Attachment J.7 - Overcoming the limits of institutional reform in Uganda.pdf

Dev Policy Rev. 2018;36:O159–O182. | O159wileyonlinelibrary.com/journal/dpr

Received: November 2014 | Accepted: March 2016

DOI: 10.1111/dpr.12353

A R T I C L E

Overcoming the limits of institutional reform in Uganda

Matt Andrews Associate professor

© The Authors 2017. Development Policy Review © 2017 Overseas Development Institute

Kennedy School, Harvard University, Boston, MA

Correspondence Matt Andrews Email: matt_andrews@hks.harvard.edu

Abstract Uganda has been a public sector reform leader in Africa, producing many best practice laws and structures. Many reforms are limited to such gains, however, producing new forms that function poorly (laws are often not imple-mented and processes poorly executed). This article sug-gests adjustments to the reform agenda that address these limitations and close gaps between what Uganda’s system looks like and how it functions. Adjustments involve adopting problem driven iterative adaptation (PDIA), which requires Ugandans to work together actively on their reforms, coming to terms with limits and engaging iteratively to find and fit ways to overcome these limits.

K E Y W O R D S development, government, PDIA, Uganda

1 | INTRODUCTION

Uganda was an early adopter of many reforms that are now ubiquitous on the continent. These in-clude independent revenue agencies, medium- term expenditure frameworks, public expenditure track-ing surveys and results- oriented management mechanisms. A number of these reforms are respected across the continent because they yielded best- in- class new laws, organizational arrangements and operational mechanisms. Many of the reforms have been limited to these kinds of gains, however, producing new institutional forms that function poorly and yield limited impact. New anti- corruption laws are largely unimplemented, for instance, so corruption continues; impressive looking budgets are poorly executed, such that public expenditure results are still constrained; the revenue agency collects less revenue than possible, so government lacks resources to independently provide the country’s services and infrastructure needs; and more. This article explores the limits of institutional reform

O160 | ANDREWS

in Uganda and suggests an approach that could help the government pursue less limited reforms in future.

The article has three main sections. The first provides a brief historical narrative of Uganda’s public sector reforms. This argues that government has pursued a full reform agenda, with support from international organizations, and created many best practice institutional mechanisms through projects considered successful by government and donors. A second section discusses results of these interventions. It references recent evaluations and assessments including the Public Expenditure and Financial Accountability (PEFA) data, Open Budget Index (OBI). These sources show that Uganda’s reforms have been successful in creating new institutional forms but have been quite unsuccessful in improving actual government functionality. This leaves limits and gaps that currently undermine government’s impact. A third section suggests that government should reframe its reform agenda, addressing real problems of dysfunction instead of adopting hopeful best practice solutions. A fourth section argues that the gaps between form and function identified in this article offer entry points for this new kind of approach.

The article contributes to the institutional reform discussion in Uganda. On the face of it, the analy-sis might appear overly critical of past reforms. Those who have toiled in these reforms may even take umbrage at the narrative of reforms that have been limited to gains in form with little improvement in functionality. This perspective will be unfortunate, as the narrative intends to offer a commentary of what has and has not worked. The fact that reforms have yielded gains in terms of new forms indicates that Uganda’s reformers have the capability to make change happen. The fact that reforms have been limited shows that Uganda’s reformers need to think about how it makes change happen in a more tangible way, as the next step of the country’s reform journey.

The article also contributes to literature on institutional reform in developing countries beyond Uganda. This literature often discusses specific kinds of interventions—like decentralization or dem-ocratic governance or multi- year budgeting. These interventions are seldom seen as part of broad reform agendas and they are often poorly evaluated, leading to many apparently ‘successful’ reforms that do not yield more functional governments. This article presents an approach to investigating and evaluating reforms that takes a broad perspective—through the historical narrative—and ensures a structured empirical view—using multiple indicators and an ‘institutional limits and gap analysis’ to assess reform progress in a systematic way that should help focus discussion and strategies in future.

The approach builds on Andrews (2013) and Andrews, Pritchett, and Woolcock (2012).

2 | A BRIEF RETROSPECTIVE OF UGANDA’S PUBLIC

SECTOR REFORMS1

In 1986, President Museveni’s National Resistance Movement (NRM) ascended to power and in-herited a largely dysfunctional public service. Maladies were too numerous to mention: government bureaucracy and red tape had become major bottlenecks in service delivery; the public service was bloated; motivation in the public service—especially of teachers and health workers—was low; there was no strategic direction in government; the State was in fiscal turmoil, with uncertain spending pat-terns and large debt burdens; there was limited revenue to cover debt repayments or to fund necessary public sector engagements; corruption was considered a major problem.

1This section draws on descriptions of Uganda’s reform experience by (among others) Karakire (2012), Olum (2003), Williamson (2003), Robinson (2006) and Kyarimpa (2009).

| O161ANDREWS

Faced with these challenges, the NRM government promised Uganda’s citizens a fundamental change in the management of public affairs, adopting a 10- point reform programme. This was fol-lowed by an Economic Recovery Package (which sought to liberalize and stabilize the economy) in 1987, the establishment of the Public Service Review and Reorganization Commission (PSRRC) in 1988, the adoption of this Commission’s report as a blueprint for reform in 1989 and the launch of a Civil Service Reform Programme (CSRP) in the early 1990s. These steps were among the first in a constant and busy process of reform over the past 30 years.

Reforms in the 1990s involved creating the Uganda Revenue Authority (URA), introducing a new Constitution, initiating pay reforms, rationalizing the civil service and adopting budget re-forms. Many of the reforms attempted during this time had a ‘cutting edge’ dimension to them:

Uganda was one of the first countries to adopt a semi- autonomous revenue agency in Africa, for instance, and medium- term budgeting reforms predated efforts to do so in other African countries (including South Africa, where Medium Term Expenditure Frameworks (MTEFs) were only intro-duced in the late 1990s).

The mid- to- late 1990s saw a variety of initiatives aimed at streamlining the civil service, for-malizing the public financial management system and focusing the public administration. This pe-riod laid the foundation for efforts to decentralize and improve basic service provision (especially universal primary education). Service delivery reforms were guided by the Poverty Eradication Action Plan (PEAP), which was the first of various plans Uganda adopted in the past 20 years. The PEAP saw a push to deliver various services by local governments, which was a common reform in Africa at the time (Andrews & Shaw, 2003). The major decentralization mandate emerged from the 1995 Constitution, the Decentralization Statute No. 15 of 1995 and the Local Government Act of 1997.

Government revived civil service reforms again in the late 1990s, with the Public Service Reform Programme, a new commitment to pay reform and efforts to introduce performance management mechanisms. Once again, these efforts were at the vanguard of reforms in Africa, where few countries seriously considered introducing a results focus in government (Andrews, 2003b). Uganda’s expendi-ture tracking efforts were also novel and have since become ‘best practice’ across Africa.

The 2000s saw a flurry of new laws passed as part of continuing public sector reforms. These included laws aimed at further improving the public financial management system. The Budget Act, Public Financial Accountability Act and Public Procurement and Disposal of Public Assets Act stood out as impressive legislative interventions. A host of new laws and regulations were also introduced to combat corruption. These included the Inspectorate of Government Act, the Leadership Code Act and the Anti Corruption Act. Various organizational adjustments were made in concert with these new laws, and new bodies were created to provide financial, procurement and other oversight. These bodies create and monitor rules of the game, conduct audits and oversee the implementation of laws.

Public financial management (PFM) reforms continued apace in the 2000s as well, financed through projects like the Financial Management and Accountability Programme (FINMAP). New laws and regulations informed PFM in local and national governments and oversight bodies were created (like the Budget Monitoring and Accountability Unit (BMAU)). Efforts were taken to in-troduce IT- based systems through which PFM practices would be made more formal. Pilots were launched to adopt an Integrated Financial Management Information System (IFMIS) as early as 2003/2004 and an Integrated Personnel and Payroll System (IPPS) was introduced in 2008. The IPPS went ‘live’ in 2013 as the basis for paying all salaries. Its implementation led to calls for new budget laws, which were being considered at the time of writing, with new anti- corruption and civil service laws.

O162 | ANDREWS

3 | SUCCESSFUL CHANGE IN FORMS, BUT LIMITED

IMPROVEMENT IN FUNCTIONALITY

The discussion gives a partial perspective on Uganda’s public sector reform experience over the past quarter century. It intends to show how much activity this experience has involved. This is also evi-dent when examining the history of project engagements with the World Bank (just one development partner). Since 1990, Uganda has partnered with the World Bank on at least 20 projects focused on administrative and civil service reform, 14 projects addressing PFM reform, 15 aimed at fostering decentralization and seven targeting improvements in tax policy and administration.2 Over two thirds of the World Bank projects that have been evaluated were considered satisfactory, suggesting they met expectations. But where have these and other projects and reform interventions left Uganda today? This section argues that the results are mixed: Uganda’s government looks better after the re-forms, but many reform limits undermine government’s post- reform functionality and impact.

3.1 | Uganda’s government looks better after reforms3

Table 1 summarizes identifiable gains in Uganda’s reforms. The first column on the left provides perspective on gains and achievements in the civil service and public administration domain, for in-stance. In the 1990s, the civil service was reduced from 320,000 on the payroll to about 160,000 in 1998 (numbers are approximate because it is difficult to ascertain exact figures).4 This was a direct and positive gain given that Uganda faced a bloated bureaucracy in 1988, where public sector employ-ment accounted for about two thirds of all formal jobs—a fiscally unsustainable position. In contrast, by 1998 the public sector accounted for only about a third of all formal sector jobs—a more balanced reality. Furthermore, payroll and establishment controls were put in place and benefits monetized by the late 1990s, which led to improved public sector wages.

Additional civil service and public administration reform gains in the 2000s centred on the govern-ment’s introduction of ‘best practice’ performance management mechanisms. These emerged in various guises since the late 1990s and involve output and outcome based budgeting, performance targeting in national planning mechanisms (like the 2010 National Development Plan), and civil service perfor-mance contracts. Altogether, the mechanisms allow transparency for government aims and ties perfor-mance targets to conditions of employment for most senior civil servants. The coverage of these reforms is remarkable, given that at the end of 2010 all senior government officials, 92% of head teachers and all hospital directors were on performance contracts that allowed this kind of performance orientation.

PFM reforms have also yielded major gains and achievements. In particular, they have resulted in the adoption of new laws, management systems and administrative processes. These were established from the early 1990s and include the MTEF, Public Expenditure Tracking Surveys (PETS), Outcome/ Output oriented budgeting, and the new IFMIS. The impressive nature of these mechanisms is re-flected in the country’s strong performance on globally recognized measures of PFM system quality.

Uganda’s laws and formal processes satisfy 70% of the requirements in the PEFA measure used by most development partners.5 This suggests that the legal and formal process quality of Uganda’s sys-tem is 70% compliant with what experts consider ‘good international practice.’ The country registers

2This is based on the author’s calculations, given data in the World Bank project database.

3Most references are provided in association with Table 1 to minimize the need for footnotes.

4The 1995 Constitution capped line ministries at 21 and Cabinet ministers at 40.

5This is based on the analysis of the PEFA data using methods developed in Andrews (2011) where one isolates dimensions relating to de jure performance. Uganda scores 2.8 out of 4 on average on such dimensions.

| O163ANDREWS

T A

B L

E

Ex am pl es o f t he

K ey

G ai ns a nd A ch ie ve m en ts o f U ga nd a’ s R ef or m s

R ef or m a re a C iv il se rv ic e re fo rm a nd p ub lic ad m in ist ra tio n

Pu bl ic fi na nc ia l m an ag em en t re fo rm

R ev en ue a dm in ist ra tio n re fo rm

Pr oc ur em en t re fo rm

A nt i-co rr up tio n re fo rm

G ai ns a nd ac hi ev em en ts s r ef or m s ‘w id el y co ns id er ed su cc es sf ul nu m be r o f m in ist rie s a nd c iv il se rv ic e ha lv ed

(t o m in ist rie s.

in ist er s, an d

0, em pl ov ee

s) . P ub lic in ist ra tio n hi rin g dr op pe d fro m

7% o f t ot al fo rm al e m pl oy m en t t o ab ou t 3 5%

M aj or p ay cr ea se s f or re ai ni ng c iv il se rv an ts.

N at io na l p la ns st ra te gi es h av e be en pl e si nc e

. T he

N at io na l

D ev el op m en t P la n ha s pe rf or m an ce ta rg et s tie d to a p er fo rm an ce ag em en t a nt iv e sy st em go ve rn m en t.

B y en d

. e ve n o f he ad te ac he rs a nd a ll ho sp ita l d ire ct or s w er e on p er fo rm co nt ra ct s;

th is is

‘b es t p ra ct ic e’ in E as t

A fr ic a.

La w s, fo pr oc es se s a od er ni ze d, d co m pl y w ith

PE

FA

‘g oo d in te rn at io na l pr ac tic es

’ i n

Pu bl ic

Fi na nc ia l M an ag em en t, Th e

Pu

Fi na nc ia l M an ag em t’s sy

’s tra ns pa re nc y ra te d at

5% o n th e O pe n

B ud ge t I nd ex

2nd b es t i n

A fr ic a, U R

A o ne o f A fr ic a’ s f irs t au to no m ou s r ev en ue a ge nc ie s:

U R

A sc or es w el l o n m ea su re s of d el eg at ed a ut on om y

(s co rin g of p oi nt s o n th e

IT

D m et ric . b et te r t ha n al l ot he r A fr ic an c ou nt rie s)

To ta l T ax R at es a re re la tiv el y co m pe tit iv e at

7%

(c om pa re d w ith

3% g lo ba lly tim e to pl y w ith ta x re gu la tio la tiv el y co m pe tit iv e

(2 ho ur s v s.

av er ag e gl ob al ly

; p ro du ce d str on g ta x co lle ct io n re su lts to s, w he n re ve nu e to

G D

P ra te in cr ea se d by

(f ro

6, 8% in la te

0s to

1.

1%

9)

Pr oc ur em en t L aw s sc or e o f 4 po in ts o n

O

EC

D as se ss m en t

(c om pl yi ng

8%

‘g oo d pr ac tic e’

Th e

PP

D

A fu nc tio na l co nd uc ts nu er ou s a ud its an al ys al ly

; o ve r o f p ro cu re m en t s pe nd in g su bj ec t t o op bi dd in g.

A nt i-co rr up tio n la w s co ns id th be w or ld sc or in ou t o f 1 on G lo ba l In te gr ity sc e

D at a u se d to in fo rm ch ils er vi ce re fo rm a nd pu bt ic a dm in is tr at io n ca lc ul at io ns an d pe rs pe ct iv es ar e d ra w n fr om th e U ga nd a L ab or

F or ce S ur ve y (2

2)

. R ob in so n (2

4)

. L ab ou r M ar ke t I nf or m at io n St at us ep or t (

Th er ki ld se n an d

Ti de m an d (2

7)

. U ga nd a

B ur ea u of S ta tis tic s S ta tis tic al

A bs tra ct

(2

0) . A fr ic an D ev el op m en t B an k

(2

2) , U ga nd a C om pu te r S er vi ce s, M in is try o f F in an ce . P la nn in g an d

D ev el op m en t ( an d

B at eg ek a (2

3)

. T he d at a ar e co nf irm ed in v ar io us a rti cl es

: i nc lu di ng b y A nd re w M w en da in

T he

In de pe nd en t 1

Ja nu ar y

, t itl ed

U ga nd a ha s 3 rd la rg es t c ab in et in th e w or ld

D at a us ed to in fo rm p ub lic fi na nc ia l m an ag em en t r ef or m c al cu la tio ns a nd p er sp ec tiv es a re d ra w n fr om th e r

: a nd

Pu bl ic E xp en di tu re a nd F in an ci al

A cc ou nt ab ili ty (P

EF

A

) a ss es sm en t an d th e

O pe n B ud ge t I nd ex a ss es sm en t.

Se e w w w

.p ef a.

or g an d w w w

.in te rn at io na lb ud ge t.o rg T he a ve ra ge sc or e fo r d im en si on s t ha t a ss es s c om pl ia nc e w ith d e ju re e le m en ts o f t he P

EF

A a ss es sm en t is

.8

D at a us ed to in fo rm e ve nu e ad m in ist ra te r ef or m c al cu la tio ns a nd p er sp ec tiv es a re d ra w n fr om , a m on g ot he r s ou rc es . K id d an d

C ra nd al l 2

6.

A fr ic an D ev el op m en t B an k

. v ii

In te rn at io na l T ax ia lo gu e

; P W

C re po rt

(P ay in g

Ta xe s 2

3.

a va ila bl e at w w .p w c.

co m

/p ay iii gt ax es

D at a us ed to in fo rm p ro cu re m en t r ef or m c al cu la tio ns a nd p er sp ec tiv es a re d ra w n fr om th e

O

EC

D b as el in e pr oc ur em en t a ss es sm en t o f U ga nd a’ s s ys te m . h ttp w w w .o ec d.

or g/ de ve lo pm en t/e ff ec tiv e-ne ss

/3

9.

pd f T he d at a ar e sh ow n as a ve ra ge sc or es o ut o f 3

. T he a ve ra ge sc or e in

U ga nd a fo r t he L eg al a nd

R eg ul at or y

Fr am ew or k in p ro cu re m en t i s 2

.9 3.

D at a us ed to in fo rm a nt i-c or ru pt io n re fo rm c al cu la tio ns a nd p er sp ec tiv es a re d ra w n fr om th e a nd

G lo ba l I nt eg rit y su rv ey s, w w w

.g 1o ba lin te gr ity .o rg

O164 | ANDREWS

a similar score (65%) on the 2012 OBI, which assesses the degree of PFM system transparency. This score was higher than all other east African countries. Kenya scored 49%, which is considered ‘mod-erate.’ Uganda’s score fell into the ‘substantial transparency’ category of countries scoring between 61% and 80% (the same level as various Organisation for Economic Co- operation and Development (OECD) countries).6

Reform gains are also evident in revenue administration, especially in the URA. Created in the early 1990s, the URA ramped revenue collections up from about 7% of gross domestic product (GDP) (in the late 1980s) to about 11% of GDP in 1998 (African Development Bank, 2010, p. vii; Kidd & Crandall, 2006; Therkildsen, 2004). This addressed a key problem experienced by the Ugandan gov-ernment—of low resource mobilization and high dependence on outside financing. The government was considered ‘a trail blazer among the countries of the region [given the act of] establishing an autonomous revenue agency’ (African Development Bank, 2010, p. 6). The URA has continued to im-press, with a high formal level of autonomy and large scope of activities. It scores 9 of 9 points for the International Tax Dialogue assessment of revenue agency autonomy, and scores 7 of 9 when looking at the areas over which it exercises policy and administrative influence (International Tax Dialogue, 2010; Kidd & Crandall, 2006, p. 89). It also ensures a system that performs well on global indicators measuring the ease of paying taxes. A 2012 analysis showed that the URA has managed to keep tax rates low (with total tax rates of 37%, compared with a global average of 43%) and has minimized the time needed to pay taxes (at 213 hours, compared with the average of 276 hours globally). This highly autonomous and efficient tax agency has been studied and lauded across the African continent.

The same could be said of the legal and organizational infrastructure created for public procurement.

The framework emerged through the 1995 Constitution, 1997 Local Government Act (amended in 2006), 2003 Public Procurement and Disposal of Public Assets (PPDA) Act, 2006 Local Government PPDA Regulations, and 2008 Local Government PPDA Guidelines (Basheka, 2008). These (and other) interventions have led to a best practice deconcentrated model in the country, where local gov-ernments and ministries are responsible for doing procurement planning and managing procurement transactions; with a central agency (the PPDA) creating rules and providing monitoring and oversight.

The promising nature of this system—and gains already achieved through such—is described in re-cent comments from Anja Kramer (of the German aid agency KfW) and Howard Centenary (of the World Bank). In discussing the PPDA’s 2012 Public Procurement Symposium and Exhibition, they note that, Procurement is now decentralized to the spending agencies and the Public Procurement and Disposal of Assets Authority (PPDA) is now well established regulator … These re-forms have emphasized transparency in procurement … [and] over 80% of government expenditure through procurement is subject to open bidding with these contracts adver-tised in public media… Establishment of a procurement grievance handling mechanism is another achievement… PPDA has improved its monitoring system and has to date conducted 329 audits in 221 procuring agencies since its establishment in 2003 (

New Vision, 2012).

Procurement reforms also yield strong scores on donor- driven assessments. Uganda’s legal and regu-latory framework scored 41 of 42 points in the 2007 OECD procurement assessment (averaging 2.93 out of 3). This suggests that laws complied about 98% with international ‘good practice’ in 2007, which was an improvement from the score in 2005 (where the average was 2.5 out of 3). This data is corroborated

6Rwanda scored 8, Tanzania 47 (‘Uganda beats Kenya’, 2013).

| O165ANDREWS

by Global Integrity’s assessments of the procurement system, which give full marks (100 of 100) for a variety of legal innovations.

Uganda’s procurement reforms have always been tied to the larger goal of anti- corruption. There are other dimensions of the anti- corruption reforms as well, however, which build on the 1970 Prevention of Corruption Act. This was amended in 1988 as one of the new government’s first public sector reform interventions; to prevent people accused of corruption from disposing of money in their bank accounts. The 1995 Constitution also directly addressed corruption, stipulating that, ‘all lawful mea-sures shall be taken to expose, combat and eradicate corruption and abuse or misuse of power by those holding political and other public offices’ (Constitution of the Republic of Uganda, Article XXVI).

Between 2000 and 2005, the government introduced ‘[f]urther legal instruments for fighting corrup-tion includ[ing] the Inspectorate of Government Act, the Government Finance Regulations, the Public Service Act, the Public Service Standing Orders, the Local Government Act and the Leadership Code Act’ (Ittner, 2009, p. 82). It continued introducing such institutional structures after 2005, including a Code of Conduct and Ethics, Anti- Corruption Act, Whistleblowers Protection Act, and Anti- Money Laundering Bill.

In reflecting on these kinds of interventions in 2011, Global Integrity gave Uganda 98 points out of 100 for the quality of its legal framework. This was higher than any other country in the world.

The United States scored 90. Kenya scored 84. Uganda registered full marks on 128 of 135 questions about the nature of laws considered important for establishing an accountable government capable of addressing corruption problems. This equates with having full compliance with legal requirements in 95% of the areas analyzed. The United States only registered 90% on the same measure and countries like Kenya and Ghana were further behind on 84% and 79% respectively. This strong legal compliance ensured that Uganda got a score of 72 out of 100 for the overall strength of its ‘integrity system’, which is higher than most other comparator countries (Kenya scored 68 and Ghana 67). This performance once again reflects positively on the gains and achievements of the past three decades of reforms.

On paper, at least, Uganda looks like it has an impressive government that should be able to function effectively.

3.2 | Uganda’s government is not as functional as it now looks Unfortunately, Uganda’s government is not as good as it looks. In practice, reforms in many areas have been limited to appearance and do not yield the effectiveness needed for a developing country with major problems getting services delivered or addressing poverty challenges or socio- economic deficiencies. Table 2 shows reform limits and gaps in the five areas explored above.

Civil service and public administration reform limits began to manifest in the late 1990s, when government employment started expanding (even after the 1990s’ cuts). Various policy decisions (like decentralization and universal primary education) led to hiring pressure (in local governments and schools, for instance) and contributed to a major growth in the number of government employees.7 Precise data is difficult to come by, given the many different conditions of employment for public servants, but various sources suggest that government now employs between 290 and 330 thousand people—double the amount they had in place in the late 1990s.

Public hiring would not be a worry if Uganda could finance its employment in a sustainable man-ner and if employment led to better service delivery. Unfortunately, evidence does not support such a positive version of the government expansion. The government has not improved its capacity to afford

7Andrews (2003a) shows that reforms are commonly in tension; decentralization reforms especially often yield conflicting messages for those trying to introduce fiscal stability.

O166 | ANDREWS

A

B L

Li its a nd

G ap s i n

U ga nd a’ s P ub lic S ec to r R ef or m

R es ul ts

R ef or m

A re a C iv il

Se rv ic e

R ef or m a A nd

G ov er nm en t

Pe rf or m an ce

Pu bl ic F in an ci al

M an ag em en t

R ef or m b

R ev ue dm in ist ra tio n

R ef or m c Pr oc ur em en t R ef or m d

A nt i- C or ru pt io n

R ef or m e

C ur re nt li st itu tio na l ch an ge

G ov er nm en t s iz e ha s gr ow n si nc e

U ga nd a ha s 7 ca bi ne t m in is rs no

, a nd th ou sa go rn m en t em pl oy ee s;

be rs o f g ho ke rs se em s hi gh er th an b ef or e, f ar re ar m on pl ac e, ro ls ck in g

N at io na l p la ns h av e no t be en

‘c on sc io us ly im em en te d’ pe rf or m an ce is lo w ;g o f N at io na l

D ev el op m en t P la n pe rf or m an ce ta rg et s ar e no t e ve n as se ss ed o f t ar ge ts ss ed a nd a ch ie d;

of

‘n at io na l c or e pr oj ec ts ’ h as b ee n ac hi ev ed

, a nd

(6 3%

) a re o ff tr ac k e is n o as se ss m en t

A ct ua l i m pl em en ta tio n co m pl ie s o nl y w ith P

EF

‘g oo d in te rn at io na l pr ac tic es ’h ;C ur sp di ng e xc ee ds do es tic re ve nu es th e co un try h as n o do m es tic fi sc sp e fo r i nf ra st ru ct ur e sp en di ng

Tr an sp ar en cy o f bu dg et ex ec ut io n ly

8% o n th e O

B I;

co m pa re d w ith

3% fo r tra ns pa re nc bu ar at io n.

Th e is li m ite d in fo rm at io tu al sp en di ng

Ta x co lle ct io n ha s st ag na te d si nc e la te

0s

. H as he d o f G

D P

(c om pa re d w ith A fr av ag e of

0% ) i n an y ye ar b ut

/1

(d e-of f a dd iti on s fr om o il re ve nu es

Sc or es fo r c ap ac ity op er at io ns

, a tra pa re nc y in th e sy st em a re

0, o r 6 6% i ; b id s ta ke o ve r a y ea r t o co m pl et e;

o nl

1% c on tra ct pl et et a nd o nl y a re c om pl ti m e;

o nl y c on tra ct s i n

M in is tri es

D ep ar tm en ts ge nc ie s h pl et e pr oc ur em en t re co rd s i n

/1 2j

Im pl em en ta tio s s co re s 5 ou t o f 1 o n

G lo ba l I nt eg rit y

Sc al e.

C os tly rr up tio n cr ar co m m on pl ac e an d co rr up tio id rv as iv e an d hi gh

(c on tin s)

| O167ANDREWS

ef or m

A re a C iv il

Se rv ic e

R ef or m a A nd

G ov er nm en t

Pe rf or m an ce

Pu bl ic F in an ci al

M an ag em en t

R ef or m b

R ev dm in ist ra tio n

R ef or m c Pr oc ur em en t R ef or m d

A nt i- C or ru pt io n

R ef or m e

G ap s i n cu rr en t pr oc es se er fo rm ag ai ns t g oa ls

Th e ga p be tw ee n th e si ze o f U ga nd a’ s go ve rn m en t a nd it s go al s s et in th e ea rly

0s si gn ifi ca nt

; T be r o f m in is go ve rn m en t em pl oy ee s i s be tw ee n

0% h ig he r t ha n ta rg et s s et k a ro ls ef fe ct iv e o f N at io na l

D ev el op m en t P la n ta rg et s a re e ith er ss ed o r n ev ed

. T hi s m ea g ov er nm en t h rf or m an ce p la n bu t ha lf of it h as n o va lu e (w ith n o as se ss m en

t) an d a qu ar te r o f i t re fle ct s e xp lic it im pl em en ta tio n fa ilu re ; t he re ila r f ai lu re ra te fo r o f ‘ na tio na l c or e pr oj ec ts

Th e ga p be tw ee n la w s a nd im pl e-m en ta tio n of th e la w s i s 2 0%

Th e ga p be tw ee n tra ns pa re nc ul at io n an d ex ec ut io

G ov er nm en t re ve nu e is lo w w ith a

5% de nd en ce o n ex te rn al fi na nc in g an d a fin an ci ng ga th at li fr as tru ct ur e an d ot he r e xp en di

(w ch is lo in m os t o ic an c ou nt rie s) l

Th e ga p be tw ee n la w s a nd c ap ity pl em en t l aw s is

3%

; t he ap be tw ee n la w al im pl em en ta tio n is e ve n la rg er

In

1, 9% o f co nt ra ct s w er e ov er b ud ge t, e ov er ti e, th e av er ag e bi d to ok se ve n m on th s l on ge r t o co m pl et e ta et ed

Th e ga p be tw ee n la w em en ta tio n is a bo ut ou t o f

0, o ne o f t es t i n th e w or ld . N ew la w s an d sy st em s h av e no t c ur be d co rr up t be ha vi ou r

PF

M re su lts a re n eg at iv el y af fe ct ed b y cu rr en t g ap s:

A ct ua l s pe nd in g ha s re gu la rly d ev ia te d si gn ifi ca nt ly et s ( ex ce ed in g

Th e si tu at io n is w or se fo r l oc al g ov er nm en ts w al sp en di ng d ev ia te s s ig ni fic tly om b ud ge ts m a D at a ar e dr aw n fr om th e

U ga nd a

C om pu te r S er vi ce s, M in is try o f F in an ce

, P la nn in g an d D ev el op m en t a nd th e U ga nd a B ur ea u of

S ta tis tic s (

Th ey a re c on fir m ed in v ar io us a rti cl es

, i nc lu di ng o ne by nd re w M w en da in

T he

In de pe nd en t, Ja nu ar y

, t itl ed

‘U ga nd a ha s 3 rd la rg es t c ab in et in th e w or ld b D at a ar e dr aw n fr om th e

P

EF

A a ss es sm en t a nd th e

O pe n

B ud ge t I nd ex a ss es sm en t.

Se e w w w .p ef a.

or g an d w w w .in te rn at io na lb ud ge t.o rg c In fo rm at io n dr aw n fr om v ar io us so ur ce s i nc lu di ng

In te rn at io na l T ax

D ia lo gu e (2

0) d D at a ar e dr aw n fr om th e

O

EC

D b as el in e pr oc ur em en t a ss es sm en t o f U ga nd a’ s s ys te m

. h ttp w w w

.o ec d.

or g/ de ve lo pm en t/e ff ec tiv en es s/

.p df e D at a ar e dr aw n fr om th e a nd

G lo ba l I nt eg rit y su rv ey s.

w w w .g lo ba lin te gr ity

.o rg f U ga nd a

B ur ea u of S ta tis tic s, Se pt em be r 2

2, sh ow s t he o ff ic ia l s iz e of th e ci vi l s er vi ce a t 2

,0 g D at a is d ra w n fr om R ep ub lic o f U ga nd a, 2.

h Th e av er ag e sc or e fo r d im en si on s t ha t a ss es s c om pl ia nc e w ith d e ju re e le m en ts o f t he P

EF

A a ss es sm en t i s 2 i Th e th re e ar ea s as se ss ed o th er th an le ga l a nd r eg ul at or y fr am ew or k ar e tit le d;

I ns tit ut io na l F ra m ew or k an d M an ag em en t C ap ac ity

, P ro cu re m en t O pe ra tio ns a nd M ar ke t P ra ct ic es a nd I nt eg rit y an d

Tr an sp ar en cy o f t he P ub lic

P ro cu re m en t S ys te m

. U ga nd a av er ag es o ut o f 3 , 2

.2 o ut o f 3 , a nd

.9 o ut o f 3 fo r t he se th re e ar ea s.

j G ov er nm en t A nn ua l P er fo rm an ce R ep or t, FY

1/

, V ol um e

1, k Th e nu m be r o f m in is te rs is h ig he r t ha n al l c ou nt rie s i n A fr ic a ex ce pt fo r K en ya l G ov er nm en t A nn ua l P er fo rm an ce R ep or t, FY

1/

, V ol um e

1, 8.

D om es tic re ve nu e m ob ili za tio n re m ai ns in ad eq ua te , l ea di ng to c on tin ue d re lia nc e on lo an s a nd g ra nt s.”

m A s p er

Ja nu ar y lo ca l g ov er nm en t P

EF

A a ss es sm en t.

T A

B L

(C tin d)

O168 | ANDREWS

such a large workforce and service results do not appear better in the past decade. Furthermore, there seems to be little de facto control in the civil service regime, reflected in a festering problem of ghost workers and arrears across government. Audits have found ghost workers in fields as varied as health, policing and education. Arrears are also commonplace, again in many different areas of government activity. Dating back to the early 2000s, one can find civil society reports of such arrears bubbling onto the streets in the form of strikes or rallies by policemen, university professors, teachers, doctors and other government employees. Official reports do not suggest that arrears are a major fiscal prob-lem to the state, but the regularity of salary arrears (lasting for months at a time) indicates control and management weaknesses in the system.

These problems persist despite many institutional ‘solutions’ (including commitment control mechanisms and the IFMIS).8 The IPPS is meant to enhance control, but it is too early to determine if it will succeed or produce the same limited results of past technical interventions.

The limits of past reforms are further evidenced when looking at performance management and plan implementation in Uganda. The country led in introducing results management and planning, but has had limited success in implementing plans. Performance mechanisms (including budgets for organizations and contracts for managers) are still not used effectively (with budget allocations still not tied to results, for instance) (Williamson, 2003, p. viii). Beyond this, national plans have not been ‘consciously implemented’ and performance is low (Republic of Uganda, 2012); 56% of National Development Plan performance targets are not even assessed; 22% of the targets that have been as-sessed have not been achieved; only 1 of 16 ‘national core projects’ has been achieved, and 10 (63%) are off track or there is no assessment.

In short, there is a gap between the systems that exist in form and their functionality or impact. This gap is most evident when considering the expectations citizens have of service delivery and execution by government. Data on this is difficult to come by, with Afrobarometer surveys providing the most accessible source.9 According to this source, citizens have little faith in government’s ability to do key things; 79% think that government manages the economy ‘very badly’ or ‘fairly badly’; 77% answer similarly when asked if government handles the problems of poverty well; 82% think government manages equity issues badly; 51% and 44% view government health care and education service provi-sion negatively; 54% feel that government handles water and sanitation services ‘very’ or ‘fairly’ badly; 67% and 68% note that government handles maintenance of national roads and bridges and local roads poorly; 78% think similarly on electricity supply; and more. These data need to be treated with care and are used here simply to show that there is a deficit of trust and expectation on the part of citizens—meaning they do not see government as effective even after three decades of reforms and the adoption of ‘best practice’ mechanisms in the civil service and planning realms.

Similar gaps exist in PFM. Whereas reforms have produced a PFM system where laws and pro-cesses comply with over 70% of the ‘good international practices’ measured in the multi- donor PEFA assessment, implementation complies only 50% with these practices.10 This means that Uganda has many areas in its PFM system where laws and processes are not being properly used or implemented.

The PEFA scores are high (A or B) when asking if a budget calendar exists but the scores are low (D)

8A February 2006 IMF Country Report (No.06/43) reflected on the failure of the first of these and potential of a second. On page 2, it reads, “The Commitment Control System (CCS) was adopted in the late- 1990s to limits spending commitments at the ministry level to within their cash allocations. However, arrears continue to accumulate under the CCS. IN 2004/05, an esti-mated 32 billion Ugandan shillings of new arrears were accumulated. The government is currently rolling out an IFMS that should allow better tracking and control of arrears. The new system appears to be effective in pilot ministries.”

9These are 2012 data. www.afrobarometer.org/files/documents/summary_results/uga_r5_sor.pdf 10This is based on the analysis of PEFA data using methods developed in Andrews (2011), where one identifies dimensions of the PEFA analysis that relate to de facto performance. Uganda scores 2 out of 4 on such dimensions.

| O169ANDREWS

when assessing if budget preparation has been completed in the calendar period.11 The gap between de jure form and de facto functionality is significant and can be found throughout the PFM system.

Various authors have written about it. In 2003, for instance, Williamson (2003, p. viii) identified that whereas government had invested in adopting an output/outcome budgeting system, and ‘Sectors are required to analyze past performance of outputs and outcomes … and to set future output and outcome targets to be achieved over the MTEF period … The Ministry of Finance has tended to continue with incremental budgeting … and not allocate funds on the basis of performance.’

Another gap exists between the quality of upstream budget preparation mechanisms and down-stream execution mechanisms. This is reflected in PEFA scores (where the budget preparation mech-anisms score an average of 2.8 of 4 but execution mechanisms score 2.1) and in the recent OBI data.

The OBI data show that Uganda has significant transparency in the way it produces budgets and communicates what it plans to spend (scoring 73% on the questions about transparency in budget preparation). However, it only satisfies 48% of the requirements associated with execution transpar-ency (where the focus is on transparency of what was spent and what results ensued). This means that reforms provide an effective window onto the promises of government but leave it very difficult to see the actual activities and impact of government. It also means that PFM systems do not provide the information government needs to ensure spending is effective and efficient, goals are being met and lessons are being learned about how to execute plans. The continued weakness of implementation sys-tems—even after decades of reform—manifests in the regular deviation of actual spending from bud-gets (this deviation regularly exceeds 10%) and is a reason why service delivery remains deficient and why the country struggles to find fiscal space for important spending (like infrastructure spending).

The gap between appearance and reality is also apparent in the area of revenue administration. The country has a revenue agency that satisfies many ‘good practice’ criteria, but it is also considered corrupt: 69% of the 2012 Afrobarometer survey respondents had ‘a little’ to ‘not at all’ trust in the Uganda Revenue Authority and 90% felt that ‘some’, ‘most’ or ‘all’ URA officials and other tax col-lectors were corrupt. This perspective is reflected in various surveys and studies, with a 2007 Transparency International study showing the URA as ‘the most corrupt tax institution in East Africa’ (Mugerwa, 2009).12 The URA performed similarly poorly on the 2010 East African Bribery Index, which portrayed it as the most corrupt public entity in Uganda (Habati, 2010).

Beyond this, the URA has not been very effective as a revenue collection agency. Collections stag-nated after 1999 and have been stuck between 11% and 13% (of GDP) in every year but 2011 (when growth was because of one- off deals and early impacts of oil revenues) (African Development Bank, 2010; Ronald, 2011, p. 1). The percentage is low compared with other African countries—where the average tax to GDP ratio exceeds 20% (Drummond, Daal, Srivistava, & Oliveira, 2012; Ronald, 2011,

p. 1).13 The problem of low government revenue means that Uganda is heavily dependent on external financing (with about 25% of its budget funded by outside agencies). The country has a perpetual fi-nancing gap that limits infrastructure and other expenditure.

Problems in the URA (with corruption and weak tax collection) fester because tax reforms have not addressed key constraints. These include continued reliance on informal markets, low ‘tax morale’, 11The relevant dimensions are 11.i and 11.iii.

12The 2005 Afrobarometer showed that 60% of the public believe tax officials are corrupt (Ittner, 2009, p. 66). Corruption in Uganda’s URA is discussed by Fjeldstad (2005).

13Interviewees in the URA disagreed with the idea that they had low levels of revenue collection in 2005. When asked if the URA had helped to increase tax revenues at the time, they allotted themselves full marks (10 of 10) (Kidd & Crandall, 2006, p.91). The record at the time did not reflect such sentiment at all, with revenue to GDP ratios in 1997 being 12% and sitting at about 13% in 2004 (Kidd & Crandall, 2006, p. 93).

O170 | ANDREWS

low trust in government and resulting failures to pay tax, low political support for revenue reforms, and bureaucratic and political influence of the URA (Kidd & Crandall, 2006, p. 32). In respect of the last issue, observers note that the ‘semi- autonomous’ URA is anything but ‘prone to close oversight and patronage by the national political leadership’ (African Development Bank, 2010, p. 6; Fjeldstad, 2005; Therkildsen, 2004). There is a further perception that tax incentives and exemptions have been imposed on the URA by political leaders—even though the URA’s semi- autonomous status should limit such influence. While many perspectives on this issue are possible, it seems fair to conclude that Uganda’s revenue reforms have not yielded the results that were needed—reform limits leave gaps be-tween appearances on paper and practical functionality. Ali Ronald (2011, p. 19), a Makerere student, wrote as much in 2011, observing ‘The continued poor domestic revenue mobilization despite many tax reforms established since the 1990s …’

The procurement area offers similar evidence. For instance, Cornelia Sabiiti from the PPDA noted that new processes introduced since 2002 were not being implemented.14 An assessment showed de-creased compliance with process requirements between 2008/09 and 2011/12. Compliance was around 40% for contract award and record keeping (meaning that only 40% of contracts were awarded and records kept according to the mandated processes). The PPDA targeted 100% compliance with com-plete record keeping requirements but in 2011/12 only 17.1% of the contracts complied with this. (The highest level of compliance was 32% in 2008/09, which shows that new formal processes have never been well implemented and that the gap between form and function is growing). The assessment also found that the average time taken for all kinds of bids was longer than government- set targets. The target for open international bidding is 155 days but the average performance in 2011/12 was 341 days.

The lowest level was 206 in 2008/09. This is an average delay of 7 months above the target of 5 months—a significant performance gap. The target for open domestic bidding is 132 days but the av-erage performance in 2011/12 was 334. The lowest level was 162 in 2008/09. In 2011/12 only 0.7% of assessed procurements were implemented within the budget estimate (monetary allocation for the contract) and only 29.4% of the contracts were completed within the original contract time; 99.3% were over budget and 71% were over time and late.

Beyond this, the PPDA’s 2016 Procurement Integrity survey reports that suppliers spend up to 20% of contract amounts on corrupt payments, up from 10% in 2008. This amounts to about $240 million annually in bribes. Such problems exist despite the many ‘best practice’ reforms introduced in Uganda. This is reflected in comments from Anja Kramer of KfW and Howard Centenary (from the World Bank) who note that various new mechanisms are not working well (the limited parts of reform are shown in bold):

[The] [e]stablishment of a procurement grievance handling mechanism is another achievement … However, the use of this grievance mechanism is still limited, with only about 30 complaints received annually at PPDA for possible fear of bidders being persecuted through denied future opportunities … PPDA has improved its monitoring system and has to date conducted 329 audits in 221 procuring agencies since its es-tablishment in 2003 … The task ahead is to increase monitoring and audit coverage to ensure that a substantial proportion of the budget expended through procurement is audited by both PPDA and the Auditor General … While these achievements are

14Cornelia Sabiiti, the Executive Director of the PPDA gave a Powerpoint presentation titled ‘Performance of the Public Procurement and Disposal System in Uganda’ at the Public Procurement Symposium & Exhibition, 19–20 September 2012.

The information in this presentation is reinforced in the PPDA’s public announcement supporting the 2012 Public Procurement Symposium & Exhibition (‘PPDA Supplement 33’ (2012, September 19) New Vision.).

| O171ANDREWS

laudable, some challenges that constrain the achievement of the main objectives of value for money and efficiency in service delivery still remain. Corruption especially through bribery in procurement is one of the major vices that impedes achievement of value for money … Procurement delays also contribute to ineffective service delivery … There are weaknesses in contract implementation especially in performance by contractors …

(New Vision, 2012).

These sentiments are reflected in various articles, including Agaba and Shipman (2007) and Basheka (2008). The latter notes that procurement laws are not being implemented in local governments (in par-ticular). Basheka’s survey across districts showed fundamental challenges were not dealt with, including the lack of functional and effective procurement and disposal units (created in name, but with no appoint-ments) and problems in carrying out market surveys (and ensuring local markets could facilitate the kinds of procurement being planned) and collusion in bid preparation among district staff. Such constraints are reflected in the weak scores Uganda gets on OECD procurement assessments of capacity, operations and transparency. In 2007 the scores were 80 out of 120 (or 66%) for these factors. Given that laws satisfied 99% of OECD survey requirements, the gap between laws and the capacity to implement laws was 33%.

The gap between laws and actual implementation is even larger, given that (as already shown in 2011) 99% of contracts were over budget, 71% were over time, and the average bid took seven months longer to complete than targeted. This reveals a major challenge of reform implementation that is also borne out in Global Integrity data (where Uganda scores lower on questions asking about practice than it does on questions about legal form).

The situation is similar when looking at anti- corruption reforms. As already shown, Global Integrity give Uganda 98 out of 100 on the quality of anti- corruption laws. The country is in full compliance with 95% of the questions asking about the quality of laws. But it scores 51 of 100 when the survey asks about implementation of laws, with full compliance on only 11% of the questions about implementation and practice. The gap between form and function is 47, which is nearly twice as high as in places like Sierra Leone (Kenya is at 30), showing that Uganda has the best anti- corruption laws in Africa and one of the poorest records of implementation. This is a problem with real implica-tions. Costly corruption crises are commonplace, and corruption is considered pervasive (with crises in the health sector, a scandal about bicycle procurement, malfeasance in microfinance, a pensions scandal and fund losses in the Peace Recovery and Development Plan). Transparency International’s Corruption Perceptions Index provides further evidence of corruption’s pervasiveness. The score was 29 in 2012 (out of 100) and the country is ranked 130 out of 176 countries (dropping from 111 of 179 in 2007). Reflecting on the situation, the think tank Global Integrity describes Uganda as a place where there has been ‘Tough talk, much noise, no improvement’ (Turner, 2009). Ittner (2009, p. 64) writes similarly: ‘This trend indicates that the perceived level of corruption has hardly, if at all, im-proved during the last decade despite numerous and highly publicized attempts at reform.’

Ittner (2009) notes that reforms have had limited success partly because they did not deal with the contextual realities of corruption in Uganda. He argues, for instance, that the main enforcement body is the police, which is widely considered one of the most corrupt entities in Uganda. This is reflected in government’s own data, like the 2003 National Integrity Surveys, which show that half or more respondents see the police as corrupt—given actual experiences of having to pay bribes (not perceptions).