicial Development," Working Paper 3. HaNoi,Vietnam, April. Gillespie, John, 2002. "Continuity and Change inVietnamese 'Socialist' Legal Thinking," Law and Governance:Socialist Transforming Vietnam Conference, 1-37, 2002."Transplanted Company Law: An Ideological and Cultural Analysisof Market-Entry in Vietnam," International and Comparative LawQuarterly 51, 641-672.

76. This indicator was based on a survey of Chinese entrepreneurs. SeeTsai, Kellee. 2006. "Capitalists without a Class: Political Diversity AmongPrivate Entrepreneurs in China," Comparative Political Studies 39.



Legal system providedmechanism for firms toappeal officials’ corruptbehavior (% Always orUsually)


Firm confident that legalsystem will upholdproperty rights andcontracts (%StronglyAgree or Agree)

PCI Survey Questions:[6*F10.1(if Court) +4*F10.2(if Court) + 2*F10.3(ifCourt) + 3*F10.1(if ProvincialGov.)+2*F10.2(if ProvincialGov.)+1*F10.3(if ProvincialGov.)]

Use of Legal Institutionsas Primary Modes ofDispute Resolution

People's Supreme Court

Cases filed by by non-state entities at ProvincialEconomic Court per 100firms.

People's Supreme Court

Cases filed by by non-state entities as apercentage of total casesfiled at ProvincialEconomic Court

Min

Median

Max


Min

Median

Max


Min

Median

Max


Min

Median

Max


Min

Median

Max


11.25

23.22

41.46

NA

67.03

83.99

94.32

47.51

94.82

208.87

0

0.41

9.49

0

54.7

100

20

33.74

52.3

--0.26*

66.67

81.2

94.19

0.68*

30.58

64.4

138.89

0.56*

0

0.58

8.12

0.84*

0

50

100

0.49*

19.25

32.74

53.62

0.61*

69.49

81.21

92.68

0.46*

13.33

46.13

82.88

0.38*

0

1.29

6.97

0.32*

0

63.3

100

-0.05

Indicator Source Measure 2006 2007 2008

Table 6.12: Comparison of Legal Institutions (2006-2008)

* Significant at 5% Level; NA = Not ApplicableAll values are at the provincial-level.Legal Sub-Index did not exist in 2005


Figure 6.11: Legal Policy and Institutions Sub-Index

Firm perceptions of legal policy account for 60% of the index, hard data of bias accounts for 40%.

DATASETS USED IN ANALYSIS

DATASETS USED IN ANALYSIS

Association for Small and Medium Size Enterprise Development (ASMED) for Circular 05 DraftingCommittee: 2008. Report on Implementation of the One Stop Shop for Business Registration,Tax Registration,and Stamps and Licenses for business establishment in accordance with the Enterprise Law. Ha Noi,Viet Nam.

General Department of Vocational Training. 2005-2007. Central and Local Vocational Training Centers. Hanoi,Vietnam.

General Statistical Office (GSO). 1995-2007. Statistical Handbook. Hanoi: Statistical Publishing House.

GSO. 1995-2005. The Real Situation of Enterprises Through the Results of Surveys Conducted in 2001, 2002,2003, 2004, 2005, and 2006. Hanoi: Statistics Publishing House.

Labor Exchange Bureau. 2008. Number of Labor Exchange Centers in Provinces. Hanoi,Vietnam.

Ministry of Labor, Invalids, and Social Affairs (MOLISA). 2007. Statistical Yearbook of Invalids and Social Affairs.Hanoi: Labor-Social Publishing House. http://www.molisa.gov.vn/

Ministry of Planning and Investment (MPI). 1995-2008. Registered Firms and Capital. Hanoi,Vietnam.

Ministry of Finance (MOF). 2008. Land Prices in Provincial Capitals. Hanoi,Vietnam.

Ministry of Natural Resources and the Environment (MONRE). 2008. Total Available Land and Land with Land Use Rights Certificate. Hanoi,Vietnam.

Ministry of Trade (MOT). 2006. Trade Fairs in 2007 and Planned for 2008. Hanoi,Vietnam.

People's Supreme Court (PSC). 2005, 2006& 2007. Cases Held in Provincial Economic Courts. Hanoi,Vietnam.

"The Vietnam Provincial Competitiveness Index: Measuring Economic Governance for Private Sector Development." US AID's Vietnam Competitiveness Inititative, Hanoi.

2007 Final Report,Vietnam Competitiveness Initiative Policy Paper #12




APPENDICES


App

endi

x 1A

:M

ultip

le R

egre

ssio

n A

naly

sis o

f Gov

erna

nce

and

Key

Priv

ate

Sect

or O

utco

me V

aria

bles

OLS

Reg

ress

ion;

Robu

st s

tand

ard

erro

rs in

par

enth

eses

;Ro

bust

mod

els

repl

aces

Tel

epho

nes

per

1,00

0 C

itize

ns fr

om 1

995

with

the

200

7 va

lue.

***

p<0.

01,*

* p<

0.05

,* p

<0.

1;ln

indi

cate

s na

tura

l log

tak

en.D

umm

y de

note

s di

chot

omou

s de

pend

ent

varia

bles

.IV

-2SL

S de

note

s an

Inst

rum

enta

l Var

iabl

es -

Tw

o St

aged

Lea

st S

quar

ed p

roce

dure

s,in

stru

men

ting

usin

g th

e pr

ovin

ce's

dist

ance

in k

ilom

eter

s fro

m t

he 1

7th

para

llel (

See

Mal

esky

and

Tau

ssig

200

8 fo

r ju

stific

atio

nof

thi

s in

stru

men

t)

0.32

5***

(0.0

80)

-

0.00

0001

34

(0.0

00)

2.55

7***

(0.4

01)

0.00

252*

(0.0

01)

0.19

8*

(0.1

08)

64 0.84

3

0.13

4

0.33

2***

(0.0

64)

0.00

0017

6

(0.0

00)

2.51

9***

(0.3

23)

0.00

142

(0.0

02)

0.00

787*

*

(0.0

03)

-0.1

77

(0.2

05)

64 0.86

3

0.12

6

0.30

9***

(0.1

09)

0.00

0073

2

(0.0

00)

0.00

0150

(0.0

01)

0.00

687*

*

(0.0

03)

2.72

8***

(0.3

02)

-0.2

51

(0.1

94)

64 0.86

5

0.12

5

0.34

6***

(0.1

20)

0.00

0122

(0.0

00)

-0.0

0282

(0.0

03)

0.02

94*

(0.0

17)

2.52

2***

(0.3

92)

-1.3

17

(0.8

66)

64 0.70

2

0.17

7

2.78

7

3.00

4

0.67

4**

(0.2

96)

-0.0

0041

8

(0.0

00)

3.88

4***

(0.9

42)

0.00

762*

(0.0

05)

0.81

4**

(0.3

60)

64 0.60

0

0.45

5

0.69

0**

(0.2

90)

-0.0

0037

4

(0.0

00)

3.79

6***

(0.9

23)

0.00

508

(0.0

05)

0.01

82*

(0.0

10)

-0.0

542

(0.5

77)

64 0.62

3

0.44

5

0.47

3*

(0.2

72)

-0.0

0026

3

(0.0

00)

0.00

217

(0.0

04)

0.01

59*

(0.0

09)

5.01

5***

(0.9

19)

-0.1

74

(0.5

33)

64 0.67

8

0.41

1

0.44

0

(0.2

78)

-0.0

0030

6

(0.0

00)

0.00

478

(0.0

07)

-0.0

0391

(0.0

41)

5.19

6***

(0.9

82)

0.76

3

(1.9

68)

64 0.65

1

0.40

8

2.78

7

3.00

4

-0.4

12

(56.

024)

-0.0

760

(0.0

47)

54.4

4

(178

.061

)

0.44

3

(0.8

99)

103.

2

(66.

706)

64 0.03

1

115.

4

5.56

9

(62.

830)

-0.0

591

(0.0

52)

20.4

9

(152

.255

)

-0.5

33

(0.9

44)

7.01

2***

(2.4

38)

-231

.0*

(135

.898

)

64 0.16

2

108.

2

57.2

7

(48.

891)

-0.0

665

(0.0

52)

-0.2

59

(0.9

40)

7.22

3***

(2.4

24)

-235

.5

(196

.226

)

-229

.1*

(132

.660

)

64 0.17

6

107.

3

108.

8

(124

.349

)

0.00

0463

(0.1

59)

-4.3

15

(3.1

03)

38.0

2*

(21.

320)

-515

.6

(550

.916

)

-168

4

(104

0.92

5)

64

-2.3

41

205.

8

2.78

7

3.00

4

Inde

pend

ent

Vari

able

s/D

epen

dent

Va

riab

les

Ente

rpri

ses

per

1,00

0 C

itize

ns 2

007

(ln)

OLS (1)

OLS (2)

Robu

st

(3)

IV-2

SLS

(4)

OLS (5)

OLS (6)

Robu

st

(7)

IV-2

SLS

(8)

OLS (9)

OLS

(10)

Robu

st

(11)

IV-2

SLS

(12)

Inve

stm

ent

per

Cap

ita 2

007

(ln)

Prof

it pe

r En

terp

rise

200

7

Nat

iona

l City

Dum

my

Dist

ance

from

Han

oi o

rH

CM

C (

km)

Tele

phon

es p

er 1

,000

Citi

zens

(19

95)

% S

econ

dary

Sch

ool

Gra

duat

es

Unw

eigh

ted

Prov

incia

lC

ompe

titive

ness

Inde

x(2

008)

Tele

phon

es p

er 1

,000

Citi

zens

(200

7)

Con

stan

t

Obs

erva

tions

R-sq

uare

d

Root

Mea

n Sq

uare

d Er

ror

Cra

gg-D

onald

F-S

tatis

tics

ofIn

stru

men

t Stre

ngth

And

erso

n C

anno

nica

lC

orre

latio

n L

R St

atist

ic


0.300**

(0.113)

-0.000401***

(0.000)

1.523**

(0.684)

0.00542**

(0.002)

1.833***

(0.084)

1.388***

(0.164)

64

0.701

0.245

0.305***

(0.114)

-0.000378***

(0.000)

1.490**

(0.621)

0.00400*

(0.002)

0.0102*

(0.005)

1.792***

(0.078)

0.901***

(0.312)

64

0.720

0.239

0.298**

(0.118)

-0.000349***

(0.000)

0.00329

(0.002)

0.00973*

(0.006)

1.575***

(0.519)

1.755***

(0.078)

0.854**

(0.336)

64

0.717

0.240

0.415

(0.366)

-0.000153

(0.000)

-0.00956

(0.009)

0.106

(0.066)

0.878

(1.345)

1.417***

(0.274)

-3.720

(3.261)

64

-1.000

0.604

2.565

2.817

-0.000402***

(0.000)

1.775***

(0.432)

0.00491**

(0.002)

1.632***

(0.047)

0.0435

(0.084)

1.233**

(0.541)

1.420***

(0.247)

64

0.510

0.314

Independent Variables/Dependent Variables OLS

(1)

OLS

(2)

Robust

(3)

IV-2SLS

(4)

OLS

(5)

GDP per Capita 2007 (ln)Interaction Effect

National City Dummy

Distance from Hanoi or HCMC (km)

Telephones per 1,000 Citizens (1995)

% Secondary School Graduates

Unweighted Provincial Competitiveness Index (2008)

Telephones per 1,000 Citizens (2007)

Bia Ria - Vung Tau Dummy

High PCI Dummy (Score above 2006 PCI Median (55.1))

Interaction between High PCI*Standardized Infrastructure

Constant

Observations

R-squared

Root Mean Squared Error

Cragg-Donald F-Statistics of Instrument Strength

Anderson Cannonical Correlation LR Statistic

Appendix 1B: Multiple Regression Analysis of Governance and Provincial GrossDomestic Product

OLS Regression; Robust standard errors in parentheses; Robust models replaces Telephones per 1,000 Citizens from 1995 with the 2007 value.*** p<0.01, ** p<0.05, * p<0.1; ln indicates natural log taken. Dummy denotes dichotomous dependent variables.IV-2SLS denotes an Instrumental Variables - Two Staged Least Squared procedures, instrumenting using the province's distance in kilometers from the17th parallel (See Malesky and Taussig 2008 for justification of this instrument)


0.00688***

(0.001)

0.339**

(0.145)

-0.00200***

(0.001)

-0.350***

(0.020)

0.0267***

(0.002)

YES

YES

192

64

0.990

0.0472

0.0263***

(0.006)

0.718

(0.538)

-0.00777***

(0.002)

-0.786***

(0.082)

0.0694***

(0.005)

YES

YES

192

64

0.983

0.147

0.0106

(1.145)

228.1***

(51.145)

-0.432*

(0.254)

50.97***

(12.722)

4.040***

(1.339)

YES

YES

192

64

0.890

44.75

0.0155***

(0.005)

1.142***

(0.362)

-0.00237**

(0.001)

-0.645***

(0.095)

YES

YES

192

64

0.991

0.0938

Independent Variables/Dependent Variables

GDP perCapita 2007

(ln)

(4)

Profit perEnterprise

2007

(3)

Investment perCapita 2007

(ln)

(2)

Enterprises per1,000 Citizens

2007 (ln)

(1)

Unweighted Provincial Competitiveness Index

Telephones per 1,000 Citizens


Population (1,000s)

GDP (Millions of VND)

Panel Fixed Effects

Year Fixed Effects

Observations

Panels

R-squared


Appendix 1C: Panel Analysis of Governance and Key Private Sector OutcomeVariables (2006–2008)

OLS Regression with Panel Corrected Standard Errors in parentheses, implemented in STATA using XTPCSE. Time period covers all 64 provinces forthree years.*** p<0.01, ** p<0.05, * p<0.1; ln indicates natural log taken.; Dummy denotes dichotomous dependent variables.Panel-Specific AR1 process used to address auto-correlation.% Secondary School Graduates Unweighted Provincial Competitiveness Index (2008)Telephones per 1000 Citizens (2007)


OLS Regression with Panel Corrected Standard Errors in parentheses, implemented in STATA using XTPCSE.Time period covers all 64 provinces for three years.*** p<0.01, ** p<0.05, * p<0.1; ln indicates natural log taken.; Dummy denotes dichotomous dependent variables.Panel-Specific AR1 process used to address auto-correlation.

-46.56***

(3.201)

14.03*

(7.260)

-0.0212***

(0.007)

81.49***

(23.449)

0.0413

(0.053)

82.40***

(12.813)

YES

125

64

0.085

3.804

-47.62***

(3.383)

10.50

(7.435)

-0.0226***

(0.006)

69.03***

(24.677)

0.0393

(0.054)

78.81***

(12.989)

0.0435

(0.084)

1.233**

(0.541)

YES

125

64

0.087

3.762


Interaction

(5)

GDP Growth

Baseline

(1)

Lagged GDP per Capita (ln)

National City Dummy


Telephones per 1,000 Citizens


Bia Ria - Vung Tau Dummy

High PCI Dummy (Score above 2006 PCI Median (55.1))

Interaction between High PCI*Standardized Infrastructure

Year Fixed Effects

Observations

Panels

R-squared


Appendix 1D: Panel Analysis of GDP Growth (2005–2007)


1 Entry Costs

2 Land Access & Security of Tenure

3 Transparency

4 Time Costs of Regulatory Compliance

5 Informal Charges

6 Competition Environment & SOE Bias

7 Proactivity

8 Private Sector Development Policies

9 Labor Policies

10 Legal Institutions

6.55

(0.028)**

0.47

-0.031

3.66

(0.016)**

4.35

(0.020)**

3.10

(0.028)

0.43

(0.038)

2.81

(0.015)*

4.60

(0.019)**

3.85

(0.012)***

1.66

(0.017)

21.77

(0.106)*

0.00

(0.092)

8.40

(0.035)**

3.53

(0.077)

0.00

(0.25)

27.76

(0.108)**

8.65

(0.042)*

15.14

(0.054)**

6.07

(0.044)

-0.30

-0.017

-12.43

(27.42)

64.92

(22.93)***

23.39

(15.16)

16.74

(16.58)

83.79

(34.84)**

11.40

(26.66)

36.80

(11.48)***

42.40

(15.77)***

17.66*

(10.49)

19.49*

-10.7

Sub-IndexProfit per Enterprise

(Millions of VND) New Investment per

Capita (% Change)†

Enterprises per1,000 Citizens (% Change)†

Appendix 1E: Predicted Effects of One-Point Change in Sub-Indices on DependentVariables

(Cross-Sectional Analysis on 2008 Data)

OLS Regression; Robust standard errors in parentheses. All dependent variables are from General Statistical Office 2007*** p<0.01, ** p<0.05, * p<0.1; ln indicates natural log taken.All are results of multiple regression countrolling for National City Dummy, Distance from Hanoi and HCMC (km),Telephones per Capita (1995), andPercentage of Secondary School Graduates. They replicate the second model in Appendix 1A for each dependent variable, replacing the full PCIscore with each sub-index.† Regrression coefficients (") with natural logs are converted to percentages using the following formula (e"-1)*100.


1 Entry Costs

2 Land Access & Security of Tenure

3 Transparency

4 Time Costs of Regulatory Compliance

5 Informal Charges

6 Competition Environment & SOE Bias

7 Proactivity

8 Private Sector Development Policies

9 Labor Policies

10 Legal Institutions

3.10

(0.005)***

5.24

(0.009)***

1.85

(0.001)***

2.47

(0.008)***

4.01

(0.007)***

5.36

(0.013)

1.81

(0.006)***

-3.16

(0.006)***

-1.77

(0.003)***

1.56

(0.006)***

14.00

(0.024)***

19.96

(0.035)***

8.31

(0.009)***

9.72

(0.032)***

9.82

(0.021)***

20.68

(0.047)***

5.88

(0.019)***

-11.13

(0.016)***

-6.14

(0.007)***

5.88

(0.026)**

-4.953

(3.218)

8.964

(3.713)**

-8.743

(3.274)***

-0.877

(5.073)

0.707

(4.147)

13.26

(2.398)***

-2.684

(3.637)

-3.309

(6.323)

5.873

(1.896)***

4.608

(4.969)

Sub-IndexProfit per Enterprise

(Millions of VND) New Investment per

Capita (% Change)†

Enterprises per1,000 Citizens (% Change)†

Appendix 1F: Predicted Effects of One-Point Change in Sub-Indices on DependentVariables

(Panel Analysis on 2006–2008 Data)

OLS Regression with Panel Corrected Stanadard Errors in parentheses. All dependent variables are from General Statistical Office 2007*** p<0.01, ** p<0.05, * p<0.1; ln indicates natural log taken.All are results of replicate the models in the Apendix 2C panel analysis, replacing the full PCI with sub-index scores. All controls are the same.

† Regrression coefficients (") with natural logs are converted to percentages using the following formula (e"-1)*100.


-0.0

0095

8**

(0.0

00)

0.00

0358

(0.0

02)

3.23

2***

(0.2

38)

-0.0

0099

4*

(0.0

01)

-0.0

0116

(0.0

02)

0.00

0605

(0.0

02)

-0.0

0063

2**

(0.0

00)

0.00

103

(0.0

01)

4.02

2***

(0.6

33)

?0.0

0132

*

(0.0

01)

0.00

0371

(0.0

02)

0.00

660

(0.0

09)

-0.0

0046

4

(0.0

00)

0.00

0195

(0.0

01)

-0.0

0000

0568

(0.0

00)

-0.0

0085

6*

(0.0

01)

0.00

875

(0.0

10)

3.36

3***

(0.6

34)

0.00

103*

*

(0.0

01)

?0.0

0227

*

(0.0

01)

-0.0

0185

(0.0

04)

0.00

0082

8

(0.0

00)

-0.0

0030

0

(0.0

01)

-6.9

8e-0

9

(0.0

00)

-0.0

0015

4

(0.0

00)

-0.0

0526

(0.0

05)

1.56

4***

(0.5

28)

(0.0

01)

?0.0

0227

*

(0.0

01)

-0.0

0185

(0.0

04)

0.00

0082

8

(0.0

00)

-0.0

0030

0

(0.0

01)

-6.9

8e-0

9

(0.0

00)

-0.0

0015

4

(0.0

00)

-0.0

0526

(0.0

05)

1.56

4***

(0.5

28)

652

0.00

0013

5

(0.0

00)

-0.0

0057

1

(0.0

00)

-0.0

0100

(0.0

02)

-0.0

0007

56

(0.0

00)

0.00

0081

9

(0.0

00)

0.00

0000

0531

(0.0

00)

-0.0

0010

7*

(0.0

00)

-0.0

0076

7

(0.0

01)

0.00

0025

8

(0.0

00)

-0.0

0045

9

(0.0

00)

-0.0

0060

7

(0.0

01)

0.00

0067

2

(0.0

00)

-0.0

0015

2

(0.0

00)

0.00

0000

0674

(0.0

00)

-0.0

0007

57

(0.0

00)

-0.0

0159

**

(0.0

01)

Inde

pend

ent V

aria

bles

/Dep

ende

ntVa

riab

les

Tota

l Wai

ting

Peri

od fo

r Fi

rm

Regi

stra

tion

in 2

007

Re-R

egist

ratio

n

(Day

s)

Num

ber

of

Lice

nses

Expe

rien

ced

Diff

icul

ties

Long

er th

an a

Mon

th to

Reg

ister

NBR

EG

(1)

NBR

EG

(2)

NBR

EG

(3)

NBR

EG

(4)

NBR

EG

(5)

PRO

BIT

(6)

PRO

BIT

(7)

Tim

e sin

ce O

SS w

as Im

plem

ente

d in

Pro

vince

(day

s)

Perc

enta

ge o

f Firm

s U

sing

OSS

Tota

l Num

ber

of P

rovi

ncia

l Bur

eauc

rats

Ave

rage

Sal

ary

of P

rovi

ncia

l Offi

cial

s

Num

ber

of A

dmin

istra

tive

Uni

ts in

Pro

vinc

e

Dist

ance

from

Han

oi o

r H

CM

C (

km)

Tele

phon

es p

er 1

,000

Citi

zens

(19

95)

% S

econ

dary

Sch

ool G

radu

ates

Con

stan

t

App

endi

x 2A

:Im

pact

of O

ne-S

top

Shop

(OSS

) Im

plem

enta

tion

on B

usin

ess R

egist

ratio

n W

aitin

g Pe

riods


689

64

-276

9

4.04

5

689

64

-276

9

9.65

3

689

64

-276

9

15.9

5

652

64

-145

8

15.0

5

793

64

-145

8

15.0

5

793

64

-423

.5

15.2

7

839

64

-316

.4

67.0

2

Inde

pend

ent V

aria

bles

/Dep

ende

ntVa

riab

les

Tota

l Wai

ting

Peri

od fo

r Fi

rm

Regi

stra

tion

in 2

007

Re-R

egist

ratio

n

(Day

s)

Num

ber

of

Lice

nses

Expe

rien

ced

Diff

icul

ties

Long

er th

an a

Mon

th to

Reg

ister

NBR

EG

(1)

NBR

EG

(2)

NBR

EG

(3)

NBR

EG

(4)

NBR

EG

(5)

PRO

BIT

(6)

PRO

BIT

(7)

Obs

erva

tions

Clu

ster

s

Log

Like

lihoo

d

Chi

-Squ

ared

Mod

els

1-5

are

Neg

ative

Bin

omia

l Reg

ress

ions

to

addr

ess

the

disp

erse

d co

unt

data

inhe

rent

in t

he m

easu

res.

Mod

els

6 an

d 7

are

PRO

BIT

mod

els

with

mar

gina

l pro

babi

litie

s di

splay

edRo

bust

sta

ndar

d er

rors

(cl

uste

red

at p

rovi

nce-

leve

l) in

par

enth

eses

.**

* p<

0.01

,**

p<0.

05,*

p<

0.1

Dep

ende

nt v

aria

ble

(Mod

el 1

-3):T

otal

wai

ting

perio

d fo

r re

gist

ratio

n in

day

s (P

CI Q

uest

ion

C1)

Dep

ende

nt v

aria

ble

(Mod

el 4

):Tot

al w

aitin

g pe

riod

for

re-r

egist

ratio

n in

day

s (P

CI Q

uest

ion

C2)

Dep

ende

nt v

aria

ble

(Mod

el 5

):Tot

al li

cens

es/d

ocum

ents

req

uire

d fo

r re

gist

ratio

n in

day

s (P

CI Q

uest

ion

C3)

Dep

ende

nt v

aria

ble

(Mod

el 6

):Fi

rm w

aitin

g lo

nger

tha

n on

e m

onth

to

be fu

lly le

gal (

dich

otom

ous,

PCI Q

uest

ion

C4)

Dep

ende

nt v

aria

ble

(Mod

el 7

):Fi

rm e

xper

ienc

ed d

iffic

ultie

s in

reg

istra

tion

(di

chot

omou

s,PC

I Que

stio

n C

6)


Confidence in Apellate System (% change)1

Use of Courts in Dispute Settlement (% change)2

Confidence in Legal System (% change)3

National City Dummy

South of 17th Parallel


Change in Number of Foreign Invested Enterprises

Change in GDP per Capita (ln)

Total Number of Cases in Provincial Economic Court

Constant

Observations

R-squared


-5.812

(9.666)

5.989

(13.733)

0.0105

(0.021)

0.956

(3.296)

-198.2

(128.163)

0.00627

(0.008)

21.21

(16.364)

64

0.067

37.31

159.0**

(69.575)

-0.213

(9.580)

7.570

(13.777)

0.00937

(0.021)

0.844

(3.142)

-247.9**

(116.107)

0.000669

(0.008)

21.93

(14.437)

64

0.123

36.65

167.0**

(69.524)

2.671***

(0.892)

-10.50

(10.361)

12.67

(12.345)

0.00770

(0.017)

0.588

(3.272)

-189.9

(120.013)

0.00849

(0.007)

29.71

(18.828)

64

0.269

33.90

125.6*

(62.914)

145.4*

(75.098)

-3.128

(13.355)

5.931

(13.934)

0.00750

(0.021)

1.144

(3.145)

-277.0**

(113.721)

-0.00300

(0.010)

25.81*

(15.227)

64

0.171

36.09

142.7**

(66.879)

2.475**

(0.914)

103.3

(72.104)

-11.81

(12.013)

11.13

(12.639)

0.00650

(0.018)

0.819

(3.209)

-214.9*

(125.952)

0.00531

(0.009)

31.88

(19.699)

64

0.292

33.79

Independent Variables/DependentVariables

Change in Court Cases Filed by Private Firms (%)

OLS

(1)

OLS

(2)

OLS

(3)

OLS

(4)

OLS

(5)

Appendix 2B: Determinants of Change in Economic Court Usage

OLS Regression; Robust standard errors in parentheses.*** p<0.01, ** p<0.05, * p<0.1; ln indicates natural log taken. Dummy denotes dichotomous dependent variables.

1) PCI Question G8: Firms believing they can go to a superior official to appeal unjust treatment (% change)2) PCI Question F9.1-9.3: Firms listing courts as one of three favored dispute settlement mechanisms (% change)3) PCI Question H7.9: Firms expressing confidence in legal system (% Change)


Appendix 3: Determinants of Transport Costs (VND)

OLS Regression; Robust standard errors in parentheses.*** p<0.01, ** p<0.05, * p<0.1Red River Delta is the untested comparative region

Distance from Port (km)

Asphalted Road (% of Total Road)

Northern Uplands

North Central Coast

South Central Coast

Central Highlands

North Southeast

Mekong Delta

Constant

Observations

R-squared

9689***

(2693)

-2708973**

(1836696)

3728029***

(1339595)

-859915

(1318825)

-5887624***

(1840104)

-1351440

(1468793)

97189

(692014)

456377

(922477)

5861267***

(1563455)

64

0.633


OLS

(1)
