1

Productivity and Performance Effects of Business Process Reengineering: A Firm-Level Analysis

Revised - September 2010

KEMAL ALTINKEMERa • YASIN OZCELIKb • ZAFER D. OZDEMIRc

kemal@purdue.edu; yozcelik@fairfield.edu; ozdemir@muohio.edu

a Krannert School of Management, Purdue University, West Lafayette, IN 47907 b Dolan School of Business, Fairfield University, Fairfield, CT 06824

c Farmer School of Business, Miami University, Oxford, OH 45056

Key Words: Business process reengineering, business value of information technology,

productivity, panel regression.

2

Author Biographies

Kemal Altinkemer is an Associate Professor at the Krannert School of Management, Purdue

University. He was a guest editor of special issues of the journals Telecommunication

Systems and Information Technology and Management. He is an Associate Editor in five

journals related to telecommunications, e-commerce and MIS. His research interests are in

design and analysis of computer networks, infrastructure development, distribution of

priorities by using pricing as a tool, infrastructure for e-commerce and pricing of information

goods, bidding with intelligent software agents, strategy from brick-and-mortar to click-and-

mortar business model. He has more than forty publications in prominent journals such as

Operations Research, Operations Research Letters, Management Science, INFORMS

Journal on Computing, Transportation Science, European Journal Operational Research,

Computers and Operations Research, Annals of Operations Research, and more than forty

publications in various conference proceedings. He is a member of AIS, INFORMS and

ACM.

Yasin Ozcelik is an Assistant Professor of Information Systems and Operations Management

at the Dolan School of Business, Fairfield University. He obtained both his Ph.D. in

Management Information Systems and master’s degree in Economics from Purdue

University. His articles have appeared in journals including Journal of Management

Information Systems, Information Systems and E-Business Management, International

Journal of E-Business Research, International Journal of Business Information Systems,

International Journal of Services and Operations Management, International Journal of

Project Management, International Journal of Information Systems in the Service Sector, and

Journal of Global Business Issues. His research was also published in several books and

conference proceedings, including AMCIS, ECIS, MWAIS, INFORMS, and

EURO/INFORMS.

Zafer D. Ozdemir is an Associate Professor at the Farmer School of Business, Miami

University. He received his doctorate from the Krannert School of Management, Purdue

University. His research focuses mainly on economics of e-health and information systems

and has appeared in scholarly journals such as Journal of Management Information Systems,

International Journal of Electronic Commerce, Decision Support Systems, Information &

Management, Communications of the ACM, and Communications of the AIS. In addition, he

has presented his research in several national and international academic conferences.

3

Abstract In this paper, we empirically investigate whether business process reengineering (BPR) is

associated with enhanced firm productivity and overall performance. We analyze firm-level

panel data that covers the period between 1987 and 2008 using fixed effects and first

differencing, which are standard methods to account for unobservable firm-level effects. We

employ standard variables for measuring firm productivity and performance. We find that

one of our key performance variables, return on assets, drops significantly during the project

initiation year. According to fixed effects results, the performance and productivity measures

improve (in a decreasing manner) after project initiation. We also find that enterprise-wide

BPR projects are associated with more negative returns during project initiation. However,

there is no clear evidence about their superiority over functionally-focused BPR projects in

terms of performance improvements after project initiation. This may be because grand

projects are risky and sometimes lead to grand failures.

1. Introduction

Business process reengineering (BPR) is the redesign of processes, typically using

information technology (IT), in order to gain significant improvements in key areas of

performance such as service, quality, cost, and speed. BPR initiatives usually aim to integrate

separate functional tasks into complete cross-functional processes.

During the early 1990s, BPR was consistently ranked as the most important issue for

chief information officers (CIOs) because of the opportunities it was believed to create for

businesses [ 13]. Perceptions toward BPR started to change thereafter, as more executives

expressed skepticism about the success of their BPR initiatives and as researchers reported

disappointing results on those initiatives, especially during mid-1990s. For example, Caron et

al. [ 24] and Murphy [ 95] reported failure rates of 50% and 70%, respectively. Holland and

4

Kumar [ 73] claimed that 60% to 80% of BPR efforts were unsuccessful, while Hammer [ 61]

and Champy [ 25] argued that 70% of companies obtained no benefit from BPR. More

recently, Devaraj and Kohli [ 46] outlined a number of unsuccessful BPR projects.

Although BPR by itself is no longer hotly debated among information systems

scholars, process changes and reengineering continue to enjoy significant popularity and

relevance in practice, as it impacts organizational restructuring efforts, quality improvement

programs, enterprise resource planning (ERP) software implementations, and the redesign of

business processes for e-commerce [ 3, 34, 109, 111]. In fact, according to a 2008 CIO Insight

survey of the IT executives, 36% of the firms consider BPR as the most widely employed

management tool.1 Another 2008 CIO Insight survey revealed that IT executives view

improving business processes as their number one priority.2

69

In addition, a survey by the Ziff

Davis company concluded that business process improvement was the single most important

project type for IT executives in 2007 [ ].

So, why the renewed interest? There seem to be a number of contributing factors.

Managers may be rediscovering BPR as they seek new ways to solidify their business

processes for competitive advantage. Today, more than ever, firms face a rapidly changing

business environment and high consumer expectations. In such an environment, the design

and implementation of sound business processes are extremely important in achieving the

required business performance and flexibility. The need for establishing inter-organizational

links through e-business and supply chains have made BPR an even more strategic tool for

firms [ 2]. In addition, reengineering is necessary to facilitate processes across the

organizational boundaries and to integrate back- and front-office operations [ 53, 84].

1 See http://www.cioinsight.com/c/a/Foreward/Old-Trends-Never-Die/. 2 See http://www.cioinsight.com/c/a/Research/CIOs-Rank-Their-Top-Priorities-for-2008/

5

Since BPR projects typically involve substantial investments, justification of their

expected returns is necessary to acquire funding. In 2002, for example, the research company

Prosci published a report titled “Best Practices in Business Process Engineering and Process

Design” based on a survey of 327 organizations from 53 countries.3

45

According to this report,

73% of the respondents indicated that they met or exceeded their objectives with their BPR

projects – a clear sign of success for BPR. However, the same study also reported that 75% of

the time a business case with a projected return on investment had to be submitted for each

BPR initiative to secure funding. Given the negative publicity that still surrounds BPR,

practitioners would clearly benefit from empirical research that investigates the relationship

between BPR implementations and firm performance, since the results of such research could

provide guidance in estimating and documenting the potential impact of projects. When

outlining future research directions on BPR in 2000, Devaraj and Kohli [ ] commented:

“…the literature in BPR implementation is rife with anecdotal evidence and short on rigorous

empirical evidence of performance impacts of BPR. There is a definite need to better measure

BPR implementations through objective measures, and to relate BPR to organizational

performance in the context of other variables that also affect performance.”

This paper aims to fill this gap. We systematically study the productivity and

performance effects of BPR using a rich data set compiled from primary and secondary

sources. Following Kohli and Devaraj’s [ 81] call to gather larger samples to assess the lagged

effects of IT on productivity, we take a long-term view with a data set that spans 22 years

(1987-2008) for large U.S. firms and investigate the relationship between firm performance

and BPR implementations. We find that one of our key performance variables, return on

assets, drops significantly during the project initiation year. After project initiation, however,

3 See http://www.prosci.com/bpr-benchmarking.htm.

6

various performance and productivity measures improve in a decreasing manner. We also

find that enterprise-wide BPR projects are associated with more negative returns during

project initiation. There is no clear evidence about their superiority over functionally focused

BPR projects in terms of performance improvements after project initiation.

The remainder of the paper is organized as follows. In Section 2, we briefly discuss

BPR and explain its past and contemporary definitions. We then discuss relevant literature in

Section 3, and present our hypotheses in Section 4. We describe our data in Section 5 and

empirical methods in Section 6. Finally, we provide the regression results in Section 7, and

discuss our contributions, limitations, and directions for future research in Section 8.

2. A Brief History of Business Process Reengineering

BPR as a management paradigm first emerged in the early 1990s with the works of

Davenport and Short [ 38], Hammer [ 60, 61], Harrington [ 66], Hammer and Champy [ 65],

Davenport [ 34], Johansson et al. [ 77], Klein [ 80], Dixon et al. [ 47], Manganelli and Klein

[ 86], and Ovans [ 101]. According to the 1993 definition by Hammer and Champy [ 65], BPR

is “the fundamental rethinking and radical redesign of business processes to achieve dramatic

improvements in critical, contemporary measures of performance, such as cost, quality,

service, and speed” (p. 32).

Many early BPR implementations have reported real benefits, ranging from

immediate financial benefits to overall customer satisfaction and growth sustenance. For

example, the CIGNA Corporation successfully completed a number of BPR projects in the

early 1990s [ 88] and realized savings of $100 million by improving its customer service and

quality while reducing operating expenses [ 24]. Similarly, reengineering the accounts payable

process at the Ford Motor Company increased the speed of payments and improved company

relations with long-term suppliers [ 65]. On the other hand, many projects simply failed to

7

deliver tangible results. Reengineering experts argued that such poor outcomes could be due

to expecting too much too soon [ 65], a lack of partnership between IT and business [ 88],

undertaking projects without a complete understanding of costs and benefits, and not exactly

knowing how to redesign a set of related activities [ 7]. Another explanation for the failure is

that seriously troubled companies often consider BPR to be the solution for all their woes

[ 54]. In such cases, however, the problems run far deeper than what can be fixed by

reengineering alone.

The poor outcomes during the 1990s may also be attributed to a poor understanding of

what BPR was really all about. Even the pioneers of reengineering confessed having made

mistakes when defining and promoting BPR. For example, Hammer told the Wall Street

Journal in November 1996 that “[he] was insufficiently appreciative of the human

dimension” [ 124, p.A1]. Similarly, Davenport wrote in the Fast Company in 1995 that “The

rock that reengineering has foundered on is simple: people. Reengineering treated the people

inside companies as if they were just so many bits and bytes, interchangeable parts to be

reengineered. But no one wants to ‘be reengineered’” [ 35, p.70].

Another shortcoming of Hammer and Champy’s approach was that they viewed

radical change as “throwing away the old” which received substantial criticism [ 48] because

certain existing processes could be working well and hence did not need to be discarded. In

2001, Hammer [ 62] asserted that the most important word in the definition of BPR is

“process” rather than “radical”. It therefore seems reasonable to suggest that reengineering

has evolved in the minds of BPR pioneers [see, for example, 37, 62, 63, and 64], and now

represents a continuum of approaches to process transformation. This new view of BPR is

also shared by its critics [ 9, 51, 56, 78, 103, 112, 114, 115]. BPR, as understood today,

focuses on business processes and includes both radical and incremental process changes,

which should be customized to the problem and the context [ 92].

8

Consequently, more recent literature views BPR as a management tool that can aim

for modest yet meaningful process changes. Reflecting the changes in perspectives on BPR,

Hauser and Paper [ 67] proposed in 2007 that Hammer and Champy’s definition be modified

by replacing “radical” with “meaningful”, further emphasizing the notion that the

appropriateness of process changes for the business context is more important than whether

they are drastic or incremental [ 92]. We adopt this modified definition of BPR for the

purpose of this study. Broadening the definition of BPR may be one of the major factors

behind the recent revival of BPR initiatives, albeit with different labels such as “business

process management”. In fact, both the usage of BPR as reported by Bain’s annual

management tool survey and the number of articles published on BPR has increased since

2001, providing evidence that such a revival is indeed the case [ 123].

3. Literature Review

The IS literature on BPR can be grouped into four streams of research [ 119]. The largest

research stream investigates lessons learned from BPR initiatives and outlines critical success

factors [ 15, 22, 24, 29, 36, 49, 52, 58, 68, 85, 97, 109, 115]. Most of the papers in this

category are case studies. Another research stream focuses mainly on interorganizational

aspects of BPR [ 26, 27, 82, 106], while a third stream examines the effectiveness of BPR

methodologies, tools, and techniques [ 33, 42, 43, 72, 79, 98, 99]. The research stream most

relevant to this study, however, is the fourth one that investigates organizational impacts of

BPR.

Some studies in this fourth category aim to explain the relationship between process

changes and IT investments, and thereby provide a rationale as to why BPR, supported by

closely aligned IT investments, generate value for the firm. In 1995, Barua et al. [ 7] proposed

the theory of business value complementarity based on the theory of complementarity that

9

was originally introduced in economics by Edgeworth [ 50]. The notion of complementarity

suggests that the economic benefit of a factor increases with the use of its complementary

factors. In the context of reengineering, IT allows for innovative business processes, new

skills, and new organizational structures [ 18]. Hence, Barua et al. [ 7] and Barua and

Whinston [ 8] argue that IT is complementary with organizational characteristics and

processes, and that investments in IT are less likely to succeed if done in isolation.

The theory of business value complementarity was further supported by research that

has shown that the impact of IT on firm performance is indeed contingent on the coordinated

execution of closely aligned organizational changes. Brynjolfsson and Hitt [ 17] argued that a

significant component of the value of an IT investment is its ability to enable complementary

changes in business processes and work practices of firms, which may eventually lead to

increases in productivity by reducing costs or improving intangible aspects of existing

products, such as timeliness, quality, and variety. Devaraj and Kohli [ 45] showed in 2000 that

IT investments contribute to a higher level of revenue when they are combined with BPR

initiatives. Bresnahan et al. [ 14] examined the effect of three related innovations (IT,

workplace reorganization, and new products and services) on demand for skilled labor, and

found firm-level evidence that the demand for skilled labor was complementary with all three

innovations. Bertschek and Kaiser [ 11] analyzed a cross-sectional data set to understand the

relationship between IT investment, non-IT investment, labor productivity, and workplace

reorganization. They found that workplace reorganization induces an increase in labor

productivity that is attributable to the complementarities it has with IT. Finally, evidence

from the banking industry suggests that the level of impact of IT on bank performance

depends on the extent to which firms support their IT investments with BPR [ 74, 94].

Some other studies on the organizational impact of BPR use mathematical modeling

to understand the conditions and contexts that favor various types of reengineering efforts.

10

Buzacott [ 20] shows analytically that reengineering principles are most relevant when the

task time variability is relatively high, typical of situations where the task involves cognitive

capabilities rather than routine manual procedures. Such situations are more likely to be

found in services rather than manufacturing. Orman [ 100] adopts a decision model paradigm

to quantify the impact of IT on organizational structures and information-processing tasks

and to specify the appropriate reengineering efforts that maximize the benefits from IT. His

approach applies primarily to information-processing tasks that involve decision making

under uncertainty. Seidmann and Sundararajan [ 110] develop an economic model to study the

interplay between IT, employee empowerment, and task consolidation during process

reengineering. They show that, unlike the popular notion in the reengineering literature,

centralized decision making and bundled tasks are superior to other process designs in a time-

based competitive marketplace.

While studies in this fourth research stream have achieved significant milestones in

helping us understand various organizational issues that surround process reengineering, none

of them empirically documented the impact of BPR on organizational performance at the

firm level. In other words, we have a fairly good understanding of the factors that facilitate

BPR and the precursors for BPR success, but we do not know whether BPR projects in

general have a noticeable impact on the productivity and performance of firms as measured

by standard approaches. We contribute to this stream of BPR research by systematically

investigating the productivity and performance effects of IT-driven process changes using a

rich data set. In doing this, we have benefited greatly from previous work on the business

value of IT, in particular research on the firm-level effects of information technology

investments [ 12, 18, 70] and enterprise resource planning implementations [ 71], both in

specifying and refining our empirical models and methods.

11

4. Hypotheses

Assuming that firms behave rationally, we expect firms engaging in BPR to improve their

performance. Firm performance could be measured in various ways, such as labor

productivity, financial performance ratios, and stock market valuation. As discussed in the

next section, we investigate the effect of BPR implementations on a number of performance

measures.

Because of the relatively high costs of BPR, it is possible for a firm to observe a drop

in performance and productivity when initiating a project. The monetary costs of a BPR

project include purchasing new equipment, hiring talented new personnel who are capable of

using the equipment, training existing employees to handle new roles, and payments to

consulting firms if outside expert advice is needed. It has been reported that organizations

implementing BPR projects may have to increase their training budgets by 30 to 50 percent

[ 3]. Projects may also entail non-pecuniary costs. Given that a BPR project modifies existing

business practices of a firm, such organizational changes may lead to an initial turmoil, and

thereby adversely affect firm performance and productivity during the initiation phase of the

project [ 59]. The problems that may adversely affect performance include (i) communications

barriers between functional areas [ 89], (ii) lack of communication between CIOs and CEOs

[ 90] as well as between BPR teams and other personnel [ 55], (iii) resistance from employees

who may potentially be affected by the changes brought about by BPR initiatives [ 108], (iv)

management reluctance to commit resources to BPR projects while expecting quick results

[ 32, 90], and (v) failing to address employee habits during implementation [ 57]. All of these

factors collectively suggest the following hypothesis:

Hypothesis 1. Firms experience a drop in performance and productivity in the year they

initiate BPR projects.

12

Once implementation risks are successfully resolved, employees are likely to become

more comfortable with the new and improved process design, and hence firms will be able to

operate more efficiently. It may take some time for IT investments to translate into

performance effects, partially due to the time required to learn and effectively use new

systems [ 81]. Brynjolfsson et al. [ 19] suggest that the organizational impact of IT investments

is most pronounced two to three years after the IT investment is made, but in some cases the

benefits can be observed sooner than that. Thus, we expect firm performance and

productivity to surpass pre-BPR levels after project initiation, although precisely when this

would happen is likely to depend on the type of project. This expectation is also consistent

with the anecdotal evidence on the positive effects of BPR on firm performance in the

literature, which was detailed in the previous sections. We therefore hypothesize:

Hypothesis 2. Firm performance and productivity starts improving after the project initiation

year.

While we expect firms to derive certain benefits from their BPR implementations,

such benefits may not last indefinitely, or at least their impact on firm performance and

productivity may not be felt as strongly after a while. The literature on business value of IT

suggests that competitive advantages gained from successful implementation of IT are

difficult to sustain and that any financial advantage is short-lived [ 116], since competitors

will eventually adopt comparable technologies to support their processes [ 41]. Exceptions to

this rule are cases when the IT-enabled investment involves scarce resources, leverages

unique processes within the organization, or is dependent on a series of antecedent

investments [ 28]. Although the investments in IT for the purpose of supporting BPR efforts

may be more difficult to imitate since the associated IT implementations potentially enable

unique and proprietary business processes, we would still expect their impact on firm

13

performance and productivity to reduce over time in general. We thus have the following

hypothesis:

Hypothesis 3. The positive impact of BPR projects on firm performance and productivity

decreases over time.

A fourth issue of interest is the effect of project scope on firm performance. The scope

of BPR projects varies considerably; some projects focus on a few business functions, such as

order fulfilment and accounts payable, while others are implemented at the enterprise level.

The scope of these projects may potentially affect the level of impact on firm productivity

and performance. Empirical studies in the literature do not provide conclusive evidence on

the direction of the impact. It may be that BPR projects with a large scope make the highest

possible impact [ 10]. It may also be that the application of BPR across the enterprise does not

actually produce as much benefit as a functionally-focused initiative [ 40]. In order to

understand this issue empirically, we suggest the following hypothesis:

Hypothesis 4. Enterprise-wide BPR has a more negative effect during project initiation year

and a more positive effect thereafter on firm performance and productivity than functionally-

focused BPR.

Finally, the diversity in the functional focus of BPR projects across firms may also

help explain the relationship between BPR and firm performance, since certain areas of the

firm may more readily benefit from process redesign than other areas. Innovation diffusion

literature provides guidance in understanding which functional areas may be associated with

a higher level of improvement in firm performance due to BPR. The primary concern of

innovation diffusion research is how innovations are adopted and why some innovations are

adopted faster than others [ 5]. Rogers [ 107] and Tornatzky and Klein [ 122] identified

“relative advantage”, the degree to which an innovation is perceived as better than the one it

14

supersedes, as one of the most important primary characteristics of innovations that help

explain differences in adoption rates. The higher the relative advantage of an innovation, the

more likely it is to be adopted. We believe that this principle is also at work in the context of

BPR initiatives in that, the higher the relative advantage of engaging in process redesign

targeting a particular functional area, the more frequently it is to be performed. Assuming that

perceptions of BPR teams are rational, we would therefore expect the most frequently

targeted functional areas to produce better outcomes for firms upon reengineering, and thus

suggest the following hypothesis:

Hypothesis 5. The association between firm performance and BPR that impacts the more

frequently targeted functional areas is greater than the average association between firm

performance and BPR.

5. Data

5.1. Primary Data on Business Process Reengineering

Our research focuses on BPR projects conducted by large U.S. firms in the Fortune 500 list.

We obtained the press announcements for these projects from leading news databases,

ABI/INFORM, Lexis/Nexis, Business Source Complete, and Factiva, all of which together

cover all prominent U.S. newspapers, professional magazines as well as business wire news.

In addition, we searched through the annual reports of companies in our sample that were

digitally available in Lexis/Nexis. We carried out a separate search for each company in our

sample. In addition to the company name, we made sure that the word “process” appeared at

least once in every article as well as one or more of the following key words: reengineer,

reorganize, restructure, improve, revamp, overhaul, “process reengineering”, “process

improvement” , “process redesign”, and “redesign process”. The number of articles resulting

from this standard query depended on the popularity of each company. When confronted with

15

too many articles about a company, we narrowed the results by searching within results

where one or more of the additional key words listed above was required to appear in the

articles. In many cases, our standard search method and key words led us to news articles and

SEC filings that discussed valid BPR projects, but we were missing some key details such as

the starting year. We then looked for additional informative key words within the information

sources such as a code name for a project. Using such additional key words in a second round

of search for a specific company proved to be particularly useful in recovering details about a

project.

Every news article that seemed to mention a specific BPR project was considered

carefully. We observed that not all articles referred to such projects as BPR or reengineering.

Regardless of how the projects were named in the news sources, we used the definition we

adopted in Section 2 in qualifying an announcement as a valid BPR project. Two co-authors

of this study independently read each BPR announcement to make sure that the description

and details of the project adhered to the definition of BPR. Given our intent to measure

performance effects over the years, we eliminated announcements that were not clear about

the starting year of the projects. Consistent with the literature, an announcement that

contained information about BPR projects at multiple firms was counted as announcements

of multiple implementations, each relating to one of the firms involved [ 117]. When there

were multiple announcements about a particular project, all of them were used to gather

information about the project’s starting year and nature. We took 1998 as the baseline year

and included every Fortune 500 firm of that year in our search. We used Standard and Poor’s

Compustat database to extract the data for these firms. We removed 10 of these firms from

our panel since they had missing or nonexistent data, reducing our sample size to 490 firms

for the period between 1987 and 2008. Overall, 237 firms (48.4%) were found to have

implemented BPR projects.

16

We also collected information about the focus of the BPR projects in terms of the

functional areas being affected (or targeted) whenever such information was available. BPR

projects in our sample impacted one or more of the following functional areas: accounting,

customer service, finance, human resources, information technology, manufacturing,

operations, purchasing, sales and marketing, and supply chain. Table 1 lists the number of

BPR projects that affected each functional area. We found that manufacturing and operations

was the most frequently impacted function, followed by accounting and finance. Indeed,

these two areas are among the top three areas most impacted by BPR according to a 1994

survey by the consulting firm CSC Index [ 31], with the third area being marketing and sales.

Ranganathan and Dhaliwal [ 105] also report manufacturing and operations to be the most

frequently impacted functional area, whereas the top area according to the consulting firm

Prosci is customer service (this function is ranked third in our study). The similarity between

our findings and the results of previous research lends validity to this study.

[Insert Table 1 here.]

A BPR project was considered to be enterprise-wide when (i) it was explicitly

referred to as being “enterprise-wide” or “company-wide” in the information source, and (ii)

we had sufficient details about the project that showed that the project impacted virtually all

functional areas listed in Table 1. Enterprise-wide projects involved major restructuring and

strategic rethinking of business, ERP implementations coupled with comprehensive process

redesign, or other substantive changes. Most projects, however, impacted either one or two

functional areas, in which case we classified them as being functionally focused. Overall, 37

of the 237 projects in our data set are enterprise-wide, 111 are functionally focused, while 89

are categorized as “unclassified” since we did not have sufficiently clear information about

their scope. Table 2 presents the distribution of BPR projects and their scope by industry.

The industries with the highest percentage of BPR projects are construction and mining,

17

manufacturing, and public administration. On the other hand, the BPR projects implemented

in public administration, finance, insurance, and real estate industries are more likely to have

an enterprise-wide scope.

[Insert Table 2 here.]

5.2. Productivity and Performance Measures

We constructed various measures to calculate labor productivity, financial and operational

firm performance, stock market valuation, and firm output (all serve as dependent variables)

using standard approaches found in previous work on the business value of IT [ 12, 16, 17, 71,

102]. Labor productivity is calculated by dividing total sales by the number of employees.

Financial firm performance is measured by return on assets (ROA) and return on equity

(ROE). Operational performance is measured through inventory turnover, which is the cost of

goods sold divided by the total cost of inventory. Following Hitt et al. [ 71], we measure stock

market valuation of BPR projects by using a simplified version of Tobin’s q that relates

market value of a firm to its assets. Finally, firm output (value added) is calculated by sales

minus materials. Table 3 outlines how these measures and the control variables are

constructed.

Our regression analyses featuring various performance ratios allow us to capture a

wide variety of value creation within the firms. However, some economics and IS researchers

have expressed concern over the ability of these measures to capture future success of the

firm [ 12]. The ability to capture future success is especially important in our context because

it may take several years before investments in IT translate into bottom line performance

effects. Stock market valuations are less prone to this problem since they incorporate current

as well as future tangible and intangible benefits that the firm may receive. Tobin [ 120, 121]

proposed a measure (Tobin’s q) that incorporates stock market valuations to predict firms’

18

future performance, and since then analyses of this measure has been shown to improve

statistical power [ 17]. We, therefore, include Tobin’s q as an additional dependent variable in

our analysis.

[Insert Table 3 here.]

5.3. Control Variables

Our model includes three firm-level control variables: firm size, advertising expenditure, and

market share. Empirical studies in economics, finance, marketing, and strategy have

identified these controls as key determinants of firm performance; see [ 23] and [ 118] for

meta-analytic reviews. As standard in the literature, we use the natural logarithm of the

number of employees as a proxy for firm size. There is ample evidence in the economics,

marketing, and strategy literature supporting a positive relationship between advertising

expenditure and firm performance [ 1, 30, 91, 96, 109]. Finally, market share is included as a

control variable given that efficiency theory [ 21, 39], market power theory [ 87, 113], and

studies on product quality assessment [ 111] provide evidence for a relationship between

market share and firm performance. Jacobson [ 75] and Jacobson and Aaker [ 76] emphasize

further that market share can serve as a proxy for other firm-specific assets not specifically

captured in this study (such as managerial skill). Based on the results of previous research,

we expect a positive relationship between market share and firm performance.

We control for industry at the “1 ½ digit” SIC level to eliminate variation in firm

performance and productivity due to idiosyncratic characteristics of industries. We also use

separate dummy variables for each year to capture transitory, economy-wide shocks that may

affect firm performance [ 71]. The use of such dummies helps remove possible correlation

between macroeconomic trends and firms’ performances during the sample period.

19

5.4. Descriptive Statistics

Table 4 provides descriptive statistics for our data. We deflated the following variables by

using industry-specific implicit price deflators obtained from the Bureau of Labor Statistics,

with the base year being 2002: sales, income, asset, equity, inventory, market value, labor

cost, ordinary capital, value added, and advertising. We deflated cost of goods sold by using

the Producer Price Index Industry Data obtained from the Bureau of Labor Statistics. The

average firm in our sample has a market value of nearly $17 billion and annual sales of

$15 billion. The market share of the average firm is close to three percent, implying that the

market is relatively fragmented. The correlations among the independent variables in each

regression are all significantly below 0.7, suggesting that multicollinearity does not pose a

serious problem to our analyses [ 83].

[Insert Table 4 here.]

6. Empirical Methods

We use the logarithm of the numerator of each performance measure as a dependent variable,

and the logarithm of its denominator as a control variable. This formulation relies on a

property of the logarithm function, and was used commonly in past research [ 71] as it

provides flexibility in the relationship between the numerator and the denominator, while still

retaining the interpretation as a performance measure.4

log (performance measure numerator)it = intercepti + log (performance measure denominator)it

Thus, the general form of the

regression models is:

+ BPR implementation variablesit+ firm controlsit + industry dummiesit + year dummiest + εit (1)

4 This property is log(x/y) = log(x) – log(y).

20

We also perform productivity regressions based on the economic concept of

production functions. Economics literature posits that firms have a production process

(represented by a production function) that relates their output to various inputs. As is

common in the literature, we use the Cobb-Douglas form as our production function, where

value added (sales minus materials) is the output, and capital and labor are the inputs.

Representing the Cobb-Douglas production function in its log-log form yields the following

specification:

log (value added)it = intercepti + log (capital)it + log (labor)it

+ BPR implementation variablesit + industry dummiesit + year dummiest + εit (2)

Let BPRit denote the project initiation year for firm i. The BPR implementation

variables in equations (1) and (2) are functions of BPRit and include a dummy (BPR-D) that

takes a value of zero until BPRit and one thereafter, a linear term (BPR-L) that takes a value of

zero until BPRit and BPR

it t− thereafter, and a quadratic term (BPR-Q) that takes a value of zero

until BPRit and ( )2BPR

it t− thereafter. Recall that Hypothesis 1 is concerned with whether there is

any negative impact on firm performance during project initiation. According to Hypothesis

2, we expect BPR projects to impact firm performance positively as firms adjust to and

leverage their redesigned processes. However, we also expect such impact to reduce over

time per Hypothesis 3. Thus, these three hypotheses suggest a negative coefficient for BPR-

D, a positive coefficient for BPR-L, and a negative coefficient for BPR-Q.5 Taken together,

the BPR implementation variables allow us to examine the hypothesized influence of BPR

projects on firm performance and productivity. Also, BPR-L and BPR-Q jointly capture the

effect of BPR projects on firm performance and productivity over time.

21

Recall that Hypotheses 4 and 5 are about the impact of project scope and functional

focus, respectively. To test these hypotheses, each implementation variable discussed above

is separated into additional variables. For example, consider a project with a functional scope.

BPR-D for projects with an enterprise-wide scope is always zero, while the corresponding

variable for those with a functional scope matches BPR-D for the firm. Other scope-related

implementation variables as well as variables for each functional area are constructed in a

similar fashion. Note that, since Hypothesis 5 is concerned with whether targeting certain

functional areas with BPR is associated with a performance improvement that is above the

average improvement associated with BPR, we include the original implementation and post-

implementation variables in the specification of the panel data regression models when

investigating the effect of projects with a specific functional focus on firm performance and

productivity.

We report the results for both fixed effects (FE) and first differencing (FD), which are

two standard methods that account for the unobservable firm effects (such as organizational

structure and management style) that persist within observations [ 125, pp. 248-252]. If not

accounted for, these unobservable effects correlate with the error terms leading to biased

regression coefficients. FE involves time-demeaning, while FD involves differencing the

data. Similar to FE, FD differences out firm-specific effects but also allows for drift of the

firm-specific coefficient, whereas FE assumes that the firm-specific effects remain constant

over time.6 The choice between FE and FD depends on the relative efficiency of the

estimators. The FE estimator is more efficient if the idiosyncratic errors are serially

uncorrelated, while the FD estimator is more efficient when the errors follow a random walk.

According to Wooldridge [ 125, p. 284], the truth is likely to be somewhere in between. Since

5 A significantly negative sign for BPR-D indicates a drop in firm performance and productivity during the project initiation year. The impact on other early years can be tested separately for 1BPR

it t= + , 2BPRit t= + , etc.

22

differencing introduces autocorrelation in the error terms, we allow for correlation between

the error terms for the adjacent time periods for each firm using xtgee command of Stata.

7. Results and Discussion

In this section, we present our panel data regression results for the specifications presented in

Equations (1) and (2) above (see Tables 5-7). Different measures of firm performance and

productivity are regressed on the implementation variables by controlling for various firm

characteristics as well as time and industry. Please note that, for ease of exposition, we only

report the coefficients and standard errors for the key implementation variables.7

The results of our original specification in Equation (1) are presented in Table 5. As

expected, firm size, advertising expenditure, and market share are generally positively

associated with performance and productivity measures. Similarly, capital and labor are

positively associated with firm output (value added).

7.1. The Effect of BPR on Firm Performance and Productivity

According to the results in Table 5, we find mixed evidence supporting Hypothesis 1. Recall

that a negative coefficient for BPR-D suggests that the related performance measure drops

during the project initiation year. The estimated coefficient of BPR-D is almost always

negative (the only exception is when estimating labor productivity using FE), but statistically

significant only when estimating ROA using FE and when estimating ROA, ROE, and value

added using FD. The F-test for BPR-D + BPR-L + BPR-Q is statistically significant when

estimating ROA and ROE using FD, suggesting that the negative impact is carried into the

second implementation year. Finding a negative association when estimating accounting-

based measures is intuitive in the sense that, since expenses related with BPR efforts and the

6 We thank an anonymous referee for pointing this out.

23

redesigned processes are incurred early on, accounting-based measures can better capture the

impact of such expenses.

Our FE results support Hypothesis 2. The coefficients for BPR-L in all estimations are

significantly above zero, suggesting that firm productivity and performance starts improving

after the project initiation year. The estimated coefficients according to FD results are also all

positive, but none of them are statistically significant. This could be because of the reduced

variation in our explanatory variables as a result of differencing. We should note, however,

that BPR-L by itself suggests the presence of continued improvement in firm performance

and productivity after project initiation into perpetuity. Therefore, it is perhaps more

appropriate to consider the effect of BPR-L jointly with that of BPR-Q.

Hypothesis 3 is about the rate of improvement in firm performance and productivity

associated with process reengineering. In line with the hypothesis, the estimated coefficients

for BPR-Q in all estimations are negative (i.e., performance and productivity improvements

slow down over time) except only when estimating Tobin’s q using FD. These results are

statistically significant when estimating labor productivity, ROA, ROE, and inventory

turnover using FE and when estimating labor productivity and inventory turnover using FD.

Furthermore, BPR-L and BPR-Q are jointly statistically significant in all FE estimations and

when estimating labor productivity and inventory turnover using FD. These results

underscore the importance of time and suggest that firm performance and productivity

increase in a decreasing manner after project initiation.

Based on the estimated FE coefficients, firm performance comes back to pre-BPR

levels in about four months (in the case of inventory turnover) to 3.5 years (in the case of

ROA and value added). Note that labor productivity does not drop even during the project

7 Further details are available from the authors upon request.

24

initiation year, possibly because of the automation of manual activities as part of BPR efforts.

Furthermore, firm performace and productivity is estimated to improve for an extended

period of about eight years (in the case of inventory turnover) to 20 years (in the case of

Tobin’s q). The FD estimates are somewhat similar, with the notable difference that Tobin’s

q is estimated to never decrease after the initiation year.8

[Insert Table 5 here.]

There may be a number of reasons

for the improvement in firm performance for an extended period. It may be that the benefits

derived from BPR through the redesign of processes are in certain ways unique to each

organization and therefore are difficult to imitate by competitors. Another explanation is that

firms in our sample may have continued to implement complementary changes in the

organization (unobservable to us) which also positively affected firm performance, perhaps as

a direct consequence of their BPR projects. Given that in the case of estimating Tobin’s q the

long term post-implementation variable is significantly positive while that for the short term

is not, we are inclined to prefer the latter explanation over the former.

Arguably, firms may be more willing to relay information about BPR projects with

better outcomes, potentially leading to biased estimates. During the project initiation phase,

however, firms are much less informed, although they may still have certain expectations

about potential success in the future. Therefore, we have also collected information about the

announcement date of each project whenever available in order to check whether the results

were sensitive to when the projects were announced. Despite a much smaller sample size, the

results for the restricted sample of BPR projects that were announced at most three months

into project initiation are almost identical to what we have reported for the full sample in

8 Note that, according to the FE results, the coefficient for BPR-L is significantly positive while that for BPR-Q is negative but not statistically significant. According to the FD results, both coefficients are positive but not statistically significant.

25

Table 5. Also, the ordinary least squares (with Huber-White errors clustered by firm) variants

of our analyses yield similar coefficients but with higher standard errors in general.

7.2. The Effect of Project Scope

According to Hypothesis 4, we expect BPR-related benefits and costs to be higher for

enterprise-wide projects than those for functionally focused projects, all else being equal. The

panel data regression results for the formulation described in Equation (2) are presented in

Table 6. The coefficients for the three implementation variables should be viewed as the

average net effect of the implementation of a functionally focused and enterprise-wide BPR

project on the performance and productivity of the firms, respectively.

While the coefficient for BPR-D was significantly negative only when estimating

ROA using FE (see Table 5), the same variable for enterprise-wide projects is significantly

negative when estimating ROA, ROE, Tobin’s q as well as value added. Furthermore, the

difference between BPR-D coefficients for enterprise-wide and functionally focused projects

is statistically significant when estimating ROA, ROE, and value added. These results are

also confirmed by our FD analyses. We also never found BPR-D for functionally focused

projects to be significantly less than that for enterprise-wide projects. Accordingly, our results

support the notion that enterprise-wide BPR initiatives are more costly than those with a

functional focus, possibly because projects with a relatively narrow scope are less likely to

have a noticeable effect on firm performance.

In our FE regressions, the coefficients for BPR-L are positive in all estimations for

both enterprise-wide and functionally oriented projects. The coefficient for enterprise-wide

projects is significantly larger than that for functionally focused projects when estimating

ROA and ROE, although the reverse is true when estimating inventory turnover. A similar

comparison holds for BPR-Q; it is significantly smaller (negative) for enterprise-wide

26

projects than that for functionally focused projects when estimating ROA and ROE, and the

reverse is true when estimating inventory turnover. Comparisons using FD do not yield

statistically significant results. Hence, we do not find compelling evidence for improved firm

performance after project initiation associated with enterprise-wide projects when compared

with functionally oriented projects.

[Insert Table 6 here.]

7.3. The Effect of Project Focus

According to Hypothesis 5, we expect firm performance to be more positively associated

with BPR projects that focus on the most frequently targeted functional areas, all else being

equal. The most frequently impacted functional area in our data is manufacturing and

operations followed by accounting and finance, and then by customer service. Recall that our

specification includes the original implementation variables since we are concerned with

whether targeting certain functional areas with BPR is associated with a performance impact

above the average impact associated with BPR. Therefore, the coefficients for each functional

area should be viewed as the estimated net effect in comparison to a “typical” BPR project.

The regression results are presented in Tables 7a (FE results) and 7b (FD results). Note that

each column in these tables represent seven regressions (one regression for each functional

area).

[Insert Tables 7a and 7b here.]

According to both FE and FD results, the coefficients for BPR-D are significantly

negative when estimating ROE (marketing and sales, and human resources), ROA

(marketing and sales, and human resources), and Tobin’s q (customer service and human

resources). The coefficients for BPR-L are generally positive for manufacturing and

27

operations, accounting and finance, and information technology, but statistically significant

(only FE) when estimating ROE (accounting and finance), labor productivity (information

technology), and inventory turnover (information technology) only. Interestingly, the

coefficients for BPR-Q for information technology are significantly more negative when

estimating labor productivity and inventory turnover using both FE and FD, perhaps because

of the relative ease by which similar process improvements can be replicated by competing

firms. Collectively, however, these results fail to provide conclusive evidence regarding the

relationship between functional focus and firm performance, possibly due to sample size

issues.

8. Conclusion

This paper contributes to the literature on the business value of IT as well as the growing

literature on the value of BPR implementations. In order to investigate the productivity and

performance effects of BPR projects, we consider a variety of measures, including labor

productivity, financial ratios, operational performance, and stock market performance as

measured by Tobin’s q. We also perform productivity regressions based on the economic

concept of production functions. We perform panel data regressions using two standard

methods that account for unobservable firm effects (fixed effects and first differencing) and

explicitly consider the scope and focus of BPR projects in our empirical analyses. We use a

comprehensive data set compiled from primary and secondary sources spanning the period

between 1987 and 2008.

We find that ROA drops significantly during the project initiation year. Also, various

performance and productivity measures improve (in a decreasing manner) after project

initiation according to our fixed effects regressions. This result is similar in spirit to that in

Hitt et al. [ 71] which reports improved firm productivity and performance (at a decreasing

28

rate) after ERP software implementations. Aral et al. [ 4] also report improved firm

performance after, but not during, such implementations. We also find that enterprise-wide

BPR projects are associated with more negative returns (as measured by ROA, ROE, and

value added) during project initiation, although there is no clear evidence about their

superiority over functionally focused BPR projects in terms of performance improvements

after project initiation. This is in line with the observations reported in a relevant case study

[ 40], and we believe that it reflects the risky nature of grand projects, which sometimes lead

to grand failures.

Brynjolfsson et al. [ 19] find that IT investments have the most impact on the firm

after a lag of two to three years, and suggest that because the impacts of new technology are

not fully felt immediately, this may explain why earlier studies found little or no impact of IT

in the same year that the investments are made. In our context, we find some evidence of

negative impact on firm performance in the first year of BPR projects and that in some cases

the impacts are felt immediately, especially in enterprise-wide implementations. From then

on, however, we observe that some of the firm performance measures improve in a

decreasing manner and peak within two to ten years in general.

There are certain limitations to this study. First, our observations are unavoidably

limited to BPR projects that were publicly announced. Thus, in some cases we may have

miscoded the project initiation year or the scope of projects due to missing observations about

unannounced projects or extensions of announced projects. In addition, we would ideally

prefer to have information on BPR-related constructs for each project such as the presence of

cross-functional teams, end-user involvement in process design, etc. in order to accurately

code the details of each project. Although this limitation makes the precise interpretation of

the results difficult, we believe that such measurement errors make our results more

conservative. Such errors would tend to bias our results toward finding no effect of BPR or

29

no difference between enterprise-wide and functionally focused projects due to the random

error in our coding. Second, firms may be more inclined to announce more successful

projects than less successful ones, which may potentially bias the estimates, although our

results on the restricted sample of projects that were announced early on suggest that this may

not be a serious concern. Third, because we have primarily focused on large U.S. firms, our

results may not apply to small and medium-sized businesses and firms around the world.

Fourth, our results capture the effect of BPR averaged over a wide variety of firms and their

projects. Although we report a significant association between improved firm performance

and BPR implementations for certain performance and productivity measures, it is

conceivable that some of these projects may not have led to any performance gains. Finally,

firm performance may be influenced by numerous other factors, many of which are not easily

observable, and these factors may potentially reduce the precision of the estimates or bias the

estimates if their effects are not removed by the time dummies and other control variables.

The utilization of both cross sectional and time series nature of our data helps alleviate this

issue for firm specific factors that are constant over time, since a primary motivation for

using panel data is to address the omitted variables problem [ 125].

There are interesting avenues for future research in this subject area. Effectiveness of

IT applications, especially BPR implementations, may not be uniform across all activities of

a firm. Therefore, our firm-level analysis can be extended to analyze the effects of BPR at the

strategic business unit of firms. This would be possible by defining new productivity and

performance measures for different business units and comparing the resulting differences

across them. Such an analysis may provide further and more specific insights about the

design and value of BPR initiatives to the managers of the individual business units.

Moreover, adopting a process-oriented model may help researchers observe the effects of

30

BPR on other intermediate-level productivity variables, such as capacity utilization and

relative quality [ 6].

ACKNOWLEDGMENTS: We thank Jason Abrevaya (University of Texas), Bill Even

(Miami University), Mehmet Ozbilgin (City University of New York), Sumit Sircar (Miami

University), three anonymous referees, and the Editor-in-Chief, Vladimir Zwass, for their

valuable suggestions and comments. All remaining errors are our own.

References

1. Aaker, D. Managing Brand Equity: Capitalizing on the Value of a Brand Name. New

York: Free Press, 1991.

2. Abdolvand, N., Albadvi, A. and Ferdowsi, Z. Assessing readiness for business process

reengineering. Business Process Management Journal, 14, 4 (2008), 497-511.

3. Al-Mashari, M. Editorial: Business process reengineering is very much well and alive!

Business Process Management Journal, 7, 5 (2001), 370-372.

4. Aral, S., Brynjolfsson, E., and Wu, D.J. Which came first, IT or productivity? The

virtuous cycle of investment and use in enterprise systems. Proceedings of the 27th

International Conference on Information Systems, Milwakuee, WI (2006), 1819-1839.

5. Banker, R.D. and Kauffman, R.J. The evolution of research on information systems: a

fiftieth-year survey of the literature in management science. Management Science, 50, 3

(2004), 281-298.

6. Barua, A., Kriebel, C., and Mukhopadhyay, T. Information technologies and business

value: an analytic and empirical investigation. Information Systems Research, 6, 1

(1995), 3-23.

7. Barua, A., Lee, B., and Whinston, A. The calculus of reengineering. Information

Systems Research, 7, 4 (1996), 409-428.

8. Barua, A. and Whinston, A.B. Complementarity based decision support for managing

organizational design dynamics. Decision Support Systems, 22, 1998, 45-58.

9. Becker, J., Kugeler, M. and Rosemann, M. Process Management: A Guide for the

Design of Business Processes. Berlin: Springer-Verlag, 2003.

31

10. Berry, D., Evans, G.N., Jones, R.M., and Towill, D.R. The BPR scope concept in

leveraging improved supply chain performance. Business Process Management

Journal, 5, 3 (1999), 254-275.

11. Bertschek, I. and Kaiser, U. Productivity effects of organizational change:

microeconometric evidence. Management Science, 50, 3 (2004), 394-404.

12. Bharadwaj, A.S., Bharadwaj, S.G., and Konsynski, B.R. Information technology effects

on firm performance as measured by Tobin’s q. Management Science, 45, 6 (1999),

1008-1024.

13. Brancheau, J., Janz, B., and Wetherbe, J. Key issues in information systems

management: 1994-1995 SIM Delphi results. MIS Quarterly, 20, 2 (1996), 225-42.

14. Bresnahan, T.F., Brynjolfsson, E., and Hitt, L.M. Information technology, workplace

reorganization, and the demand for skilled labor: firm-level evidence. Quarterly

Journal of Economics, 117, 1 (2002), 339-376.

15. Broadbent, M., Weill, P. and St. Clair, D. The implications of information technology

infrastructure for business process redesign. MIS Quarterly, 23, 2 (1999), 159-182.

16. Brynjolfsson, E. and Hitt, L.M. Paradox lost? firm-level evidence on the returns to

information systems spending. Management Science, 42, 4 (1996), 541-558.

17. Brynjolfsson, E. and Hitt, L.M. Beyond computation: information technology,

organizational transformation and business performance. Journal of Economic

Perspectives, 14, 4 (2000), 23-48.

18. Brynjolfsson, E. and Hitt, L.M. Computing productivity: firm level evidence. The

Review of Economics and Statistics, 85, 4 (2003), 793-808.

19. Brynjolfsson, E., Malone, T.W., Gurbaxani, V., and Kambil, A. Does information

technology lead to smaller firms? Management Science, 40, 12 (1994), 1628-1644.

20. Buzacott, J.A. Commonalities in reengineered business processes: models and issues.

Management Science, 42, 5 (May 1996), 768-782.

21. Buzzell, R.D. and Gale, B.T. The PIMS Principles: Linking Strategy to Performance.

New York: Free Press, 1987.

32

22. Candler, J.W., Palvia, P.C., Thompson, J.D. and Zeltmann, S.M. The ORION project:

staged business process reengineering at FEDEX. Communications of the ACM, 39, 2

(1996), 99–107.

23. Capon, N., Farley, J.U., and Hoenig, S. Determinants of financial performance: a meta

analysis. Management Science, 36, 10 (1990), 1143-1159.

24. Caron, J.R., Jarvenpaa, S.L., and Stoddard, D.B. Business reengineering at Cigna

Corporation: experiences and lessons learned from the first five years. MIS Quarterly,

18, 3 (1994), 233-250.

25. Champy, J.A. Reengineering Management: The Mandate for New Leadership. New

York: Harper-Business, 1995.

26. Chatfield, A.T. and Bjørn-Andersen, N. The impact of IOS-enabled business process

change on business outcomes: transformation of the value chain of Japan Airlines.

Journal of Management Information Systems, 14, 1 (Summer 1997), 13-40.

27. Clark, T.H. and Stoddard, D.B. Interorganizational business process redesign: merging

technological and process innovation. Journal of Management Information Systems, 13,

2 (Fall 1996), 9-28.

28. Clemons, E.K. and Row, M.C. Sustaining IT Advantage: The Role of Structural

Differences. MIS Quarterly, 15, 3 (1991), 275-292.

29. Clemons, E.K., Thatcher, M.E. and Row, M.C. Identifying sources of reengineering

failures: a study of the behavioral factors contributing to reengineering risks. Journal of

Management Information Systems, 12, 2 (Fall 1995), 9-36.

30. Comanor, W.S. and Wilson, T.A. Advertising and Market Power. Cambridge, MA:

Harvard University Press, 1974.

31. CSC Index. State of Reengineering Report, 1994.

32. Cummings, J. Reengineering is high on list, but little understood. Network World, 27

July (1992), 27.

33. Datta, A. Automating the discovery of AS-IS business process models: probabilistic and

algorithmic approaches. Information Systems Research, 9, 3 (1998), 275–301.

34. Davenport, T.H. Process Innovation: Reengineering Work through Information

Technology. Boston, MA: Harvard Business School Press, 1993.

33

35. Davenport, T.H. The fad that forgot people. Fast Company, 1 (1995), 70-74.

36. Davenport, T.H. and Beers, M.C. Managing information about processes. Journal of

Management Information Systems, 12, 1 (Summer 1995), 57–80.

37. Davenport, T.H., Prusak, L. and Wilson, H. What's the Big Idea? Creating and

Capitalizing on the Best New Management Thinking. Boston: Harvard Business School

Press, 2003.

38. Davenport, T.H. and Short, J.E. The new industrial engineering: information technology

and business process redesign. Sloan Management Review 31, 4 (Summer 1990) 11-27.

39. Day, G.S. and Montgomery, D.B. Diagnosing the experience curve. Journal of

Marketing, 47, 2 (1983), 44-58.

40. Dean, A.M. Managing change initiatives: JIT delivers but BPR fails. Knowledge and

Process Management, 7, 1 (2000), 11-19.

41. Dehning, B., Richardson, V.J. and Zmud, R.W. The value relevance of announcements

of transformational information technology investments. MIS Quarterly, 27, 4 (2003),

637-665.

42. Dennis, A.R., Daniels, R.M., Jr., Hayes, G. and Nunamaker, J. Methodology-driven use

of automated support in business process re-engineering. Journal of Management

Information Systems, 10, 3 (Winter 1993–94), 117–138.

43. Dennis, A.R., Hayes, G. and Daniels, R.M. Jr. Business process modeling with group

support systems. Journal of Management Information Systems, 15, 4 (1999), 115-142.

44. Dess, G.G., Ireland, R.D., and Hitt, M.A. Industry effects and strategic management

research. Journal of Management, 16, 1 (1990), 7-27.

45. Devaraj, S. and Kohli, R. Information technology payoff in the health-care industry: A

longitudinal study. Journal of Management Information Systems, 16, 4 (2000), 41-67.

46. Devaraj, S. and Kohli, R. The IT Payoff - Measuring the Business Value of Information

Technology Investments. New Jersey: Prentice Hall Press, 2002.

47. Dixon, J.R., Arnold, P., Heineke, J., Kim, J.S. and Mulligan, P. Business process

reengineering: Improving in new strategic directions. California Management Review,

36, 4 (1994), 93-108.

34

48. Earl, M. and Khan, B. How new is business process redesign? European Management

Journal, 12, 1 (1994), 20-30.

49. Earl, M.J., Sampler, J.L. and Short, J.E. Strategies for business process reengineering:

evidence from field studies. Journal of Management Information Systems, 12, 1

(Summer 1995), 31-56.

50. Edgeworth, F.Y. Mathematical Psychics, London: Kegan Paul, 1881.

51. El Sawy, O. Redesigning Enterprise Processes for e-Business. Boston: McGraw-Hill,

2001.

52. El Sawy, O.A. and Bowles, G. Redesigning the customer support process for the

electronic economy: insights from Storage Dimensions. MIS Quarterly, 21, 4 (1997),

457-483.

53. Fadel, K.J. and Tanniru, M. A knowledge-centric framework for process redesign. ACM

SIGMIS CPR Conference on Computer Personnel Research, Atlanta, GA, 2005, 49-58.

54. Fitzgerald, B. and Murphy, C. Business process reengineering: putting theory in to

practice. INFOR, 34, 1(1996), 3-14.

55. Grover, V., Jeong, S.R., Kettinger, W.J., and Teng, J.T.C. The implementation of

business process reengineering. Journal of Management Information Systems, 12, 1

(1995), 109-144.

56. Grover, V. and Kettinger, W. Process Think: Winning Perspectives for Business

Change in the Information Age. Harrisburg, PA: Idea Group, 2000.

57. Grover, V., Teng, J.T.C., and Fiedler, K.D. Information technology enabled business

process redesign: an integrated planning framework. OMEGA, 21, 4 (1993), 433-447.

58. Guha, S., Grover, V., Kettinger, W.J. and Teng, J.T.C. Business process change and

organizational performance: exploring an antecedent model. Journal of Management

Information Systems, 14, 1 (Summer 1997), 119-154.

59. Guimaraes, T. Field testing of the proposed predictors of BPR success in manufacturing

firms. Journal of Manufacturing Systems, 18, 1 (1999), 53-65.

60. Hammer, M. Reengineering work: Don’t automate, obliterate. Harvard Business

Review 68, 4 (July-August 1990) 104-112.

35

61. Hammer, M. Beyond Reengineering: How the Process Centered Organization is

Changing Our Work and Our Lives. New York: Harper Collins, 1996.

62. Hammer, M. The Agenda: What Every Business Must Do to Dominate the Decade.

London: Random House, 2001.

63. Hammer, M. Process management and the future of Six Sigma. Sloan Management

Review, 43, 2 (2002), 26-32.

64. Hammer, M. Deep change: How operational innovation can transform your company.

Harvard Business Review, 82, 4 (April 2004), 84-93.

65. Hammer, M. and Champy, J. Reengineering the Corporation: A Manifesto for Business

Revolution. New York: Harper Business Press, 1993.

66. Harrington, H. Business Process Improvement. New York: McGraw-Hill, 1991.

67. Hauser, K. and Paper, D. Simulation of business re-engineering processes: Case study

of a United States motor manufacturing company. International Journal of

Management, 24, 4 (2007), 676-687.

68. Hengst, M.D. and De Vreede, GJ. Collaborative business engineering: a decade of

lessons from the field. Journal of Management Information Systems, 20, 4 (Spring

2004), 85-113

69. Hertzberg, R. Top 10 projects in ‘07. Baseline, Aug 2007 Supplement, 18-22.

70. Hitt, L.M. Information technology and firm boundaries: Evidence from panel data.

Information Systems Research, 10, 2 (1999), 134-149.

71. Hitt, L.M., Wu, D.J., and Zhou, X. Investment in enterprise resource planning: business

impact and productivity measures. Journal of Management Information Systems, 19, 1

(2002), 71-98.

72. Holden, T. and Wilhelmij, P. Improved decision making through better integration of

human resources and business process factors in a hospital situation. Journal of

Management Information Systems, 12, 3 (Winter 1995-96), 21-41.

73. Holland, D. and Kumar, S. Getting past the obstacles to successful reengineering.

Business Horizons, 38, 3 (1995), 79-85.

36

74. Hunter, L.W., Bernhardt, A., Hughes, K.L., and Skuratowicz, E. It’s not just the ATMs:

firm strategies, work restructuring and workers’ earnings in retail banking. Working

Paper, University of Pennsylvania (2000).

75. Jacobson, R. Unobservable effects and business performance. Marketing Science, 9, 1

(1990), 74-85.

76. Jacobson, R. and Aaker, D.A. Is market share all that it’s cracked up to be? Journal of

Marketing, 49, 4 (1985), 11-22.

77. Johansson, H., McHugh, P., Pendlebury, J.A. and Wheeler, W. Business Process

Reengineering: Breakpoint Strategies for Market Dominance. West Sussex: Wiley,

1993.

78. Kalakota, R. and Robinson, M. Services Blueprint: Roadmap for Execution. Boston:

Addison-Wesley, 2003.

79. Kettinger, W.J., Teng, J.T.C. and Guha, S. Business process change: a study of

methodologies, techniques, and tools. MIS Quarterly, 21, 1 (1997), 55-80.

80. Klein, M.M. IE’s fill facilitator role in benchmarking operations to improve

performance. Industrial Engineering, 25, 9 (1993), 40-43.

81. Kohli, R. and Devaraj, S. Measuring information technology payoff: a meta-analysis of

structural variables in firm-level empirical research. Information Systems Research, 14,

2 (2003), 127-145.

82. Lee, H.G. and Clark, T.H. Market process reengineering through electronic market

systems: opportunities and challenges. Journal of Management Information Systems,

13, 3 (Winter 1996–97), 113-136.

83. Lehmann D.R., Gupta, S. and Steckel, J. Marketing Research, Reading, MA: Addison-

Wesley Educational Publishers, Inc., 1988.

84. Lin, F-R., Yang, M-C. and Pai, Y-H. A generic structure for business process modeling.

Business Process Management Journal, 8, 1 (2002), 19-41.

85. Lucas, H.C. Jr., Berndt, D.J. and Truman, G. A reengineering framework for evaluating

a financial imaging system. Communications of the ACM, 39, 5 (1996), 86-96.

86. Manganelli, R.L. and Klein, M.M. The Re-engineering Handbook: A Step-by-step

Guide to Business Transformation. New York: AMACOM, 1994.

37

87. Martin, S. Market power and/or efficiency?. Review of Economics and Statistics, 70, 2

(1988), 331-335.

88. Martinez, E.V. Successful reengineering demands IS/business partnerships. Sloan

Management Review, 36, 4 (1995), 51-60.

89. McKee, D. An organizational learning approach to product innovation. Journal of

Product Innovation Management, 9, 3 (1992), 232-245.

90. McPartlin, J.P. Seeing eye to eye on reengineering. Information Week, 15 June 1992,

74.

91. Megna, P. and Mueller, D.C. Profit rates and intangible capital. Review of Economics

and Statistics, 73, 4 (1991), 632-642.

92. Melao, N. and Pidd, M. Business processes: Four perspectives. Business Process

Transformation – Advances in Management Information Systems, (eds. Varun Grover

and M. Lynne Markus), New York: M.E. Sharpe, 2008.

93. Melville, N., Gurbaxani, V., and Kraemer, K. The productivity impact of information

technology across competitive regimes: the role of industry concentration and

dynamism. Decision Support Systems, 43, 1 (2007), 229-242.

94. Murnane, R.J., Levy, F., and Autor, D. Technological change, computers and skill

demands: evidence from the back office operations of a large bank. NBER Economic

Research Labor Workshop, (1999).

95. Murphy, E. Cultural values, workplace democracy and organisational change: emerging

issues in European businesses, In Coulson-Thomas, C. (Ed.) (1994) Business Process

Reengineering: Myth & Reality, Kogan Page, London, 201-210.

96. Nelson, P.B. Advertising as information. Journal of Political Economy, 82, 4 (1974),

729-754.

97. Newman, J. and Kozar, K.A. A multimedia solution to productivity gridlock: a re-

engineered jewelry appraisal system at Zale Corporation. MIS Quarterly, 18, 1 (1994),

21-30.

98. Nissen, M.E. Redesigning reengineering through measurement-driven inference. MIS

Quarterly, 22, 4 (1998), 509-534.

38

99. Nissen ME. An experiment to assess the performance of a redesign knowledge system.

Journal of Management Information Systems, 17, 3 (Winter 2000-2001), 25-43.

100. Orman, L.V. A model management approach to business process reengineering.

Journal of Management Information Systems, 15, 1 (1998), 187-212.

101. Ovans, A. Should you take the re-engineering risk? Datamation, 41, 17 (1995), 38-42.

102. Ozcelik, Y. Essays on the effects of information technology and the internet on business

environments. Ph.D thesis, Purdue University (2005).

103. Peppard, J. Broadening visions of business process re-engineering. Omega, 24, 3

(1996), 255-270.

104. Porter, M. Competitive Strategy: Techniques for Analyzing Industries and Competitors.

New York: Free Press, 1980.

105. Ranganathan, C. and Dhaliwal, J.S. A survey of business process reengineering

practices in Singapore. Information & Management, 39 (2001), 125-134.

106. Riggins, F.J. and Mukhopadhyay, T. Interdependent benefits from interorganizational

systems: opportunities for business partner reengineering. Journal of Management

Information Systems, 11, 2 (Fall 1994), 37-57.

107. Rogers, E.M. Diffusion of Innovations. New York: The Free Press, 1983.

108. Ryan, H.W. Managing change. Information Systems Management, 9, 3 (1992), 60-62.

109. Sarker, S., Sarker, S., and Sidorova, A. Understanding business process change failure:

An actor-network perspective. Journal of Management Information Systems, 23, 1

(2006), 51-86.

110. Seidmann, A. and Sundararajan, A. Competing in information-intensive services:

analyzing the impact of task consolidation and employee empowerment. Journal of

Management Information Systems, 14, 2 (1997), 33-56.

111. Sethi, V. and King, W.R. Organizational Transformation through Business Process

Reengineering. Upper Saddle River, NJ: Prentice Hall, 1998.

112. Silver, E. Process management instead of operations management! Manufacturing &

Service Operations Management, 6, 4 (2004), 273-279.

39

113. Smirlock, M., Gilligan, T., and Marshall, W. Tobin’s q and the structure-performance

relationship. American Economic Review, 80, 3 (1984), 618-623.

114. Smith, H. and Fingar, P. Business Process Management: The Third Wave. Tampa:

Meghan-Kiffer Press, 2003.

115. Stoddard, D.B. and Jarvenpaa, S.L. Business process redesign: Tactics for managing

radical change. Journal of Management Information Systems, 12, 1 (1995), 81-107.

116. Stratopoulos, T. and Dehning, B. Does successful investment in information technology

solve the productivity paradox? Information & Management, 38 (2000) 103-117.

117. Subramani, M.R. and Walden, E. The impact of e-commerce announcements on the

market value of firms. Information Systems Research, 12, 2 (2001), 135-154.

118. Syzmanski, D.M., Bharadwaj, S.G., and Varadarajan, P. An analysis of the market

share-profitability relationship. Journal of Marketing, 57, 3 (1993), 1-18.

119. Thong, J.Y.L., Yap, C-S., Seah, K-L. Business process reengineering in the public

sector: the case of the housing development board in Singapore. Journal of

Management Information Systems, 17, 1 (2000), 245-270.

120. Tobin, J. A general equilibrium approach to monetary theory. Journal of Money, Credit

and Banking, 1, 1 (1969), 15-29.

121. Tobin, J. Monetary policies and the economy: the transmission mechanism. Southern

Economic Journal, 37 (1978), 421-431.

122. Tornatzky, L.G. and Klein, K.J. Innovation characteristics and innovation adoption-

implementation: a meta analysis of findings, IEEE Transactions on Engineering

Management, 29 (1982), 28-45.

123. Wang, P. Whatever happened to business process reengineering? Business Process

Transformation – Advances in Management Information Systems, (eds. Varun Grover

and M. Lynne Markus), M.E. Sharpe, 2008.

124. White, J.B. The next big thing. Wall Street Journal, 1996 (November 26), A1.

125. Wooldridge, J.M. Econometric Analysis of Cross-Section and Panel Data. Cambridge,

MA: The MIT Press, 2002.

40

Table 1. Number of projects that impacted each functional area

Industry No of firms with BPR projects

Percent BPR

Percent enterprise-wide

Agriculture, Forestry, and Fishing 0 0% -

Construction and Mining 7 64% 0%

Finance, Insurance, and Real Estate 31 36% 29%

Manufacturing 110 57% 15%

Public Administration 3 60% 33%

Retail Trade 24 46% 13%

Services 10 37% 10%

Transportation, Communications, and Utilities 36 48% 8%

Wholesale Trade 15 47% 27%

Total/Overall 237 48% 27%

Table 2. Distribution of BPR projects across industries

Functional Area Number of BPR Projects

Manufacturing and Operations 74

Accounting and Finance 60

Customer Service 57

Information Technology 56

Supply Chain 49

Marketing and Sales 47

Human Resources 39

41

Variables Description

Performance Measures

Labor productivity Numerator: Sales; Denominator: Number of employees. Higher ratio indicates more productivity per employee.

Return on assets (ROA) Numerator: Pretax income; Denominator: Assets. Higher ratio indicates more profit per unit asset.

Return on equity (ROE) Numerator: Pretax income; Denominator: Equity. Higher ratio indicates a high return accruing to the common shareholders.

Inventory turnover Numerator: Cost of goods sold; Denominator: Inventory. Higher ratio indicates efficient inventory management.

Tobin’s q Numerator: Market value; Denominator: Assets. Higher ratio indicates high stock market return per unit asset.

Value added Firm output measured by sales minus materials

Firm Controls

Firm size Number of employees

Advertising Advertising expense

Market share Annual sales of a firm divided by the total annual sales of all available firms within the firm’s primary industry at the 2-digit Standard Industrial Classification (SIC) level

Table 3. Description of variables

42

Variable Mean Standard Deviation

Sales a 14,948 26,202

Pretax income a 778 2,118

Cost of goods sold b 9,127 17,580

Market value a 17,041 34,548

Assets a 31,938 98,290

Equity a 5,441 10,073

Inventory a 2,275 13,705

Value added a 4,036 7,170

Labor cost a 3,842 5,075

Ordinary capital a 10,172 19,755

Advertising a 513 828

Employees 60,747 106,592

Market share 0.029 0.077

a Values in millions of 2002 dollars b Values in millions of dollars (the base year varies by industry)

Table 4. Descriptive statistics.

41

Firm Performance Firm Productivity

Dependent variable log (sales) log (pretax income)

log (pretax income)

log (cost of goods sold)

log (market value) log (value added)

Interpretation Labor productivity Return on assets (ROA)

Return on equity (ROE) Inventory turnover Tobin's q Value added

FE

BPR-D 0.0055 (0.0142)

-0.0466* (0.0202)

-0.0326 (0.0195)

-0.0057 (0.0148)

-0.0209 (0.0159)

-0.0173 (0.0091)

BPR-L 0.0213*** (0.0039)

0.0158** (0.0054)

0.0205*** (0.0052)

0.0173*** (0.0040)

0.0132** (0.0043)

0.0060* (0.0025)

BPR-Q -0.0012*** (0.0003)

-0.0009* (0.0004)

-0.0009** (0.0003)

-0.0011*** (0.0003)

-0.0003 (0.0003)

-0.0003 (0.0002)

FD

BPR-D -0.0074 (0.0100)

-0.0983*** (0.0304)

-0.0900** (0.0297)

-0.0019 (0.0126)

-0.0085 (0.0163)

-0.0193** (0.0074)

BPR-L 0.0045 (0.0078)

0.0180 (0.0141)

0.0199 (0.0137)

0.0059 (0.0097)

0.0127 (0.0105)

0.0025 (0.0046)

BPR-Q -0.0012* (0.0005)

-0.0012 (0.0009)

-0.0011 (0.0009)

-0.0014* (0.0006)

0.0003 (0.0007)

-0.0002 (0.0003)

*** p<0.001; ** p<0.01; * p<0.05. Numbers in parentheses are corresponding robust standard errors. FE: Fixed effets; FD: First differencing Control variables: Firm size, advertising, market share, industry dummies, and year dummies (performance regressions); capital, labor, industry dummies, and year dummies (productivity regressions).

Table 5. Comparisons of firm performance and productivity

42

Firm Performance Firm Productivity

Dependent variable

log (sales) log (pretax

income) log (pretax

income) log (cost of goods sold)

log (market value)

log (value added)

Interpretation Labor

productivity Return on

assets (ROA) Return on

equity (ROE) Inventory turnover Tobin's q Value added

FE

Functional

BPR-D -0.0471** (0.0182)

-0.0193 (0.0254)

-0.0225 (0.0244)

-0.0340 (0.0187)

-0.0221 (0.0205)

0.0061 (0.0114)

BPR-L 0.0253*** (0.0053)

0.0041 (0.0073)

0.0026 (0.0070)

0.0278*** (0.0053)

0.0147* (0.0059)

0.0075* (0.0033)

BPR-Q -0.0020*** (0.0004)

-0.0004 (0.0005)

-0.0002 (0.0005)

-0.0020*** (0.0003)

-0.0006 (0.0004)

-0.0001 (0.0002)

Enterprise-wide

BPR-D -0.0242 (0.0312)

-0.1568*** (0.0451)

-0.1502*** (0.0440)

0.0224 (0.0348)

-0.0748* (0.0354)

-0.0632** (0.0213)

BPR-L 0.0204* (0.0085)

0.0368** (0.0120)

0.0452*** (0.0117)

0.0061 (0.0093)

0.0166 (0.0095)

0.0136* (0.0056)

BPR-Q -0.0011* (0.0006)

-0.0024** (0.0008)

-0.0028*** (0.0007)

0.0002 (0.0006)

-0.0007 (0.0006)

-0.0009* (0.0004)

FD

Functional

BPR-D -0.0047 (0.0138)

-0.0117 (0.0411)

-0.0066 (0.0397)

-0.0009 (0.0171)

-0.0035 (0.0223)

-0.0067 (0.0100)

BPR-L 0.0199* (0.0095)

0.0105 (0.0171)

0.0023 (0.0165)

0.0248* (0.0114)

0.0092 (0.0127)

0.0074 (0.0055)

BPR-Q -0.0020** (0.0006)

-0.0003 (0.0012)

-0.0001 (0.0011)

-0.0023** (0.0007)

-0.0002 (0.0008)

-0.0005 (0.0004)

Enterprise-wide

BPR-D -0.0045 (0.0240)

-0.3230*** (0.0744)

-0.3138*** (0.0725)

-0.0108 (0.0311)

-0.0574 (0.0400)

-0.0501** (0.0185)

BPR-L -0.0004 (0.0161)

0.0472 (0.0343)

0.0467 (0.0331)

-0.0118 (0.0191)

0.0023 (0.0218)

0.0191 (0.0100)

BPR-Q -0.0003 (0.0011)

-0.0031 (0.0027)

-0.0029 (0.0026)

0.0007 (0.0013)

0.0007 (0.0015)

-0.0014* (0.0007)

*** p<0.001; ** p<0.01; * p<0.05. Numbers in parentheses are corresponding robust standard errors. FE: Fixed effets; FD: First differencing Control variables: Firm size, advertising, market share, industry dummies, and year dummies (performance regressions); capital, labor, industry dummies, and year dummies (productivity regressions).

Table 6. Comparisons of firm performance and productivity by project scope

43

Firm Performance Firm Productivity

Dependent variable log (sales) log (pretax

income) log (pretax

income) log (cost of goods sold)

log (market value)

log (value added)

BPR-D

M&O -0.0163 (0.0259)

-0.0668 (0.0369)

-0.0536 (0.0357)

0.0122 (0.0273)

-0.0852** (0.0291)

0.0109 (0.0168)

A&F 0.0065 (0.0278)

-0.0755 (0.0395)

-0.0795* (0.0382)

0.0464 (0.0302)

-0.0123 (0.0310)

-0.0499** (0.0184)

CS -0.0180 (0.0284)

-0.0607 (0.0406)

-0.0519 (0.0393)

0.0401 (0.0298)

-0.0039 (0.0318)

-0.0466* (0.0185)

IT -0.0899** (0.0286)

-0.0684 (0.0412)

-0.0607 (0.0399)

-0.0731* (0.0309)

-0.0424 (0.0322)

-0.0578** (0.0189)

SC -0.0271 (0.0294)

-0.0524 (0.0419)

-0.0547 (0.0406)

0.0376 (0.0332)

-0.0654 (0.0335)

-0.0299 (0.0193)

M&S -0.0075 (0.0305)

-0.1016* (0.0441)

-0.1120** (0.0429)

0.0027 (0.0327)

-0.0064 (0.0341)

-0.0173 (0.0201)

HR -0.0123 (0.0327)

-0.1158* (0.0469)

-0.1156* (0.0456)

0.0397 (0.0359)

-0.0543 (0.0368)

-0.0564** (0.0220)

BPR-L

M&O 0.0003 (0.0076)

0.0162 (0.0107)

0.0126 (0.0103)

-0.0074 (0.0079)

0.0157 (0.0085)

0.0063 (0.0049)

A&F 0.0052 (0.0080)

0.0198 (0.0111)

0.0282** (0.0108)

-0.0062 (0.0085)

0.0120 (0.0089)

0.0089 (0.0052)

CS -0.0078 (0.0080)

0.0123 (0.0111)

0.0122 (0.0108)

-0.0259** (0.0081)

-0.0005 (0.0088)

0.0170** (0.0051)

IT 0.0389*** (0.0081)

0.0078 (0.0115)

0.0102 (0.0111)

0.0472*** (0.0085)

-0.0046 (0.0090)

0.0092 (0.0053)

SC -0.0005 (0.0085)

0.0088 (0.0119)

0.0118 (0.0115)

-0.0141 (0.0091)

0.0072 (0.0095)

0.0058 (0.0055)

M&S -0.0066 (0.0089)

0.0264* (0.0126)

0.0326** (0.0122)

-0.0160 (0.0094)

0.0028 (0.0098)

0.0079 (0.0057)

HR -0.0014 (0.0091)

0.0288* (0.0128)

0.0369** (0.0124)

-0.0162 (0.0098)

0.0087 (0.0101)

0.0106 (0.0060)

BPR-Q

M&O 0.0008 (0.0005)

-0.0012 (0.0007)

-0.0009 (0.0007)

0.0010 (0.0005)

-0.0007 (0.0006)

-0.0005 (0.0003)

A&F -0.0001 (0.0005)

-0.0014 (0.0008)

-0.0019* (0.0007)

0.0011 (0.0006)

-0.0007 (0.0006)

-0.0004 (0.0003)

CS 0.0009 (0.0005)

-0.0010 (0.0007)

-0.0010 (0.0007)

0.0023*** (0.0005)

-0.0001 (0.0006)

-0.0012*** (0.0003)

IT -0.0034*** (0.0005)

-0.0007 (0.0008)

-0.0010 (0.0007)

-0.0032*** (0.0006)

0.0003 (0.0006)

-0.0007* (0.0003)

SC 0.0005 (0.0006)

-0.0008 (0.0008)

-0.0009 (0.0008)

0.0015* (0.0006)

-0.0004 (0.0006)

-0.0006 (0.0004)

M&S 0.0007 (0.0006)

-0.0019* (0.0008)

-0.0022** (0.0008)

0.0019** (0.0006)

-0.0003 (0.0007)

-0.0008* (0.0004)

HR 0.0004 (0.0006)

-0.0020* (0.0008)

-0.0024** (0.0008)

0.0016* (0.0007)

-0.0005 (0.0007)

-0.0009* (0.0004)

M&O: Manufacturing and operations; A&F: Accounting and finance; CS: Customer service; IT: Information technology; SC: Supply chain; M&S: Marketing and sales Control variables: Firm size, advertising, market share, industry dummies, and year dummies (performance regressions); capital, labor, industry dummies, and year dummies (productivity regressions).

Table 7a. Regression results for the effect of functional focus (fixed effects)

44

Firm Performance Firm Productivity

Dependent variable log (sales) log (pretax

income) log (pretax

income) log (cost of goods sold)

log (market value)

log (value added)

BPR-D

M&O -0.0129 (0.0201)

-0.0245 (0.0613)

-0.0356 (0.0597)

-0.0179 (0.0248)

-0.0178 (0.0328)

-0.0066 (0.0148)

A&F 0.0107 (0.0216)

-0.1044 (0.0652)

-0.0976 (0.0627)

0.0028 (0.0281)

-0.0379 (0.0354)

-0.0125 (0.0166)

CS -0.0017 (0.0219)

-0.1516* (0.0645)

-0.1253* (0.0630)

-0.0097 (0.0272)

-0.0259 (0.0359)

-0.0444** (0.0162)

IT 0.0052 (0.0220)

-0.2181** (0.0656)

-0.1945** (0.0640)

-0.0098 (0.0283)

-0.0223 (0.0359)

-0.0308 (0.0166)

SC 0.0044 (0.0228)

-0.1345* (0.0684)

-0.1253 (0.0665)

-0.0120 (0.0294)

-0.0433 (0.0383)

-0.0280 (0.0171)

M&S 0.0018 (0.0235)

-0.2245** (0.0751)

-0.2362** (0.0732)

-0.0114 (0.0294)

-0.0293 (0.0388)

-0.0256 (0.0176)

HR -0.0030 (0.0256)

-0.2550*** (0.0771)

-0.2491*** (0.0751)

-0.0177 (0.0327)

-0.0563 (0.0424)

-0.0390* (0.0195)

BPR-L

M&O -0.0033 (0.0132)

0.0174 (0.0262)

0.0150 (0.0253)

-0.0171 (0.0157)

-0.0082 (0.0179)

0.0134 (0.0079)

A&F 0.0002 (0.0142)

0.0040 (0.0278)

0.0069 (0.0266)

-0.0103 (0.0170)

0.0049 (0.0191)

0.0163 (0.0087)

CS -0.0131 (0.0141)

0.0095 (0.0257)

-0.0088 (0.0251)

-0.0229 (0.0166)

-0.0071 (0.0190)

0.0205* (0.0084)

IT 0.0271 (0.0142)

0.0353 (0.0270)

0.0234 (0.0263)

0.0262 (0.0171)

-0.0145 (0.0192)

0.0091 (0.0087)

SC -0.0097 (0.0150)

-0.0092 (0.0296)

-0.0142 (0.0286)

-0.0242 (0.0180)

-0.0001 (0.0205)

0.0117 (0.0090)

M&S -0.0107 (0.0158)

0.0123 (0.0337)

0.0125 (0.0326)

-0.0173 (0.0185)

-0.0016 (0.0213)

0.0119 (0.0096)

HR -0.0109 (0.0168)

0.0327 (0.0340)

0.0351 (0.0328)

-0.0258 (0.0199)

-0.0048 (0.0227)

0.0160 (0.0103)

BPR-Q

M&O 0.0014 (0.0009)

-0.0003 (0.0019)

-0.0005 (0.0019)

0.0022* (0.0010)

0.0005 (0.0012)

-0.0008 (0.0005)

A&F 0.0008 (0.0010)

-0.0001 (0.0020)

-0.0001 (0.0019)

0.0016 (0.0011)

0.0001 (0.0013)

-0.0009 (0.0006)

CS 0.0011 (0.0009)

-0.0002 (0.0017)

0.0008 (0.0017)

0.0021 (0.0011)

0.0009 (0.0013)

-0.0014* (0.0005)

IT -0.0039*** (0.0010)

-0.0032 (0.0020)

-0.0028 (0.0019)

-0.0031** (0.0011)

0.0011 (0.0013)

-0.0006 (0.0006)

SC 0.0010 (0.0010)

0.0006 (0.0023)

0.0010 (0.0022)

0.0021 (0.0012)

0.0002 (0.0014)

-0.0011 (0.0006)

M&S 0.0011 (0.0011)

-0.0014 (0.0027)

-0.0012 (0.0026)

0.0018 (0.0013)

0.0003 (0.0015)

-0.0011 (0.0007)

HR 0.0010 (0.0012)

-0.0029 (0.0026)

-0.0023 (0.0025)

0.0022 (0.0013)

0.0005 (0.0016)

-0.0015* (0.0007)

M&O: Manufacturing and operations; A&F: Accounting and finance; CS: Customer service; IT: Information technology; SC: Supply chain; M&S: Marketing and sales Control variables: Firm size, advertising, market share, industry dummies, and year dummies (performance regressions); capital, labor, industry dummies, and year dummies (productivity regressions).

Table 7b. Regression results for the effect of functional focus (first differencing)