Botswana Research Bulletin

8/7/2019 Botswana Research Bulletin

1/45

THE R ESEARCH B ULLETINN OVEMBER 2009

Volume 23 No. 1

R ESEARCH D EPARTMENT

B ANK OF B OTSWANA


2/45

THE RESEARCH BULLETINNOVEMBER 2009

RESEARCH D EPARTMENTBank of Botswana Volume 23 No. 1


3/45

The Research Bulletin, November 2009, Volume 23, No 1

Published by The Research Department, Bank of BotswanaP/Bag 154, Gaborone, Botswana.

ISSN 1027-5932

This publication is also available at the Bank of Botswana website:www.bankofbotswana.bw

Copyright Individual contributors, 2009

Typeset and designed byLentswe La Lesedi (Pty) Ltd Tel: 3903994, Fax: 3914017, e-mail:[email protected]

Printed and bound byPrinting and Publishing Company Botswana (Pty) Ltd


4/45

Contents

2009 Monetary Policy Statement Mid-Term Review 1 Bank of Botswana

The Importance of Sound Metadata Management for the Productionof Quality Statistics Botswanas Case 9Sabata Legwaila

Overview of Initiatives to Promote Convergence in the Contextof Regional Integration: An African Perspective 15

K.S. Masalila

Bank of Botswanas Short-term Ination Forecasting Models 2 7 James, L.V., Mokoti, T.P., and Molalapata, I.


5/45


6/45

The purpose of The Research Bulletin is to provide a forum where research relevant to theBotswana economy can be disseminated. Although produced by the Bank of Botswana, theBank claims no copyright on the content of the papers. If the material is used elsewhere, anappropriate acknowledgement is expected.

Additional copies of the Research Bulletin are available from The Librarian of the ResearchDepartment at the above address. A list of the Banks other publications, and their prices, aregiven below.

Bank of Botswana Publications

SADC Rest of Domestic Members the World

1. Research Bulletin (per copy) P11.00 US$5.00 US$10.002. Annual Report (per copy) P22.00 US$10.00 US$15.003. Botswana Financial Statistics Free US$25.00 US$40.00

(annual: 12 issues)4. Aspects of the Botswana Economy: P82.50 US$29.00 US$42.00

Selected Papers

Please note that all domestic prices cover surface mail postage and are inclusive of VAT. Otherprices include airmail postage and are free of VAT. Cheques, drafts, etc., should be drawn infavour of Bank of Botswana and forwarded to the Librarian, Bank of Botswana, Private Bag 154,Gaborone, Botswana.

Comments: The Bank would welcome any comments/suggestions on papers published onthe bulletin. Such communications should be addressed to:

The DirectorResearch DepartmentBank of Botswana Private Bag 154Gaborone

Or, e-mail to [email protected]


7/45


8/45

1

2009 M ONETARY P OLICY S TATEMENT M ID - TERM R EVIEW

2009 Monetary PolicyStatement Mid-TermReview

1. I NTRODUCTION1.1 The Mid-Term Review (MTR) of the 2009 Mon-etary Policy Statement (MPS) evaluates progress inachieving the 3 6 percent ination objective; it alsoassesses the medium-term nancial and economicconditions and other factors that are likely to inu -ence the ination outlook and policy formulation inthe remainder of the year. Overall, the assessmentinforms stakeholders of the key economic develop-ments, and consequently the appropriate monetary policy stance, with a view to fostering public expecta-tions of low, sustainable and predictable ination.

1.2 The medium-term ination objective of 3 6percent represents the Banks view of price stability and is considered to be consistent with sustain-able long-run growth of the economy. In additionto promoting growth, low ination that is close tothat prevailing in the countrys trading partnerscontributes to attaining stability of the real effectiveexchange rate (REER).

1.3 As projected at the beginning of the year, in -ation fell in the rst half of 2009, in the contextof weaker output growth and lower fuel prices, anddecreased from 13.7 percent in December 2008 to 7percent in June 2009. 1 Given the outlook for benignprice pressures over the medium-term and the riskof a decline in output expansion emanating from thedeterioration in global economic performance, theBank reduced the Bank Rate by a cumulative 400

basis points in the same period (December 2008 June 2009). Looking ahead, ination is expectedto fall within the objective range in the short termand, subject to the balance between the assessmentof expected downward trajectory and the upsiderisks, it allows scope for monetary policy to remainsupportive of the envisaged recovery of GDP.

1.4 While world economic performance continuesto be weak, with GDP growth for 2009 projected at

1.4 percent (3.1 percent in 2008), the monetary andscal stimuli as well as liquidity and nancial sup -port to banks and industry across the world appearto underpin the imminent global economic recovery.The world economy is, therefore, forecast to expand

by a faster (albeit low) rate of 2.5 percent in 2010. The improved performance, however, still largely depends on the resumption of normal nancial inter -mediation under enhanced supervisory standards,continuation of open and free international trade

and support for vulnerable economies.

1 Ination fell further, to 6 percent, in July 2009.

2. M ONETARY POLICY FRAMEWORK 2.1 The monetary policy objective is anchoredon the attainment of price stability, as dened by the medium-term objective range of 3 6 percent.

A low and predictable level of ination contributesto the broader national objectives of sustainablegrowth and development through promoting sav-ings mobilisation, fostering productive investmentand facilitating international competitiveness of thedomestic industry.

2.2 The Banks policy framework entails a forecast- based and forward-looking monetary policy strategy with a medium-term focus. The medium term is de-ned as a three-year rolling period, which is consid -ered a reasonable time frame over which monetary policy can affect price developments. The policy isguided by the medium-term forecast for ination, asderived from an assessment of prospective develop-ments for various determinants of ination. Among the inuences on ination are demand conditions,changes in administered prices and consumptiontaxes, as well as public expectations with respectto the rate of price changes. The Bank uses inter-est rates and open market operations to inuencedemand and, ultimately, price developments in thedesired direction. However, the policy response toinationary pressures is premised on an assessmentof the sources of ination. In particular, there is a distinction between factors with a transitory impact,associated with changes in administered prices andconsumption taxes, and those likely to have a lasting

inuence on ination, such as changes in demandconditions. Meanwhile, the alternative measures of ination, including headline, the 16 percent trimmedmean and ination excluding administered prices,are indicative of the sources of ination and serveto explain price developments.

2,3 Operationally, the policy framework encom-passes regular meetings of the Monetary Policy Committee, which reviews ination forecasts andthe monetary policy stance to take account of ongoing changes in economic circumstances thataffect the outlook for the determinants of ination.Therefore, the Bank is able to respond in a timely manner to prospective developments that would leadto a signicant (and lasting) deviation of inationfrom the objective range. High and volatile inationis detrimental to economic growth as it discour-ages nancial saving and generates uncertainty for investment decisions; and it quickly erodes thepurchasing power of incomes, hence reducing living standards. On the other hand, sustained periods of low or rapidly falling ination could be indicative of a decline in economic performance, which could re-quire monetary policy easing to stimulate growth.

2.4 Achieving the Banks ination objective alsocontributes to stabilisation of the real effective ex-change rate (REER), which helps to promote interna-tional competitiveness of domestic industries. In the


9/45

2

B ANK OF B OTSWANA

context of the crawling bandexchange rate arrangement,and in the event the inationobjective remains higher thanination in trading partnercountries, the Bank imple-ments a continuous adjust -ment of the nominal effectiveexchange rate to maintainstability of the REER.

2.5 An integral part of for-mulating and implementing monetary policy is commu-nicating the Banks frame-

work to stakeholders, as well as apprising the publicon economic developments,inflation outlook and thefactors considered in deter-mining the monetary policy stance. Such communica-tion is done through, among others, the Annual Reports,Monetary Policy Statements(and mid-term reviews), eco-nomic briengs and pressreleases following Monetary Policy Committee meetings.In addition to the duty of accountability, transpar-ency enhances public un-derstanding of the policy

framework, which promotespolicy credibility and thedegree to which the Bankcould succeed in inuencing inflation expectations andmaintaining price stability.

3. I NFLATION IN THE FIRST HALF OF 2009

3.1 Ination maintained a downward trend in therst half of 2009, largely reecting the impact of changes in fuel prices. In addition to the decreasein fuel prices, the effect of the large fuel price in-creases in the rst half of 2008 dropped out of theination calculation. Overall, ination declinedfrom 13.7 percent in December 2008 to 7 percentin June 2009, but was above the upper end of theBanks medium-term ination objective range of 3

6 percent. 2 Both measures of core ination (16percent trimmed mean and core ination exclud -ing administered prices) also decreased from 12.1percent to 7.5 percent and from 16.1 percent to13.8 percent, respectively. The higher level for the

latter is due to the absence of the impact of the fuelprice decrease in this measure.

2 Ination fell to 6 percent in July and, therefore, isequal to the upper end of the objective range.

3.2 Average ination for the trading partnercountries also trended downwards, falling from

6.5 percent in December 2008 to 3.9 percent inJune 2009, against the background of weak globaleconomic performance and the substantial declinein international commodity and oil prices in thesecond half of 2008. In South Africa, headline in-ation, 3 which is now the target measure for theSouth African Reserve Bank (SARB), eased from9.5 percent in December 2008 to 6.9 percent in

3 The SARB replaced the CPIX with the headlineination as the target measure of ination at the

beginning of the year and also revised the CPIX.Headline ination is compiled from all the primary and secondary urban areas in which CPI data arecollected. The revised CPIX excludes owners rent

which is based on rentals paid for similar residen-tial houses as those found in the owner-occupiermarket, while the old CPIX excluded interest rateson mortgage bonds from the headline inflation.

CHART 1: B OTSWANA INFLATION (J ANUARY 2004 - J UNE 2009)

4

6

8

10

12

14

16

18

2 0 0 4

A p r

J u l y

O c t

2 0 0 5

A p r

J u l

O c t

2 0 0 6

A p r

J u l

O c t

2 0 0 7

A p r

J u l

O c t

2 0 0 8

A p r

J u l

O c t

2 0 0 9

A p r

P e r c e n t

CPI InflationCore Inflation (16 percent trimmed mean)Core Inflation by exclusion

CHART 2: I NTERNATIONAL INFLATION (J ANUARY 2004 - J UNE 2009)

-2

0

2

4

6

8

10

12

14

16

18

2 0 0 4

A p r

J u l

O c t

2 0 0 5

A p r

J u l

O c t

2 0 0 6

A p r

J u l

O c t

2 0 0 7

A p r

J u l

O c t

2 0 0 8

A p r

J u l

O c t

2 0 0 9

A p r

P e r c e n t

SA (CPIX) Botswana

SDR countries Trading partners


10/45

3


June 2009, due to lowerrates of increase for energy and food prices. Neverthe-less, South African ination

was above the target rangeof 3 6 percent in the rsthalf of 2009. For the SDRcountries, which includeeconomies that were inrecession in the rst half of the year, ination decel -erated from 0.7 percent to

0.8 percent.

3.3 On the domestic front,the previous years upwardinuence of fuel and foodprice increases on ination

was substantially reduced in2009 (Chart 3). Notably, theannual change in the costof fuel decelerated from 4.9percent in December 2008 to

37 percent in June 2009,due to the combination of price reduction in the secondhalf of 2008 and the impactof last years large increasedropping out of the inationcalculation. Meanwhile, the

year-on-year increase in foodprices decreased from 24.9percent in December 2008

to 15.3 percent in the sameperiod.

3.4 Imported tradeablesinflation fell significantly from 11.9 percent in Decem-

ber 2008 to 0.4 percent inJune 2009, thus reecting the impact of the decreasein the cost of fuel, as well aslower food price increases.Domestic tradeables ina -tion also decreased from

26.4 percent in December2008 to 18 percent in June2009. The higher level of domestic tradeables ination reects the impactof the 30 percent additional tax on alcoholic bev-erages that was introduced in November 2008.Overall, all tradeables ination eased signicantly from 16.3 percent to 6.5 percent over the sameperiod. Meanwhile, non-tradeables ination rosefrom 6.4 percent in December 2008 to 9.8 percentin January 2009, before decelerating to 8 percentin June 2009 as the previous years increase in

electricity tariffs and transport fares dropped outof the ination calculation.

3.5 Demand pressures on ination were gener -ally low in the context of the deteriorating world

economy that had an adverse impact on domesticeconomic activity, particularly mining output. Inthe rst quarter of 2009, diamond production

was 68.6 percent lower than output for the sameperiod last year and, as a result, the rst quarterGDP contracted by 20.3 percent. In contrast, therobust performance of the non-mining sector wasmaintained, with annualised rst quarter growthof 9.3 percent and strong performance of agricul-ture (14.1 percent), nance and business services

(16 percent), and construction and social personalservices, which both expanded by 12.9 percent.

3.6 While there was virtually no contagion fromthe global nancial crisis to the domestic banking

CHART 3: D ECOMPOSITION OF THE CPI (J ANUARY 2004 - J UNE 2009)

0

5

10

15

20

25

30

2 0 0 4

A p r

J u l

O c t

2 0 0 5

A p r

J u l

O c t

2 0 0 6

A p r

J u l

O c t

2 0 0 7

A p r

J u l

O c t

2 0 0 8

A p r

J u l

O c t

2 0 0 9

A p r

P e r c e n

t c

h a n g e

i n f o o

d p r i c e

s a n

d o v e r a

l l

C P I

-60

-30

0

30

60

90

120

P e r c e n

t c

h a n g e

i n f u e

l p r i c e s

Food inflation

Inflation excluding food & fuel

Fuel inflation

CHART 4: C ONTRIBUTION OF FOOD , F UEL AND OTHER ITEMS P RICE CHANGES IN NATIONAL INFLA -TION (S EPTEMBER 2007 - J UNE 2009)

-4

-2

0

2

4

6

8

10

12

14

16

S e p - 0

7

O c t - 0 7

N o v - 0

7

D e c - 0

7

J a n - 0

8

F e b - 0

8

M a r - 0 8

A p r - 0 8

M a y - 0

8

J u n - 0

8

J u l - 0 8

A u g - 0

8

S e p - 0

8

O c t - 0 8

N o v - 0

8

D e c - 0

8

J a n - 0

9

F e b - 0

9

M a r - 0 9

A p r - 0 9

M a y - 0

9

J u n - 0

9

P e r c e n t

Inflation excluding food & fuel Food inflationFuel inflation Inflation


11/45

4

B ANK OF B OTSWANA

system, 4 the private sector credit growth was 4.3 percent in the rst six months of 2009, and muchslower than the 9.4 percent for the correspond-ing period in 2008. Indeed, lending to private

businesses fell by 4.1 percent (compared to 15.2percent growth in 2008) while household borrow-ing increased by 10.6 percent (compared to 5.5percent in 2008). Credit expansion in the twelvemonth period to June 2009 was 22.3 percent, lowerthan the 27 percent for the year to June 2008.

Although credit expansion continues to supportoutput expansion, there is evidence of a slowdownin monetary growth and this is reecting a morecautious lending environment, reduced demandfor bank nancing associated with slow economicactivity, and an increase in loan loss provisions.The upward income effect on household borrowing during this period was limited by the absence of anacross-the-board wage increase for civil servants in2009, which would normally be replicated by other

employers.3.7 The government expenditure programme for2009/10 was subject to review in the light of lowerrevenue, and thus potentially reducing the impactof public spending on economic activity and con-tributing to the lacklustre GDP expansion. Never-theless, recurrent and development expenditures inthe scal year to March 2009

were robust and 41.2 per-cent higher than spending for the corresponding periodin 2008; this compares with

the original budget estimateof a 14.7 percent increase inexpenditure. 5 For the scal

year 2009/10, the decit of P13.4 billion announced inthe February 2009 BudgetSpeech, and representing annual expenditure growthof 5.3 percent, is likely to besignicantly reduced in thecontext of the 5 7 percentcut in budgeted expenditure

and the ongoing review andprioritisation of projects.However, the Governmentstill maintained a countercy-clical stance and has securedan external loan to nance

4 The rst round effect of the global nancial crisis onthe domestic banking system has been limited dueto a low level of foreign exposure and the absence of risky derivatives in their operations. However, therecould be a second round impact on the asset quality and protability of banks emanating from the effect of

weak domestic economic activity and income losses.5 The growth in government spending by 41.2 percent

compared to the budgeted 14.7 percent was due tosupplementary expenditure authorised by Parliamentin December 2008.

the decit, 6 while domestic bond proceeds continueto be earmarked for the development of tertiary education infrastructure.

4. M ONETARY POLICY IMPLEMENTATION IN THE FIRST HALF OF 2009

4.1 The projections for both ination and output were largely unchanged from the prognosis made atthe beginning of the year. Nevertheless, there weresigns in the rst half of the year of a quicker globaleconomic recovery than earlier anticipated, which

would also imply a shorter period of output contrac-tion in Botswana (Chart 8). There was also a risk of ination being higher and a delayed return to themedium-term objective range than earlier projected,especially taking account of price developmentsin South Africa and recovery in international oilprices. 7

4.2 Overall, with the decline in the rate of pricechanges expected to continue and ination expectedto fall towards the medium-term objective range of 3-6 percent, monetary policy was eased in the rsthalf of 2009, thus providing stimulus to economicactivity in the face of downside growth risks asso-ciated with the intensication of the impact of theglobal economic deterioration. The Bank Rate was

6 The Government has negotiated a loan of USD1.5 billion (approximately P10.5 billion) from the AfricanDevelopment Bank to help nance the decit.

7 It is noted, however, that the forecast for South Afri-can ination includes the impact of the 31.3 percentincrease in electricity tariffs by ESKOM, which wouldhave no immediate direct effect on prices in Botswana (to the extent that the Botswana Power Corporationdoes not replicate the ESKOM tariff increase).

Source: World Economic Outlook, International Monetary Fund.

-2.0

-1.5

-1.0

-0.5

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

Nov 08 Jan 09 Apr 09 Jul 09Date of forecast

P e r c e n t

2009

2010

CHART 5: G LOBAL OUTPUT FORECASTS (2008 2010)


12/45

5


2008 to 1.55 in June 2009,and for the 3-month BoBCthe real rate rose from 0.5percent to 1.6 percent in thesame period (see footnote7). Similarly, the real primelending rate rose from 2.46percent in December 2008 to5.61 perc ent in June 2009,

while the real 91-day depositrate increased from 4.55percent to 0.6 percent.

4.4 As domestic inationremained higher than theaverage inflation of trad-ing partner countries, thenominal exchange rate of thePula crawled downwards.Therefore, the nominal ef-fective exchange rate (NEER)depreciated by 1.3 percentin the six month period toJune 2009. The Pula ap-

preciated by 9.9 percent against the SDR, mostly reecting the 10.7 percent strengthening against

the US dollar. With imminent prospects for globaleconomic recovery, the US dollar (which previously

beneted from the ight to quality (safe haven) phe -nomenon), weakened as investors shifted to higher

yielding emerging market portfolios. In turn, thePula depreciated by 8.1 percent against the rand,

which beneted from the same shift in portfolio

reduced by 100 basis points each time in February and April 2009 and by 150 basis points in June2009 to 11.5 percent. In response, money marketinterest rates fell (Chart 6).The average prime lending rate of commercial banksdecreased from 16.5 percentat the end of December 2008to 13 percent in June 2009,

while the 91-day deposit ratefell from 8.53 percent in De-cember 2008 to 7.64 percentin June 2009. 8 The yield onthe 14-day BoBC decreasedfrom 12.55 percent at theend of December 2008 to8.66 percent in June 2009,

while the 3-month BoBC yield fell from 13.13 percentto 8.71 percent. 9

4.3 Given the faster de-

crease in ination, real in -terest rates rose (Chart 7).The real interest rate for the14-day BoBC increased from

1.01 percent in December

8 In March 2009, the Bank directed commercial banksto maintain an anchor 91-day deposit facility paying Bank Rate minus 4 percent. This was done partly toinuence symmetric changes for both deposit andlending interest rates in response to Bank Rate ad-

justments, with a view to ensuring a more efcientpolicy transmission process, and in order for policy adjustments to have the desired effect in inuencing

saving and borrowing decisions.9 The 8.71 percent for the 3-month BoBC was recorded

at the July 2009 auction, which is the earliest auc-tion date for this instrument following the June 2009reduction in the Bank Rate.

CHART 6: I NTEREST RATES (J ANUARY 2006 J UNE 2009)

9

10

11

12

13

14

15

16

17

18

4 - J

a n - 0

6

1 - M

a r - 0 6

1 8

- A p r - 0 6

2 8 - M

a y - 0 6

2 5

- J u l - 0 6

1 9 - S

e p - 0

6

1 5 - N

o v - 0 6

1 2 - J

a n - 0

7

9 - M

a r - 0 7

9 - M

a y - 0 7

4 - J

u l - 0 7

2 9 - A

u g - 0

7

2 6

- O c t - 0

7

2 1 - D

e c - 0 7

2 1 - F

e b

- 0 8

2 1

- A p r - 0 8

1 6 - J

u n - 0

8

1 0 - A

u g - 0

8

1 9 - S

e p - 0

8

1 0 - N

o v - 0 8

5 - J

a n - 0

9

2 6 - F

e b

- 0 9

2 7

- A p r - 0 9

2 4 - J

u n - 0

9

P e r c e n t

3-Month BoBC Rate Bank RatePrime Rate 14 Day BoBC Rate

CHART 7: R EAL INTEREST RATES - INTERNATIONAL COMPARISONS FOR 3-MONTHS MONEY MAR-KET INSTRUMENTS (J ANUARY 2004 J UNE 2009)

-6

-4

-2

0

2

4

6

8

2 0 0 4

A p r

J u l

O c t

2 0 0 5

A p r

J u l

O c t

2 0 0 6

A p r

J u l

O c t

2 0 0 7

A p r

J u l

O c t

2 0 0 8

A p r

J u l

O c t

2 0 0 9

A p r

P e r c e n t

SA (CPIX) UK

USA Botswana


13/45

6

B ANK OF B OTSWANA

holdings. The real effective exchange rate (REER) 10 of the Pula depreciated by 0.6 percent in the six months to June 2009 due to both the narrowing of ination differentials with trading partner countriesand the nominal depreciation against the rand. Ineffect, the rate of crawl more than offset the ina -tion differential, thus reversing part of last yearsappreciation of the REER.

4.5 Movements in interest rates and the Pula exchange rate are reected in changes in the realmonetary conditions index, which measures therelative tightness and easiness of nancing condi -tions in the economy. As shown in Chart 10, withthe progressive reduction in interest rates, the fallin ination and narrowing of ination differentials

vis--vis the average ination of trading partnercountries, real monetary conditions loosened, po-tentially facilitating domestic demand-led stimulusto economic activity, as well as enhanced interna-tional competitiveness of the domestic industry.

5. M EDIUM -TERM INFLATION O UTLOOK 5.1 The forecast ination path is determined by a combination of, among others, prospective de-

velopments with respect to demand-pull pressuresresulting from real economic activity, importedination and other exogenous factors, such aschanges in administeredprices. The external inu -ences on domestic pricesinclude economic and nan -cial developments in South

Africa, which is Botswanasmajor trading partner, as

well as global events, suchas changes in internationalcommodity prices and de-mand in major markets. 11

5.2 South Africas GDPgrowth is projected to de -celerate from 3.3 percent in2008 to 1.1 percent in 2009,

mainly reecting the declinein manufacturing and min-ing output, as well as slug-gish domestic demand con-ditions. Despite the forecastoutput contraction and theassociated downward pres-sures on ination, there areupside risks to the inationoutlook emanating from a 31.3 percent increase inESKOM electricity tariffs.

10 The REER is calculated using the weighted averageination for the SDR countries and South AfricanCPIX ination.

11 Forecasts for external variables are obtained mainly from the Reuters survey of forecasters.

Overall, South Africas annual headline inationis projected to remain above the upper end of the36 percent ination target for the whole of 2009.In the context of sluggish economic activity, theSouth African Reserve Bank is expected to maintainthe current easy monetary policy stance, but witha reduced likelihood of further interest rate cuts inthe short term.

5.3 World economic performance remains weak, with output forecast to decline by 1.4 percent in2009, from an estimated growth of 3.1 percentthe previous year, before recovering to 2.5 per-cent in 2010. The forecast contraction in globaleconomic activity in 2009 is attributed more to

weaker performance in advanced economies andthe concomitant contraction in trade ows (Chart8). It is expected that the stimulus packages andnancial support to the nancial system and otherindustries, as well as enhanced bilateral and mul-tilateral support for vulnerable economies, wouldcontribute to a quicker economic recovery. It is,however, projected that growth will remain sig -nicantly lower than long-term potential. In thecircumstances, global ination remains low and isforecast to average 2.5 percent for 2009. However,there is an upside risk to the ination outlook duemainly to uncertainties with respect to develop-ments in international oil prices (around USD70in July from a low of USD44 in December 2008).

5.4 The domestic economy, which contracted inthe last quarter of 2008 and the rst quarter of 2009,is projected to have recovered to positive growth inthe second quarter, largely reecting developments

Notes: 1. Developed economies include USA, UK, Euro area, Japan, Norway, Canada,Australia, Sweden, New Zealand and Switzerland 2. Data for 2008Q1 2009Q1 are estimates, while those for 2009Q2 2009Q4 are projections

Source: JP Morgan

CHART 8: E XPORT , IMPORT AND GDP Q UARTERLY GROWTH ESTIMATES AND FORECASTS FOR DEVELOPED ECONOMIES

-40

-35

-30

-25

-20

-15

-10

-5

0

5

10

2008Q1 2008Q2 2008Q3 2008Q4 2009Q1 2009Q2 2009Q3 2009Q4

P e r c e n t

Exports

Imports

GDP


14/45

7


in the world economy andthe consequent beneficialimpact on the diamond sec-tor; this trend is expected to

be sustained going forward.However, it is expected thatoutput will remain below thelong-term trend, with theresultant negative outputgap signifying low demandpressures on ination overthe medium term. The March2009 Business ExpectationsSurvey (BES) revealed wide-spread pessimism among

both the domestic and ex-port-oriented rms. Overall,

business confidence haddeclined from 82 percent inthe September 2008 survey

to 40 percent in the March2009 survey. Businesses,nevertheless, anticipatedpositive output growth in 2009/10, possibly bu-oyed by government efforts to maintain spending at levels sufcient to support growth outside themining sector. While the government will incur a large budget decit to support economic activity,expenditure growth will be at a lower level giventhe ongoing project review. It is also anticipatedthat expansion of credit to the private sector willcontinue to slow in the context of lower growth ineconomic activity, more cautious lending approach

by banks and restrained demand for credit, as wellas a moderate increase in personal incomes. In thisregard, the March BES indicated much reducedexpectations of additional borrowing, from bothdomestic and external sources.

5.5 The narrowing ination differentials betweenBotswana and the trading partner countries suggestcontinuance of a marginal downward rate of crawl

with virtually no upward pressure on ination.However, market forecasts suggest that the rand

will appreciate in the short term and the resultant

depreciation of the Pula against the South Africanrand will exert some modest upward pressure ondomestic ination. Additional upside risks to theination outlook include possible large increasesin administered prices, notably an increase in elec-tricity tariffs in the context of higher power importcosts and to recover the costs of expansion of gen-eration and supply infrastructure. Meanwhile, theMarch 2009 BES shows expectations of a declinein ination, with businesses anticipating lowerrates of increase for business costs, including raw materials and wages. Overall, the Bank forecaststhat ination will, in the short to medium term,stabilise around the objective range (Chart 9).

6. M ONETARY POLICY STANCE6.1 The recent and prospective economic de-

velopments indicate a continuation of sluggisheconomic growth, due to lower external demand,a reduced rate of government expenditure growthand a slower rate of monetary expansion. While itis anticipated that output expansion will resumefollowing negative growth in the last quarter of 2008and rst quarter of 2009, in Botswana and globally,it is forecast that the ensuing rate of increase inoutput will be lower than the long-term trend.In turn, both the external and domestic demandpressures on ination will continue to be low. Inaddition, there are base effects of the large impactof last years 30 percent alcohol levy dropping outof the ination calculation; hence the forecast fora further decline in ination. On the other hand,the projected decrease in ination provides scopefor maintenance of expansionary monetary policy to support domestic economic recovery, particu-larly in an environment in which scal stimulus is

constrained by the fall in government revenues.6.2 However, there are signicant threats to thepositive ination outlook, notably the threat of an increase in fuel prices which has potential for

wide-ranging second-round effects. In addition, theBotswana Power Corporation faces an increase incosts after the main supplier, ESKOM, was givenpermission to raise tariffs. Furthermore, in order tohelp nance expansion of the power supply infra -structure, there could be a large and inationary increase in local electricity tariffs. In addition, thereremains a possibility of an upward adjustment of

other administered prices and consumption taxes,as well as government levies, particularly given thestrain on scal resources. However, such one-off price developments would be transitory and, to the

CHART 9: B OTSWANA INFLATION FORECAST (DECEMBER 2008 - M ARCH 2011)

2

4

6

8

10

12

14

16

2008:4 2009:1 2009:2 2009:3 2009:4 2010:1 2010:2 2010:3 2010:4 2011:1

Year-on-year Inflation

P e r c e n t


15/45

8

B ANK OF B OTSWANA

extent that the second-round impact and expecta-tions are contained, would not result in a continu-ous rise in ination.

6.3 Overall, output projections and prospectsfor monetary expansion indicate benign upwardpressures on ination. Except for a sustained in -crease in fuel prices, the effect of changes in other

administered prices and taxes would be transitory.The important thing, however, is taking appropriateand timely action to contain second-round effectsand to manage expectations. As reected in Chart10, going forward, the projected movements inreal exchange rates and real interest rates, as wellas narrowing ination differentials vis--vis trad -ing partner countries suggest less tight monetary conditions. 12

7. S UMMARY AND CONCLUSIONS7.1 Ination eased in the rst six months of 2009,

thus continuing a trend that started in the lasthalf of last year, mainly inuenced by the decreasein fuel prices and in the context of low domesticdemand pressures, as well as a benign externalinationary environment. While ination remainedabove the Banks medium-term objective rangeof 3 6 percent in the period under review, themedium-term outlook of low ination enabled theBank to ease monetary policy, reducing the Bank

12 The assessment of real monetary conditions measuresthe relative easiness or tightness of monetary policy and gauges the effect that monetary policy has onthe economy through changes in the exchange rateand interest rates. The real monetary conditions aremeasured by an index (RMCI) that combines, througha weighted average, the deviations of the real exchangerate and real interest rate from their trend values.

Rate by 400 basis points between December 2008and June 2009.

7.2 It is anticipated that external inationary pressures will continue to be restrained due to weakdemand and slack labour market conditions (hence,low wage pressures) and stabilisation of commod-ity prices at lower levels than in the previous year.

Weaker domestic economic activity, a lower rate of increase in business costs and restrained growthin personal incomes will also moderate demandpressures on ination. There remain upside risksto ination due to uncertainty on international oilprice movements and a possible large increase inadministered prices. Overall, it is expected thatination will stabilise around the medium termobjective range in the short to medium term.

7.3 In the circumstances,the prevailing monetary policy stance that is charac-

terised by a monetary easing bias remains appropriate inthe short term and supportsa recovery of economic activ-ity, including maintenanceof robust performance of the non-mining sectors. TheBank will continue to moni-tor economic and nancialdevelopments with a view toresponding appropriately toensure medium-term pricestability without jeopardis -ing the anticipated economicrecovery and growth.

CHART 10: R EAL MONETARY CONDITIONS INDEX (RMCI) (DECEMBER 2008 M ARCH 2011)

-4

-3

-2

-1

0

1

2

3

4

5

6

2008:4 2009:1 2009:2 2009:3 2009:4 2010:1 2010:2 2010:3 2010:4 2011:1

P e r c e n t

RMCI REER Gap RIR Gap


16/45

9

The Importance of SoundMetadata Management forthe Production of QualityStatistics Botswanas

CaseSabata Legwaila 1

A BSTRACTThe paper outlines Botswanas experi-ence of attaining modern data produc-tion standards through metadata man-agement. While the main focus is thosestatistics that are produced by the Bankof Botswana, the importance of equiva-lent developments in other areas is ac-knowledged. The timely availability of quality statistics is important for the for-mulation of appropriate macroeconomicand nancial policies. Throughout mod -ern economies, whether in Government,

private businesses or civil society, usersrequire up-to-date and accurate statis-tics, especially given that credibility iskey to policy effectiveness. Like other modern and upcoming economies, Bot-swana has come to recognise the impor-tance of quality, timely and transpar-

ent statistics, the production of whichis guided by internationally acceptedstandards. Metadata must be easy tounderstand and not only address thecontent of the data, but also provide av-enues for users to make follow-ups with

producers. An essential component for any metadata is outlining appropriately

prioritised plans for improvement and following up on their implementation. Effective cooperation between statistics producing agencies is essential in or-

der to realize plans to graduate to high-er standards of statistics production.

INTRODUCTIONThe purpose of this paper is to outline Botswanasexperience of attaining modern data producer stan-

1 Principal Economist (Statistics and InformationServices), Research Department. An earlier versionof this paper was presented at the conference of theInternational Statistical Institute (ISI), held in Durban,South Africa in August 2009. The views expressed inthis paper are those of the author and do not neces-sarily represent those of the Bank of Botswana or theISI.

dard through metadata management. 2 In addition,the paper also explains the benets that the country has achieved regarding improved statistics, andthe way forward in utilising modern day data andmetadata management systems. The main focus of the paper is on the countrys experience with theproduction of monetary and balance of paymentsstatistics. But the efforts made in statistics in otherareas aimed at addressing the many challenges thatstill exist are also important.

The timely availability of quality statistics isimportant for the formulation of appropriate mac-roeconomic and nancial policies. Therefore, theimportance of good quality statistics cannot beoveremphasised. Throughout modern economies,

whether in Government, private businesses orcivil society, users require up-to-date and accuratestatistics, especially given that credibility is key topolicy effectiveness. Like other modern and upcom-ing economies, Botswana has come to recognise

the importance of quality, timely and transparentstatistics, the production of which is guided by internationally accepted standards.

Due to the demand for high quality statistics by users, producers emphasise producing statisticsthat are timely, internationally comparable andtransparent. For there to be transparency, there isneed to document the processes of data productionthrough the use of metadata. Metadata is simply a description of how the data are produced; be itthe production processes, description of conceptsand terminology, classication and standards, sta -tistical methods and software used, as well as theeconomic units that generated the data. Metadata are essential in all stages of the production proc-ess in order to assist the user understand the data in the context of how it has been produced. Userscan utilise sources of information more effectively if they know the denition of concepts and classica -tions together with the various stages of produc-tion. There is, therefore, a wide variety of metadata systems being developed by specialists in variouselds, some simple and others more sophisticatedin relation to their compilation and management(Lukhwareni et al , 2005).

One of Botswanas objectives is to producehigh-quality statistics. This has taken the form of the countrys active participation in internationalinitiatives of data improvement, mainly based onthe standards developed by the International Mon-etary Fund (IMF). This has been made possible by nancial assistance given by the United KingdomDepartment for International Development (DFID).Such efforts have helped strengthen capacity toproduce a broad range of ofcial statistics in Bot -swana, compiled by several agencies, through a focus on metadata issues.

The rest of the paper will cover the following:overview of models of data and metadata systems,

2 Metadata describes countries practices with respectto the production and dissemination of statistics.


17/45

10

S. L EGWAILA

which gives a basis for national data systems; background information on steps leading to im-provements in Botswanas data quality; metadata management efforts and benets; lessons and chal -lenges and, nally, a conclusion.

A N O VERVIEW OF MODELS OF MODERN D ATA AND METADATA S YSTEMSData are public goods and, for this reason, theirproduction cannot be limited to use by ofcials ornationals only. The interconnections of the worldseconomies require the availability of quality data to inform timely and well-based policy decisions.This is in order, for example, to counter macr-oeconomic instability, such as the recent globalrecession and nancial turmoil. Ultimately, the

value of statistics depends upon their accessibil-ity, quality, timeliness, dissemination, validity andcontrols (World Bank, 2002), which in turn depend

on the credibility of the methods and standardsused in the statistics production. Similarly, in a globally competitive environment, business servicesdepend on availability of national, international,and transnational data for policy formulation andevidence-based decision making at national andinternational levels. In other words, national data enter the international statistical systems, wherestandardisation is necessary to produce consist-ent, international data sets. Accordingly, models of national data are, to a large extent, inuenced by international data systems and models.

International organisations, including the IMFand World Bank, play an important role in statisti-cal development by promoting and implementing internationally agreed standards, methods andframeworks for statistical activities. These, andother international institutions, have a stake innational statistical systems; hence, they help toensure that the processes for the compilation of statistics are based on common methods. Theseenable international comparisons of statistics andenhance coherence of information content andinterpretation through the use of standardisedconcepts and denitions. By providing a common

basis for comparison, international guidelines makestatistics more useful and offer more transparentreporting of outcomes. The approach also permitsthe establishment of benchmarks or standardsagainst which national statistical systems canmeasure their performance, and strive to reach evenhigher standards. This has led statistics produc-ers to adopt models that meet modern data needs,and put in place metadata systems that not only enhance transparency, but also increase ease of usability.

IMPROVEMENTS OF Q UALITY OF STATISTICS IN BOTSWANA

Background to Statistical Developments inBotswanaTo be effective, statistical systems must be backed

by legislation that provides the safeguards of condentiality for the providers of raw data andassurances of data integrity and accessibility by users. The nature and organisation of nationalstatistical agencies vary from country to country.In decentralised systems, separate agencies haveindependent mandates to compile and disseminatestatistics in particular areas. But even in highly centralised systems, responsibilities may be dividedamong different agencies and institutions. Gener-ally, a national statistical system is made up of agencies whose activities are typically coordinated

by legislation, administrative practices and profes-

sional standards, which is essential for avoiding discrepancies in data. For example, central banksusually collect data on money and banking andother areas such as the balance of payments.

In Botswana, there is enabling legislation gov-erning the collection of data by the various agencies.However, the main agency responsible for produc-ing national statistics is the Central StatisticsOfce (CSO), which is due to be transformed froma government department into an autonomousagency. 3 The Bank of Botswana produces the bal-ance of payments and monetary statistics, whilethe Ministry of Finance and Development Plan-ning (MFDP) is the source of government nancestatistics. Inter-agency coordination in statisticsproduction is provided through working groupsand a high-level Statistics Producers Committee(SPC), comprising representatives from the threeinstitutions. The overall objective is to promote goodgovernance in the production of statistics, includ-ing transparency and accountability. In turn, theseguide policy formulation and decision making by stakeholders, including government, householdsand businesses.

Although the CSO has for many years produced

relatively reliable statistics, the efforts towards a coordinated approach for an overall framework of good statistical processes started in 2001 follow-ing the IMF mission to prepare a Report on theObservance of Standards and Codes (ROSC) onstandards of national statistics in Botswana. 4 Thereport assessed statistics covered by real sector,prices, government nance, monetary and balanceof payments, against benchmarks set out in the

3 At the time of writing, the CSO operates as a govern-ment department under the provisions of the Statistics

Act (1967). However, in October 2009 a new StatisticsBill was gazetted for discussion by Parliament, which

will establish for the new autonomous body, to becalled Statistics Botswana.

4 See Appendix 1 for a summary of the mission.


18/45

11

T HE IMPACT OF V ALUE A DDED T AX ON INFLATION IN B OTSWANA

IMFs Data Quality Assessment Framework (DQAF).The report noted that Botswanas statistics weregenerally of good quality, but needed improvementsin both compilation and, more importantly, dis-semination frequency and timeliness. Accordingly,since 2001, Botswana has participated in the re-gional General Data Dissemination System (GDDS)project for Anglophone African countries, for whichtechnical assistance (TA) funded by DFID wasprovided jointly by the IMF and the World Bank.Initially, the project was intended to last for three

years, but was subsequently extended to a secondphase up to September 2009, with the possibility of a third phase.

Under the project, the following major recom -mendations were made under the various statisti-cal areas:

(a) monetary statistics : expansion of data cover-age to include other depository corporationsthat were not covered in the then monetary

survey and produce a full depository corpora-tions survey; proper application of the conceptof residency by banks; banks to submit data electronically; and the need for the developmentof statistical database software;

(b) balance of payment s statistics : improving peri-odicity to quarterly as opposed to annual data as well as improve coverage and/or methodol-ogy for various balance of payments items, mostimportantly trade in services;

(c) government nance statistics (GFS) : adopt the2001 GFS Manual and timely dissemination of

central government budgetary accounts as wellas compilation of consolidated data of generalgovernment (both local and central govern-ment); and

(d) real sector statistics : acceleration of work pro-gramme for implementing System of National

Account (SNA) 1993 with respect to nationalaccounts, production and dissemination of producer price index (PPI) as well as regular5-yearly updates of the consumer price index

weights.

Prior to 2001, except for internal working

documents, the Bank of Botswana had no publicly available documentation on how the data were pro-duced. Moreover, while the data were disseminatedon a timely and regular basis, 5 there was no publicly available metadata documentation. Thus, apartfrom relying on past experience, data users had nomeans of knowing when they could expect to receivedata updates, what the policy was regarding revi-sions; nor did they have condence that other usersdid not have more favourable access to data.

Implementing Plans for ImprovementParticipation in the Anglophone Africa GDDSProject has led to the development of metadata,

5 Principally through two Bank publications: themonthly Botswana Financial Statistics and the AnnualReport.

outlining appropriately prioritised plans for im-provement, which were rst published in the IMFDissemination Standards Bulletin Board (DSBB) in

August 2002. 6 To initiate and assist with the imple-mentation of these plans, the national authoritiesrequested TA from the IMF under both phases of the GDDS Project. A largely ad hoc, country specicapproach was used in the rst phase; however, inthe latter phase a modular approach was adoptedunder which groups of countries were provided withstructured assistance in key areas.

With regard to monetary statistics, the rst TA mission took place in 2003. During the follow-upmissions in August 2004 and 2005, the range of monetary data was expanded alongside the adop-tion of the framework recommended by the 2000

Monetary and Financial Statistics Manual (MFSM)and data were published using standardised reportforms. Other work included the adoption of instru-ment classication and denitions of institutional

units as outlined in the framework recommended by the MFSM and the MFS Compilation Guide.

For balance of payments, the major objective was to compile and publish the quarterly balanceof payments accounts; a goal which was achievedin January 2008. However, a remaining challengeis the large errors and omissions which imply signicant miss-estimation in one or more of thesub-accounts. This is likely to be due to deciencies

with information collected as part of the quarterly and annual balance of payments survey. FurtherTA is being sought to help resolve the problem.This commenced with a further IMF TA missionin early 2009, but is also likely to involve otherproviders of TA. 7

Efforts over the last few years have focused onensuring that the monetary and nancial and the

balance of payments statistics are in conformity with the MFSM and the 1993 Balance of Payments Manual, fth edition ( BPM5), which provide practi-cal guidance for compiling statistics in those areas,as well as dissemination issues. Further improve-ments in this and other areas will continue beyondPhase II of the GDDS project.

A CHIEVING D ATA Q UALITY THROUGH METADATA M ANAGEMENT Besides the need to improve documentation of data production processes, enhancement of metadata quality has to be maintained and managed on an

6 Botswana made notable progress in this regard.Prior to the ROSC mission, metadata was a largely unknown concept to compilers of statistics in thecountry. Nevertheless, Botswana was among therst countries involved in the project to produce andpublish metadata on the DSBB.

7 Notably, both the Macroeconomic and FinancialManagement Institute of Eastern and Southern Af-rica (MEFMI) and the Commonwealth Secretariat areproviding assistance on monitoring private capitalows.


19/45

12

S. L EGWAILA

ongoing basis to meet the needs of both users andproducers. Following the prompt moves to prepareand post metadata on the DSBB (see footnote 3),this achievement was followed up with deliberatesteps to implement the plans for further improve-ments and these were recorded in the updates of metadata that are undertaken annually. Focusing on metadata issues has helped to highlight whatneeded to be done to effect improvement, and hasalso added backbone and discipline to Botswanasstatistics production efforts, including improveddata quality, dissemination and coordinationamong agencies and between data sets.

Data Quality ImprovementsBy implementing plans for improvements withdetermination, Botswana has made considerableprogress in terms of improving the quality of data and its dissemination. For instance, the Bank of Botswana was able to publish the required deposi-

tory corporations survey within three years of sub-scribing to the GDDS project. Quarterly balance of payments estimates are also published regularly,although, as noted above, concerns about someaspects of the data remain. Similarly, the CSO hasimproved quality of key indicators, such as monthly trade statistics.

Improved DisseminationSimilarly, data dissemination has improved. Mon-etary statistics are submitted on a monthly basisfor publication in the IMF International FinancialStatistics (IFS), in standardised report formats,and are also published on the Bank of Botswana

website 8 each month. Metadata is published on theDSBB and currently updated on an annual basisor when there are new developments to the data.

A link to the metadata is provided on the Banks website, which is currently under review to ensureenhanced data accessibility. These efforts increaseddata reporting frequency and quality, and has alsogiven rise to closer coordination between the Bankof Botswana, CSO and MFDP, as the key data producing agencies through fora such as the SPC,and a national GDDS committee, comprising rep-

resentatives from all the three statistics agencies.The close cooperation has resulted in accountability in jointly pursuing agreed goals in terms of data production schedules.

LESSONS AND CHALLENGES Although the Bank has made signicant progressin improving its data, it is evident that, while thecountry needs to be proactive in taking action wherestatistics improvements are concerned, the plansand goals need to be so designed as to be achiev-

able in a manageable time frame. Limitations andconstraints become clear during regular reviews.For example, the second ROSC mission in 2006

8 www.bankofbotswana.bw

showed how much progress had been and could beexpected to be made. In this regard, documenting plans for improvement in the metadata is good forguiding prioritisation for technical assistance. ForBotswana, this has promoted clearer prioritisa-tion within and among the three data producing agencies.

This also creates peer pressure for Botswana to benchmark against countries that have achievedhigher data quality standards, hence instilling discipline in ensuring perseverance in data im-provements efforts. This is facilitated by the regularupdates of metadata on the DSBB, which enablesusers to compare the standards achieved in Bot-swana with the many other countries, not just in

Africa, that participate in GDDS. 9

Development and improvement of statistics is a continuous process and, therefore, requires carefulplanning taking into account resource constraints

before graduation to the next level, i.e. subscription

to the IMFs Special Data Dissemination Standards(SDDS). SDDS is more demanding than GDDSin terms of specic standards that must be bothachieved and maintained, and thus requires insti-tutional capacity development and policy review.For example, a wider range of data categories iscovered, and countries that persistently fall shortof the required standards are identied. 10 Therefore,

while Botswana has participated in the module inthe second phase of the GDDS project to help herprepare for SDDS, the nal decision to subscribeshould not be taken lightly.

There can also be a tradeoff among the differentdimensions of data standards. For instance, theGDDS emphasises timelines, which could result inan erosion of quality if data is published before itis ready in order to meet the target deadline. Thereare also practical issues, especially for countries

with limited capacity in terms of trained and expe-rienced staff, etc. For such countries in particular,cooperation between data producers and users canmake sense, but may risk violating the data dis-semination standards relating to equal access forall users. In this context, it can be a challenge forthe Bank of Botswana, which is also a major user

of statistics, to cooperate effectively with the CSOin the production of statistics (national accounts,for example), without breaching the spirit of thisprinciple.

Capacity constraints also affect suppliers of the raw information that is necessary to compilestatistics. These are typically private businessthat, in a small country such as Botswana tendto be small-scale and limited in number. As such

9 www.dsbb.imf.org. Currently, 94 coutries participatein GDDS.

10 Additional categories that are covered include foreignexchange reserves (the reserves template), interna-tional debt, labour market and production indicators,etc, while requirement in terms of frequency andtimeliness of publication are much more stringentthan with GDDS.


20/45

13

T HE IMPACT OF V ALUE A DDED T AX ON INFLATION IN B OTSWANA

they may easily suffer from survey fatigue, whichcontributes to the problem of low response ratesthat affects surveys conducted by both the CSOand Bank of Botswana.

As noted, subscription to SDDS requires thata high standard is reached simultaneously acrossa broad range of data categories. For this to beachievable, all the agencies involved need to besupportive of the others, and sensitive to the par-ticular challenges that each may face. An examplehere, although not strictly relevant to issues of subscription to SDDS, is the cooperation betweenthe Bank of Botswana and the recently-establishedNon-Bank Financial Institutions Regulatory Au-thority (NBFIRA) which is important for collecting information about nancial institutions outside the

banking sector, but which is still in the early stagesof developing the necessary capacity. 11

CONCLUSIONBotswana has gone a long way in improving itsstatistics, and has done it right the rst time, by producing and using metadata to guide the data improvements. The improved coordination among data producing agencies is notable, and the par-ticipation of Botswana in the SDDS module of thesecond phase of the GDDS project, conrms theextent of the progress that has been made. However,subscription to SDDS remains a challenge which

will require further time to deal with and should not be rushed. As well as increasing the range of data coverage, further improvements need to be madein terms of accuracy and timeliness, as well as toincrease efforts to manage the National Summary Data Page (NSDP) and produce an Advance ReleaseCalendar (ARC), both of which are key requirementof SDDS. 12

Overall, though there are challenges withrespect to graduating to SDDS and capacity con-straints, effort are continuing with the assistanceof international agencies such as the IMF and the

World Bank, in helping with the development of thecountrys statistics in general through provisionof technical assistance and training. In general,

there have been marked improvements in thecountrys statistics. In the period ahead, Botswana

will continue to learn from other countries thathave succeeded in adopting and utilising moderndata models and developing world-class metadata systems to ensure coherence and ease in data production and use.

11 This is not directly relevant to SDDS, as this does notcover data from banks. However, this could changein the future given the extent to which the recentnancial crisis has not been conned to banks.

12 The NSDP is currently published on the Bank of Botswana website, but both access and timeliness of updates could be improved further.

REFERENCES Building Statistics Capacity to Monitor Development:

Information Paper for the World Bank Board of Ex-ecutives. October 2002

Matjaz J. et al, Using Metadata in Statistical Processing Cycle The Production Tools Perspective

Lukhwareni, T.J, et al, Management of Metadata inNational Statistics Agency, Statistics South Africa

2005.Bank of Botswana Research Bulletin. December 2003


21/45

14

S. L EGWAILA

T ABLE 1: B OTSWANA ROSC R ECOMMENDATIONS

General NationalAccounts

Prices GovernmentFinance

MonetaryStatistics

Balance of Payments

Total

Short term 6 (2) 1 (1) 2 (1) 2 (1) 4 (2) 2 17

Med term 3 2 2 4 1 1 (1) 13

Long term 1 1

Total 9 3 5 6 5 3 31

Note: figures in brackets are the number of recommendations which were given high prioritySource: Botswana ROSC Report

A PPENDIX 1: S UMMARY ON REPORT ON THE O BSERVANCE OF STANDARDS AND CODE M ISSION

The ROSC Report on BotswanaThe IMF ROSC mission commenced in early October 2001 and lasted for two weeks.It focused on the following data categories: national accounts; prices (consumer

and producer); government nance; monetary statistics and balance of payments.Neither socio-demographic nor labour market statistics were included at this stage,although the subsequent GDDS project covered some of these areas through theprovision of TA.

The nal report of the mission was presented to the Government in early De -cember at the same time as the workshop to launch the regional GDDS project,

which was held in Gaborone. The report included an assessment of the quality of Botswana statistics together with key short- and longer-term recommendations forimprovement. After discussions with the Government and the Bank of Botswana,the report, together with The Response of the Authorities, was posted on the IMF

website. Overall, the quality of Botswana statistics was generally seen as good, andin some cases very good.

For example, the CPI, on which ination measures are based, was assessed as being of SDDS quality, and the report emphasised that concerted effort could quickly move Botswana within striking distance of SDDS requirements more generally. Themajor deciency in terms of quality was in the area of producer prices, which theCSO candidly admitted was a concern and that assistance would be required to ef-fect necessary improvements.

Another major area where improvements were found to be needed was that of dissemination. This was in terms of both timeliness and availability to all users.Some indication of this can be seen from Table 1, which divides the main ROSCrecommendations according to sector. Out of the 31 recommendations, nine (about30 percent) were in the general category and, of these, eight dealt with dissemina-tion issues ranging from the need to establish advance release calendars to providing more extensive information on methodology, including data limitations. (The ninth

was a recommendation regarding the need for training.) Six of these recommenda-tions were identied as achievable in the short term.

Some of the sector-specic recommendations also dealt with dissemination is -sues, notably timeliness. The relatively high number of recommendations for gov-ernment nance was due, in part, to the difculties associated with incorporating local government nances fully into the framework of regular reporting. A furtherimportant conclusion was that, in some areas, production of statistics could befacilitated by improved coordination between agencies. The report noted discrepan-cies in methodology, delays in communicating necessary information (for instance,the balance of payments relies on various inputs from the CSO such as trade data

while in turn, the national accounts require timely balance of payments data), andthe potential confusion caused by the various reporting years used across the dif-ferent sectors.

Source: Bank of Botswana Research Bulletin (2003),


22/45

15

Overview of Initiatives toPromote Convergence inthe Context of RegionalIntegration: An African

Perspective1

K S Masalila 2

A BSTRACTThis paper provides an overview of initiatives to promote economic conver-

gence that are geared towards foster-ing economic integration in Africa. It is

found that there is a specic agenda for regional integration in Africa that is supported by the development of in-stitutions, a transition programme andmonitoring mechanisms for performanceand implementation. Broadly, macroeco-nomic indicators are converging and are,therefore, less variant across Africa; al-though their levels generally do not meet the set criteria. In addition, there are fre-quent slippages in harmonisation effortsand development of institutions, whichretard the integration programme. Nev-ertheless, integration and convergenceinitiatives incorporate improvements

in performance and sustainability of good policies that in the long-term could promote higher levels of investment and durable growth in the continent.

INTRODUCTIONThe imperative for regional integration stems froma desire to minimise the cost of trade between na-tions and facilitate market access that promotesgrowth of the regions industries, as well as tostrengthen the economic power of the combined

member states vis--vis third parties. For Africa,integration is also a developmental necessity inrelation to trade, economic performance, as wellas strengthening of policy credibility and effec-tiveness conducted within a secure and peaceful

1 This paper was prepared for presentation at theSouth African Reserve Bank/Irving Fisher CommitteeSeminar on Economic and Financial Convergenceen route to Regional Economic Integration: Experi-ence, Prospects and Statistical Issues Amidst GlobalFinancial Turmoil. The Seminar was held in Durban,South Africa in August 2009. The views expressed inthis paper are those of the author and do not neces-

sarily represent those of the Bank of Botswana, South African Reserve Bank or the Irving Fisher Committee.

2 Deputy Director, Research Department, of the Bankof Botswana. The author is appreciative of usefulcomments from colleagues in the Bank of Botswana.

environment. The organisational and institutionalinitiatives towards regional integration are expectedto ultimately increase intra-regional trade, stimu-late regional infrastructure development, improveadministrative efciency of institutions, facilitatehigher levels of investment and industrialisation,enhance efciency of resource allocation, affordcross-country labour/skills mobility and reducethe scope for adverse political inuence on macro-economic policies.

Specic to macroeconomic policy convergence,it is argued that it provides efciency and growththrough the elimination or reduction of exchangerate uncertainty and transaction costs; ensuresmonetary stability (price stability and lower reallong-term interest rates); and helps member coun -tries to cultivate the discipline needed to avoidexcessive decits and unsustainable debts. Thispaper outlines specic initiatives taken towardsregional integration in Africa and efforts being

made to attain macroeconomic convergence. Thefollowing section outlines the institutional stepstowards regional integration for the continent. Thenext two sections discuss some elements of con-

vergence and the expected benets, and assess theregions performance with respect to convergencecriteria and other expectations arising from regionalintegration initiatives. The nal section covers theconcluding observations.

REGIONAL INTEGRATION IN THE A FRICAN

CONTEXTRegional integration initiatives in Africa are under-taken under the auspices of the African Unions(AU) programme of transition to an African Eco-nomic Community (AEC) enunciated in the June1991 Abuja Treaty and the Constitutive Act of the

AU adopted in 2000. The African Union has des-ignated regional economic communities (RECs) asthe building blocks towards achieving an AfricanEconomic Community. In line with the AU integra-tion and transition framework, there are eight rec-ognised RECs, namely, Arab Maghreb Union (UMA);Community of Sahel-Saharan States (CEN-SAD);Common Market for Eastern and Southern Africa (COMESA); East African Community (EAC); Eco -nomic Community of West African States (ECOW-

AS); Economic Community of Central African States(ECCAS); Inter-Governmental Authority on Devel -opment (IGAD); and Southern African DevelopmentCommunity (SADC). In addition, there are six otherinter-governmental organisations working on theintegration agenda: the Central African Monetary and Economic Community (CEMAC); EconomicCommunity of the Great Lakes States (CEPGL);Indian Ocean Commission (IOC); Mano River Union

(MRU); Southern African Customs Union (SACU);and West African Economic and Monetary Union(WAEMU). Figure 1 shows the composition of the

various RECs and other regional groupings.


23/45

16

K S M ASALILA

As apparent from Appendix Table A1, a featureof these regional inter-governmental organisations,

which potentially slows progress towards integra-tion, is their signicant overlapping of membershipmotivated by strategic and political considerations,as well as economic benets and geographical con -tiguity. The disadvantages of multiple membershipsinclude difculties and tardiness in relation to:meeting nancial obligations to the RECs; focusing on numerous agendas of each REC; low rates of rat -ication and implementation of agreed treaties andprogrammes; incompatibility of some programmes;duplication of effort; and little support for, and/orunderstanding of RECs in member countries. Theother concern with overlapping and uncoordinatedmembership is with respect to countries belonging to broader regional groupings pursuing economicintegration, which also include those that do not

belong to the monetary unions, a situation whichincreases operational problems, as well as compli-

cate the planning for possible future enlargement(Ghosh, Guide and Wolf, 2008). These concerns andimpediments to the integration process led the AUSummit held in Banjul in 2006 to put a moratoriumon the recognition of new RECs.

The AU programme envisages the gradual inte-gration of African economies through a transitionprocess from establishing free trade areas, customsunions, common markets, monetary unions, cul-minating in one continental central bank (AfricanCentral Bank) and a single currency; initially forthe RECs and then for the continent. In a freetrade area, the group of countries eliminates tariffsand non-tariff barriers on substantially all trade

between them, with each country maintaining itsown schedule of tariffs on non-members. For a customs union, the group of countries constitutesa single customs territory in which duties andother restrictive trade regulations are eliminated forsubstantially all trade between the parties and, inaddition, there is a common external tariff appliedto trade with non-members. In a common market,in addition to having a customs union, restrictionson the movement of capital and labour are removed,allowing for free movement of goods, services and

factors of production. A monetary union augmentsa common market by establishing a single monetary authority which conducts monetary policy for theunion, resulting in introduction of a single currency (SADC FTA Brochure, 2009).

Integration is undertaken based on desirablefeatures of an optimal currency area (OCA) 3 whichinclude the nature and extent of:(a) price and wage exibility (b) nancial market integration(c) factor market integration (capital and labour

mobility)

3 Optimal currency area or region describes a situa-tion whereby a geographical region would maximiseeconomic efciency by having the entire region sharea single currency (Mundell, 1963).

(d) goods market integration(e) political integration

However, the major drawback of the envisagedeconomic integration and the adoption of commonpolicies is loss of monetary policy independencefor the individual countries. In moving ahead withintegration, it is considered that the benets interms of trade, optimal policy formulation, en-hanced welfare and security, as well as the likely increase in living standards outweigh the loss of policy independence for the individual countries.Moreover, it is argued that over time the desire to

work towards economic integration brings its own benets (Ghosh, Guide and Wolf, 2008, p176). Forexample, while, initially, weak transaction links anda lack of complementarity in output provide struc-tural reasons for low integration, aggravated by uneven progress in implementing trade liberalisa-tion agreements, targeting a customs union, which

would include the elimination of intra-union tariffs

and the harmonisation of indirect taxes, leads toprogress in economic integration. On process andadministrative issues, it has also been noted thata regional approach in key structural areas, suchas tariff reduction and harmonisation, legal andregulatory reform, payment systems rationalisation,nancial sector reorganisation, investment incen -tive and tax system harmonisation, and labourmarket reform enable participating countries topool their resources and avail themselves of regionalinstitutional and human resources. As a result, thegroup of countries attains a level of technical and

administrative competence that would not be pos-sible on an individual country basis. The regionalapproach also allows countries to assert their in-terests from a stronger and more condent positionin the international arena (Outtarra, 1999).

In addition, the focus on common and optimalmacroeconomic policies ultimately leads to betteroutcomes with respect to important economic indi-cators. There is, as well, the desire to foster interre-gional trade and capital ows, to insulate monetary policy from national politics, and to minimise theadverse impact of global capital ows by pooling reserves. It is noted, for example, that,

The European experience suggests two possible benets, policy credibility and trade and nan -cial integration. Indeed, while the original moti-

vation for European monetary integration in the1970 Werner Report was fostering greater inte-gration of goods and capital markets in Europe,in the event the impetus for maintaining xed ex -change rates (and eventually adopting a singlecurrency) was to import the Bundesbanks policy credibility to aid disination efforts in the early 1980s (Ghosh, Guide and Wolf, 2008, p186).

Work is, therefore ongoing on the regional inte-gration agenda to develop policies and programmesthat would hasten the formation of the AEC; pro -mote intra-African trade; harmonise and coordi -nate policies and programmes in RECs; as well as


24/45

17

A N O VERVIEW OF INITIATIVES TO P ROMOTE C ONVERGENCE IN THE C ONTEXT OF R EGIONAL INTEGRATION : A N A FRICAN P ERSPECTIVE

coordinate the development of infrastructure andinstitutions (Mouyelo-Katoula and Nshimyumure-myi, 2007). In summary, the envisaged phases fortransition to AEC are as follows: 4

(a) creation of regional blocs to be completed by 1999;

(b) strengthening of intra-REC integration andinter-REC harmonisation to be completed in2007;

(c) establishment of a free trade area and customsunion in each regional bloc to be completed by 2017;

(d) establishment of a continent-wide customsunion and, therefore, a free trade area to becompleted by 2019;

(e) establishment of a continent-wide African Com-mon Market to be completed by 2023;

(f) establishment of a continent-wide economicand monetary union and, thus, also a currency union alongside the pan-African Parliament to

be completed by 2028; and(g) end of all transition arrangements by 2034 at

the latest.

Convergence in the Context of Regional Inte-gration in AfricaConvergence in the context of regional integration isessential to establish commonality in an economicregion, in particular relating to symmetric economicperformance, institutional development, regulationharmonisation, access to infrastructure, as wellas policy-making coordination and administrative

efciency. Since economic integration entails a common approach to policy formulation or a centralpolicy-making authority, it is important that thisis not constrained by considerations of asymmetry or disparate performance with respect to economicindicators and national institutions. Moreover, inorder to have a common effect across the region,centralised policy has to reect symmetrical de -

velopments and need to be transmitted throughsimilar institutions and administrative processes.Symmetry for the region can be assessed in termsof whether: shocks to output are in the same direc-

tion; the shocks are of comparable magnitude; and whether aggregate GDP for the region is dominated by movements of one or two members.

In addition to issues of symmetry in economicperformance and policy formulation, it is alsoimportant to have similar standards and a coor-dinated approach to the regulation of systemically important institutions such as the nancial sectorinstitutions and economic activities; for example,institutions responsible for control of monopoliesand competition. Beyond regulation, other consid-erations relate to harmonisation of legislation, in-stitutions, statistics and administrative processes;especially with respect to tax and trade incentive

4 This programme may, however, become redundantas the AU pushes for the fast tracking of the Africannancial institutions.

structures, tariffs, business and labour laws, as well as payment systems.

The focus on convergence can also improvegeneral governance, adherence to prudent policies,a monitoring framework and, in turn, performance

with respect to important economic indicators.Notably, convergence suggests the adoption of the highest standard or optimal policy over whichthere are monitoring mechanisms and sanctions fornon-performance. There is, therefore, self-imposedpolicy and regulatory discipline that motivatescompliance, a factor which engenders policy cred-ibility that, in turn, help to promote investmentand market opportunities. In essence, regionalsurveillance and the dialogue between the variouspartners reduce the risks of slippages in meeting pre-agreed performance benchmarks, resulting in a more stable, predictable environment, which is anessential condition for the private sector to our -ish. Furthermore, the conditions and obligations

associated with participation in reform programmes within a regional organisation also facilitate the work of the domestic authorities in implement-ing politically difcult but important measuresthat promote scal prudence and macroeconomicstability, as well as foster a conducive businessenvironment.

Specic initiatives taken towards the achieve -ment of economic integration in Africa includepolicies, infrastructure development, trade facilita-tion and administrative processes, as well as legalreforms.

(a) Macroeconomic Convergence: Policies andBenchmarksConvergence criteria normally centre on economicindicators that are related to macroeconomic policy formulation and performance. These include therate of economic growth, level of ination, interestrates, exchange rate performance, and ratios of the

budget decit, as well as government debt to GDP;alongside ratios of savings and investment to GDP.The African Monetary Cooperation Programme(AMCP) provides a blueprint for macroeconomicconvergence for the continent and was formulated

with the objective of ensuring the adoption of col -lective policy measures that foster a harmonisedmonetary system and commonality in the man-agement of institutions. The programme has a long-term perspective involving (a) the adjustmentof member countries exchange rate to equilibriumlevels, (b) eventual exchange control-free currentand capital account transactions, and (c) the pur-suit of market-oriented monetary policy. The ulti-mate aim is to evolve (through the regional central

banks) towards a single monetary zone by 2021, with a common currency and continental central

bank (Table 1).Broadly, macroeconomic convergence is to bepromoted through implementation of appropriatemonetary policy and scal policy. Among the initia -


25/45

18

K S M ASALILA

tives in this respect is the transition towards an Af-rican Central Bank. However, this goal is preceded

by creation of regional central banks within regionaleconomic communities. Within this framework,

there are efforts to ensure effective policy-making and programme implementation to meet the set

TABLE 1: S TAGES FOR IMPLEMENTATION OF THE AFRICAN MONETARY COOPERATION P ROGRAMME (AMCP) *

Stage I:2002 - 2003 Adoption by sub-regions of monetary integration programmes

Stage II:2004 - 2008 1. Harmonisation and coordination of macroeconomic and monetary policies

2. Gradual interconnection of payments and clearing systems

(a) Promotion of African banking networks(b) Promotion of sub-regional and regional stock exchanges

(c) Strengthening and harmonisation of banking and nancial supervision

3. Observance of the following macroeconomic indicators by 2008

(a) Budget decit/GDP ratio not exceeding 5 percent

(b) Central bank credit to government not exceeding 10 percent of previous yearstax revenue

(c) Single digit ination rate

(d) External reserves/import cover of at least 3 months

Stage III:2009 - 2012 Observance of the following macroeconomic indicators by 2012

(a) Budget decit/GDP ratio not exceeding 3 percent by 2012

(b) Elimination of central bank credit to the government

(c) Ination rate of less than 5 percent

(d) External reserves/import cover equal or greater than 6 months

Stage IV:2013 - 2015 Assessment of macroeconomic performance and negotiation for the establishment of a

common central bank (2015); consolidation of third stage achievements

(a) Ination rate of less than 3 percent

(b) Continuous assessment of macroeconomic indicators against convergence criteria;comparative analysis referred to a Convergence Council

(c) Commissioning of a study on the establishment of an African Exchange RateMechanism

Stage V:2016 - 2020 Finalisation of arrangements for launching the African Monetary Union

(a) Preparation of institutional, administrative and legal framework for setting up thecommon central bank and currency of the African Monetary Union

(b) Review of commissioned study on the African Exchange Rate Mechanism; opera -tionalisation of the mechanism

(c) Appointment of key ofcers of the common central bank

(d) Preparation for the introduction of a common currency

(e) Recruitment of staff of the Bank(f) Mid-term assessment of country performance

(g) Final assessment of countries performance against convergence criteria

Stage VI: 20211. Introduction and circulation of the common African currency

2. A transitional period during which sub-regional monetary institutions would operatealongside the African Central Bank

Source: Association of African Central Banks * Note: The AMCP was adopted by the Association of African Central Banks (AACB) in Algiers in September 2002


26/45

19

A N O VERVIEW OF INITIATIVES TO P ROMOTE C ONVERGENCE IN THE C ONTEXT OF R EGIONAL INTEGRATION : A N A FRICAN P ERSPECTIVE

goals by these regional institutions. 5 In this

Documents

Botswana Research Bulletin