Revisiting state intervention : State-sponsored micro-credit and poverty reduction in Uganda

Whereas it is often argued that state intervention with credit to reduce poverty is a futile venture, our research findings in Uganda show that while this thesis may be true in some cases, it should not always be taken for granted. Hence, we argue that state-sponsored credit schemes can also perform better provided their programme designs are fortified with best practices. We demonstrate this case using a comparative study of two state-sponsored micro-credit schemes targeted at reducing poverty. While the performance of ECS was dismal, PAP was able to register impressive performance in terms of institutional efficiency and participants’ welfare.


INTRODUCTION
There has been growing acceptance that micro-credit can play a productive role in the global fight against poverty since the 1990s (World Bank, 1990).The prominence of micro-credit can partly be explained by the realization that the poor are bankable as well as successes recorded by microfinance organizations such as the Grameen, BRI and Bancsol in Asia and Latin America respectively.It can also be explained by the increasing poverty levels in the world at a time when neo-liberal reforms have also spurned economic growth in many of the developing countries.The global population afflicted by poverty is currently estimated at about one billion (Morduch, 1999).The proportion of people living in absolute poverty is even much worrying in poor countries such as Uganda.Until recently, people living in absolute poverty in Uganda had declined from 56% in 1992 to 34% in 2000.However, this positive trend was reversed after 2000 with poverty levels increasing from 34 to 38% (Uganda National NGO Forum, 2004) but later declined again to 31% in 2006.The challenge of bourgeoning poverty has therefore increased the support for micro-credit as an anti-poverty strategy at least in the short-term (Sebstad and Chen, 1996).
While there is growing acceptance that micro-credit can play a positive role in poverty reduction, there is no consensus on the direct involvement of the state in the provision of micro-credit to the poor people.The lack of consensus is manifested in the controversial debates about state intervention in credit markets.The advocates of state-sponsored credit (Huppi and Feder, 1990;Barham et al., 1996;Khandker, 1998) justify its role on the grounds that formal financial institutions discriminate against the poor and remote areas.They also argue that state intervention with cheap credit protects the poor from loan sharks, promotes social justice, and leads to the rapid spread of the benefits of a modern economy such as investments in technology and inputs that consequently revolutionize agricultural production and income (Fuentes, 1996).It is within the same context that Entandikwa 1 Credit Scheme (ECS) and Poverty Alleviation Project (PAP) were conceived.Nonetheless, critics of concessional lending (Adams and Vogel, 1986;Adams et al., 1984;Hoff and Stiglitz, 1993;Braverman and Guasch, 1993;Baydas et al., 1994;Yaron, 1994) contend that it distorts credit markets and subsequently leads to inefficiencies and inequities.It is also criticized for perpetuating patronage practices, serving elite interests (Robinson, 1994;Basu, 1997), and also for perceiving the poverty challenge as a simple problem that can be tackled by capital injections instead of overhauling E-mail: wmuhumuza@yahoo.com.the whole socio-economic structure (Buckley, 1997).It is further argued that subsidized credit schemes have had disappointing results.They are frequently associated with high arrears and big losses both to the financial institutions administering the programmes as well as for the government and donor agencies ( (Buckley, 1996).
The criticism against state-sponsored credit has even found fertile ground in the renaissance of classical neoliberalism especially since the 1980s.The neo-liberal policy environment prescribes to the state a role of creating a conducive policy environment for the credit markets to thrive unfettered by state interventions.While the neo-liberal climate has reigned for over two decades, the liberalization of credit markets that has witnessed a proliferation of private players has not made any new remedies to the traditional obstacles that continue to prevent the poor's access to credit.Private sector financial institutions engaged in micro-credit regardless of whether they are owned by private actors or NGOs have continued to be driven by a financial systems approach that continues to discourage the poor, and have neither effectively penetrated the remote areas nor served the real poor.They have also not willingly supplied agricultural credit.It is because of the down-side of private sector-managed credit that the state periodically intervenes with cheap credit.
While the state has continued to directly intervene in credit markets, the results have generally been disappointing.Nonetheless, the state's dismal performance need not be used as a justification to completely rule out its role in sponsoring credit.What needs to be addressed is how to make state-sponsored credit effective rather than dismiss it.This is necessary because states in poor countries such as Uganda remain major determining factors especially where the capacity of financial institutions is still limited.Besides, the structures and processes of a modern economy are yet to blossom so as to allow market forces to be fully unleashed.The full monetization of the economy remains a challenge especially where land markets are undeveloped and the informal sector still predominant.Given such circumstances, the role of the state in credit markets cannot simply be wished away.
This paper therefore argues that instead of diminishing the role of the state in credit markets because of poor performance, its capabilities should instead be strengthened.This is because the state has more potential advantages than the private sector in credit management because of its legitimacy, vast resources, and the formallegal infrastructure backed with the threat or use of force.Hence, what is required is strengthening the effectiveness of the state in managing credit.The paper further argues that where the state's technical capacity was strengthened with technical support, its performance was fundamentally enhanced as the case of Poverty Alleviation Project (PAP) demonstrates.Due to technical support from the African Development Bank (ADB), PAP was able to outperform ECS, a parallel state-sponsored credit scheme.
The proceeding sections focus on the two experiences with a view of justifying the contention that state-sponsored credit can be effective if the state's technical capacity is bolstered.The paper begins with a brief background of the two credit schemes and that of the participants after which it gives a historical overview of state-sponsored credit schemes.It further examines the programme designs of the two credit schemes and their impact on performance.It then assesses the impact of the two credit schemes on the recipients' economic wellbeing as well as the policy implications of such programmes.

BACKGROUND TO STATE-SPONSORED CREDIT
The popularity of micro-credit in Uganda's policy circles has fundamentally grown under the National Resistance Movement (NRM) regime.Micro-credit has been used both as an antipoverty strategy and a political instrument since 1988 with varying outcomes as will be discussed in the proceeding sections.Two state-sponsored microcredit schemes namely, PAP and ECS were initiated in 1994 and 1995 respectively.Whereas ECS was solely funded by Government of Uganda (GOU) (Republic of Uganda, 1999;Republic of Uganda, 1996:6) with Ush.6 billion (US$6 million), PAP was funded with an ADB loan of US$13.5 million and co-funded with US$1.5 million by the GOU (Republic of Uganda, 1998).
The origin of the two government-directed micro-credit schemes needs to be understood within the broader poverty policy perspective of accelerating poverty reduction through broad-based economic growth while focusing public expenditure on the key sectors of the economy that enhance income earning opportunities for households of the poor people, raise labour productivity, and improve people's welfare through the provision of basic social services.While the overall anti-poverty strategy is dominated by the need to raise household incomes, the anti-poverty action planthe Poverty Eradication Action Plan (PEAP) 2 -took cognizance of the fact that the high economic growth rates were not having a "trickle-down" effect on the poor as anticipated (Republic of Uganda, 1999).Hence, PAP and ECS were set up to stimulate the creation of income-generating activities among the poor and consequently reduce poverty.
It is against this background that this paper presents findings based on research conducted on ECS and PAP using a case study of two districts of Mbarara and Mpigi in Uganda.It involved a multi-phased sample of 40 heads of households benefiting from state-sponsored microcredit and 20 non-benefiting household heads (control group) in the two districts.Their socio-economic characteristics showed that access to credit largely (57%) went to the age bracket of 31-50 years that comprised people actively engaged in rural economic activities.According to Uganda's state-sponsored credit guidelines, eligibility for credit access was 16 years and above (Republic of Uganda, 1995).Credit recipients within the age range of 16-30 years and 51-65 years were few (27 and 17% respectively).More particularly, the 51-65 years group was hesitant to acquire loans because they considered it a risky venture that could easily lead to vulnerability.Whereas 53% of credit recipients were primary school leavers and therefore literate, 3 most of them lacked socioeconomic-related skills that are associated with higher education such as basic accountancy, record keeping and management skills.On the other hand, 46% had completed senior one and above.Only 1% had no formal education due to reasons related to absolute poverty and cultural obstacles.These statistics on educational status show that many less educated people participated in credit, which creates an impression that it was well targeted since the very poor tend to be characterized by lack of or low educational levels.The fact however is that the two credit schemes targeted the "active" poor 4 as opposed to the absolute poor.In terms of family size, the majority (87%) had four and above dependants.This can be attributed to the phenomenon of extended family which is peculiar to rural household.This cultural responsibility especially in the absence of a state welfare policy has been exacerbated by the advent of HIV/AIDS in Uganda, which has left behind many widows and orphans.

A HISTORICAL REVIEW OF STATE-SPONSORED CREDIT SCHEMES
State-sponsored credit schemes in Uganda have had similar objectives irrespective of being initiated by different regimes.Credit has been used both as an instrument to transform farmers from subsistence to modern agricultural practices as well as a strategy to reduce poverty.The first experience with a statesponsored credit scheme was in 1961.The Co-operative Credit Scheme (CCS) was introduced with the aims of improving farmers' agricultural methods and income as well as transforming and expanding the country's economy through increased agricultural production.It provided short-term loans to farmers through primary cooperative societies, with a revolving fund of Ushs.13 million.The loans given were in-kind.Farmers were provided with agriculture in-puts such as pesticides, fertilizers, spraying equipment, tractor hire services and other equipment and materials.Apart from disbursing credit through or in association with the co-operative movement in Uganda, the CCS was also linked to the extension services from the Department of Agriculture.Whereas the credit scheme initially recorded an impresssive repayment rate of 100%, this exemplary performance was short-lived as loan recoveries continued to deteriorate in the proceeding years.Defaults ranged from 0.007% in 1962 to 4% in 1971 and 8.86% in 1973 when it was subsequently abandoned.It was later revived in the late 1980s.For instance, Ush.80 million under the PL-480 programme was injected into the CCS in 1988.Additional funding came from the Swedish Co-operative Centre.The CCS recorded a moderate recovery rate of between 70 and 90% (Bank of Uganda, 1994).The impressive repayment performance can be explained by the rise in the world prices of coffee around that period.It can also be attributed to the role of Cooperative societies that Muhumuza 123 dominated the marketing of agricultural commodities in Uganda prior to the implementation of World Bank economic reforms.Hence, the elaborate co-operative infrastructure coupled with versatile mechanisms simplified credit recoveries since loans were directly recovered from sale proceeds of crops produced by members.Therefore, the best practices exhibited by cooperative societies not only allowed for the screening of borrowers but also minimized administrative costs especially where clients were dispersed in the countryside.Another state-sponsored credit scheme that operated side by side with CCS was the Progressive Farmers' Loan Scheme (PFLS) which was launched in 1964.It targeted progressive farmers selected by district agricultural officers.Loans were made available from the Uganda Credit and Savings Bank (later named Uganda Commercial Bank (UCB) with a 50% guarantee by the district authorities.This credit scheme involved 850 loans amounting to Ush.6 million.Loans extended were in-kind and farmers only accessed agricultural in-puts from selected suppliers who were directly paid by the bank.The scheme was however suspended because of high default rates.The failure was attributed to wrong mechanisms used to identify beneficiaries, a cumbersome appraisal procedure which resulted in untimely disbursement, lack of sufficient staff to supervise credit, and the absence of a readily available market for farmers' produce.A similar credit scheme, the Rural Farmers Scheme (RFS) which was launched under the auspices of UCB in April 1988 not only shared similar objectives with its predecessor state-sponsored credit schemes but also fell prey to similar problems.The RFS was intended to help farmers expand their operations, increase their output, and transform them from subsistence to commercial production.A revolving fund of Ush.6 billion (US$6 million) was made available by UCB.The RFS was characterized by favourable loan terms.It comprised character loans with an interest rate of 11 and 12% per annum for secured and unsecured loans respectively.The loan had a grace period of six months and was to be repaid in 18 months.The good intentions and favourable loan terms notwithstanding, the RFS was plagued by high default rates and recorded no significant impact on the welfare of participants.Operational weaknesses responsible for the failure of RFS included top-down deficiencies characterized by bureaucratic delays and irregularities in loan processing; questionable integrity of loan officers; poor or no monitoring and supervision of loan use; poor disbursement policy (in-kind); high transaction costs which restricted the farmer's choice to bargain for quality and price of 20% of the total disbursed; lack of collateral security for the loans; crop failure; political interference; as well as absence of efficient markets for farmers' produce (UCB, 1994).It is significant to note that whereas the CCS recorded better performance, the other two were ineffective.While there could have been other extraneous explanations, this disparity was largely attributed to the kind of programme designs and operational mechanisms used by the credit schemes in question.

TERMS AND CONDITIONS OF ECS AND PAP
According to government policy guidelines, all ECS loan applicants must have been residents of the area for at least six months and had to be guaranteed by two sureties of high reputation but not politicians.Loan applications were channelled through Chairpersons of the Village (LC1), Parish (LC2) and Sub-county Councils (LC3).The ECS involved character loans (collateral free) and initially charged an interest rate of 12% but later increased it to 16% per annum.It was charged on declining balances of the principal.The repayment period was twelve months including three months grace period.Each of the 214 Counties was to receive Ush.30 million (US$18,600) out of which 30% of the funds were to be extended to the women and youths.
In contrast, the procedure for accessing PAP credit was simple and fast.It took between two weeks and two months to disburse credit although disbursement was sometimes affected by delays in remittances of funds which consequently affected field operations.The procedure required individuals within solidarity groups to apply for loans from the Intermediary Entity (IE) 5 within that locality or directly from the Area Project Officer (APO) in case of non-existence of IEs.Whereas solidarity groups were previously composed of 7-15 members, they were later enlarged to 15-30 members to guard against having a family organize itself into a group and also to lower group contribution costs.The procedure also insisted on the feasibility of the micro-economic ventures and for applicants to acquire credit for on-going activities for which they had proven skills and experience.The credit scheme had no gestation period and also emphasized monthly repayments and mandatory savings.Members of solidarity groups were required to save 12% of the loan they applied for.PAP also charged an open market interest rate of 22%, which catered for transaction costs as well as risks involved in credit intermediation.

LINKING PROGRAMME DESIGN WITH ORGANIZATIONAL EFFECTIVENESS
It is widely acknowledged that the design of a credit scheme has major implications for its performance (Zeller, 1998;Snow and Buss, 2001).Unless the design is well crafted there is a risk of the credit scheme being ineffective.The importance of programme designs has been recognized by the World Bank which even went to the extent of conducting a research programme to enable it build case studies for best practices for micro-credit programmes (http://www.worldbank.org/).The Ugandan example focuses on two state-sponsored schemes whose varying programme designs led to different outcomes.
The organizational structure and the coordination of ECS process (Republic of Uganda, 1995) were complex.They consisted of Entandikwa Secretariat 6 which was responsible for the national coordination of ECS.Below it was the District Steering Committee (DSC) which was composed of district councilors and heads of departments (civil servants), and chaired by the Chief Administrative Officer (CAO).It was responsible for overseeing ECS activities and advising the Entandikwa Secretariat on all matters pertaining to the ECS in the district.Under the DSC was County Steering Committees (CSCs) which were largely composed of local councillors and very few technocrats.The CSC was responsible for monitoring, supervising and enforcing repayment but also monitored the performance of IAs and reported to DSC.The representative of the Intermediary Agency (IA) responsible for implementing the programme at the grassroots was also the secretary for CSC.In addition, Sub-county chiefs were responsible for following up defaulters and ensuring loan repayments.
Apart from coordinating and monitoring the ECS operations, the Entandikwa Secretariat's other responsibility was to provide a link between the Ministry of Finance and Economic Planning (MFEP) and all government departments that were implementing antipoverty credit programmes.However, there existed no clear coordination system between the Secretariat and the various anti-poverty credit schemes that were spread across different government departments.For instance, credit activities of NGOs were registered and coordinated under the Ministry of Internal Affairs while PAP and the Northern Uganda Reconstruction Project (NURP) were being coordinated under the Office of the Prime Minister (OPM).Likewise, Karamoja Development Agency (KDA) was under the President's Office and the Cotton Subsector Development Project (CSDP) under Bank of Uganda (the Central Bank).The point here is that the NRM government failed to harmonize the coordination of micro-credit policy nation-wide.One vivid indication that there was weak co-ordination between the Secretariat and other government departments engaged in antipoverty programmes was their inconsistency about repayment figures.Whereas the Commissioner for ECS put the repayment rate for ECS at 55%, the official poverty status report from the OPM put the recovery rate at 40% (Republic of Uganda, 1999) which was more authentic.It was, therefore, evident that the Secretariat's activities were not well coordinated with those of other government departments with similar objectives.
While the operational guidelines gave the national responsibility for monitoring and coordinating of ECS to the Entandikwa Secretariat, its effectiveness was lacking.It however attributed poor performance to inadequate staffing and facilitation.Indeed, the task of coordinating and monitoring the activities of ECS was challenging since it involved 214 Counties nationwide.Nonetheless, the Entandikwa Secretariat's perception of monitoring and co-ordination was a bit exaggerated in that it was equated with nation-wide physical travel which was indeed a costly exercise.This was unnecessary given that the ECS had the implementing machinery on the ground.
It was evident that the ECS structures were replicated and roles assigned to each of them were duplicated.There was lack of clarity on which office was fully responsible and accountable for what activity.In other words, the lack of clear demarcation of roles between the different agencies involved in coordinating the ECS adversely affected its operations.Therefore, the poor programme design and bureaucratic mode of operation of ECS was not only responsible for poor monitoring, supervision and recovery of loans but also led to wasteful expenditures.According to the design of ECS, each implementation agency was given a commission to cater for operational costs.The IAs were initially given a commission of 4% but later increased to 6% of the total amount of credit allocated to each County.Likewise, 1 and 0.5% were given to the CSC and DSC respectively as illustrated in Table 1.
Table 1 shows that Ush.615,600,000 (US$342,000) was spent on administrative costs of ECS between May 1995 and July 1997.This was 6.2% of the total loans disbursed.This transaction cost was unjustified since most of the loan beneficiaries defaulted and the repayment rate was only 40%.The Commissioner for Entandikwa Secretariat attributed the increase in commission to the need to motivate IAs to improve their effectiveness which objective was never achieved.The ineffectiveness of IAs was attributed to capacity limitations.Most of the IAs in Mbarara district neither had qualified staff in credit management nor possessed adequate office facilities to handle credit.For example, two-thirds of IA offices in Mbarara district did not have basic equipment such as a typewriter and were manned by unqualified personnel.The capacity limitations of IAs can also be explained by the fact that they were hurriedly created by politicians to manage Entandikwa microcredit.This was perceived to be good business since the IAs were entitled to a commission.Apart from Mpigi District where experienced NGOs (e.g.World Vision) and co-operative societies were engaged to manage Entandikwa, most IAs in Mbarara District were created by politicians.For example, a half of the ten Counties that constituted Mbarara District had IAs that had been created by members of parliament.This political interference compromised their capacity to manage credit.
The mentioned capacity limitations of IAs contributed to the mismanagement of Entandikwa micro-credit as confirmed by the Auditor General's audit report.It observed the absence of records at the districts showing amounts received; ledger accounting showing status of each loan; repayments; among others.It also noted that the management of the scheme was left in the hands of IAs.It concluded that it was difficult to satisfy whether all amounts disbursed reached the rightful beneficiaries and whether loans were being repaid (Auditor General's Report, 1996).Therefore, the Auditor General's report not only uncovered serious financial irregularities but was also a clear testimony that the capacity for the implementing agencies was lacking.Therefore, the high administrative costs incurred did not positively correlate with the performance of ECS.Hence, attributing poor performance of ECS to lack of effective co-ordination and monitoring of its activities by the Secretariat staff in Kampala was not convincing.The main explanation was the poor design of ECS.
In contrast, PAP was coordinated at the national level by the Office of the Prime Minister (OPM).The two departments responsible for PAP activities within the OPM included the National Steering Committee (NSC) and the Income Generation Support Unit (IGSU).The structure also provided for a District Loans Committee (DLC) to scrutinize and approve credit disbursements.PAP's framework was later restructured for purposes of efficiency and effectiveness in credit delivery.The IGSU and DLC were scrapped leaving only the NSC and the Area Project Coordinators (APCs) at the national and district levels respectively.Hence, PAP's administration was decentralized to four zonal offices of Arua, Gulu, Luwero and Kabale.These programme coordinating offices were presided over by APCs while Area Project Officers (APOs) directly supervised the activities of IEs at the grassroots.To qualify, IEs must have been working with the community in the targeted areas for not less than three years and had to be acceptable to the local leadership.Therefore, the programme design of PAP drastically differed from that of ECS which was bloated and wasteful as illustrated in Table 2.
It is evident from Table 2 that PAP's structure helped it cut costs in the delivery of financial and non-financial resources to the poor which was not the case with ECS.
Whereas PAP had a simple and flexible loan access procedure where loan applicants either dealt with the APOs directly or with the IEs, applicants for ECS went through a long bureaucratic and rigid procedure that involved local councillors.Even in circumstances where technical officers were represented on the CSC, their role was only advisory.It was revealed that local councillors ignored their technical advice on various occasions: There is a lot of conniving among local politicians sitting on the CSC and for us we watch helplessly because if you interfere with their agenda they can make your work difficult.These councillors hold private meetings and agree in advance on how to support each other's loan  (1994)(1995)(1996)(1997)(1998).

Year
Cost applicants.They come to CSC when they already have predetermined positions.The CSC is just for window dressing purposes because loan applicants without Godfathers lose out.The CSC will continue to be unfair to some loan applicants unless politicians are removed from it. 7 Allowing local councillors to identify credit participants adversely affected the transparency of ECS, which was unable to get honest clients.Local politicians used their influence to access loans, recommended their relatives and friends, and their political supporters especially their campaign agents.Similarly, IAs took advantage of ECS' laxity to engage in fraud as was the case with Mpigi District where two IAs embezzled credit for Kyadondo and Gomba Counties.It would have been appropriate for Entandikwa loans to be managed by a competent body or qualified and experienced personnel.One that could undertake feasibility studies of projects submitted for consideration by loan applicants.This was ignored to the peril of ECS.Therefore, the poor design of ECS as evidenced by the incorporation of local politicians among others contributed to its ineffectiveness.
The elaborate programme design and mode of operation of PAP that freed it from the pitfalls of bureaucracy and political meddling was also due to the adoption of recognized "best practices" from successful micro-finance programmes such as the Grameen Bank, Bangladesh Rural Advancement Committee (BRAC), Bank Rakyat Indonesia (BRI) and Banco Solidario (BancoSol) of Bolivia (Mohammad, 1998;Montgomery and Hulme, 1996;Sharma and Zeller, 1999).The success of these microfinance institutions (MFIs) is attributed to the adoption of innovative credit delivery mechanisms.These include group co-guaranteeing, peer pressure and monitoring, effective supervision by credit officers, targeting of women, mandatory savings, monthly meetings, and loan cycles.In regard to loan cycles, PAP induced its clients with a five loan cycle system as a means of motivating them to make timely repayments.The first cycle involved a maximum loan of Ush.200, 000 (US$112) with an increment of Ush.200, 000 for each of the four subsequent cycles.Loan cycles were effective in encouraging thriftiness and motivating loan repayment among PAP clients.Access to repeat loans has been widely recognized as a strong incentive for repayment and many micro-credit schemes have adopted this innovation.The World Bank Development Report (1989) also revealed that borrowers acknowledge that their greatest incentive for strict repayment is the expectation of getting another loan.It is also argued that one of the innovations that made the Grameen Bank a successful micro-credit institution was the initiation of repeat loans (Stiglitz, 1990).
Moreover, PAP was able to build the capacity of IEs where it was lacking by enhancing their credit management capacities as well as those of the beneficiaries through training.Credit beneficiaries were given pre-loan training for eight weeks in which they were equipped with basic knowledge about group dynamics and management; budgeting for small projects; project management skills such as bookkeeping and record keeping; and micro-project appraisal techniques.It also involved more training of IE staff and provision of equipment such as office cabinets for record keeping, typewriters, and transport (motorcycles and bicycles) for the liaison staff.In contrast, ECS did not have a training component for the staff of IAs and neither did it undertake sensitization campaigns for its credit beneficiaries.It was apparently clear, therefore, that no effective innovations had been undertaken to improve the performance of ECS as was the case with PAP.Hence, these innovations enhanced the capacity of PAP beneficiaries to effectively utilize the loans.Training credit beneficiaries is particularly recognized as an important component that determines the success of the scheme.Berenbach and Guzman (1994) argue that training allows credit beneficiaries to improve their management and administrative techniques necessary for the success of their micro-enterprises.They observe that training is one of the integral elements that have contributed to the effectiveness of the solidarity group methodology in credit delivery world-wide.
In addition, PAP had an arrangement with government extension officers stationed at the district to assist those beneficiaries engaged in agriculture.Whereas clients appreciated these services, they complained about the cost of facilitating extension service officers with daily allowances.Moreover, there was concern among some PAP clients engaged in agriculture about inadequate marketing opportunities.They required assistance to access markets where they would be paid better prices to meet their loan obligations and also improve their welfare.Even though PAP provided market information to its clients, this largely favoured big producers of cash crops instead of food crops.This challenge brings in the role of the state in connecting farmers to markets through providing good infrastructure and information about markets.In the case of Uganda, the state lacked capacity as a result of many years of political and economic decline.Besides, the state's blind embrace of economic neo-liberalism and the disbanding of the co-operative movement negatively affected the agriculture sector.
Other important innovations pursued by PAP included extension of credit to women who constituted 60% of all clients.More women accessed PAP credit because of their perceived honesty and better utilization of credit.Moore and Schoombee (1995) also confirm the credibility of women in credit utilization.They attribute the Grameen Bank's preference of women customers to being the needy and reliable customers.Besides, PAP also gave credit to those people engaged in micro-economic projects for which they had prior experience as opposed supporting the terminally poor.It should however be noted that while the targeting of the economically active people led to positive results, such an approach discriminated against the absolutely poor.This practice contradicted the government rhetoric that statesponsored credit was driven by the motive to empower the poor and consequently reduce their poverty.The concern about distortions within state-sponsored credit schemes such as ECS are frequently echoed in other contexts.Basu (1997) and Robinson (1994) assert that experience shows that government-directed credit benefits local elites who could have obtained loans at commercial rates.A comparative experience with ECS can be found in Buckley's work on Malawi (Buckley, 1996).He notes that one of the weaknesses that contributed to the poor performance of the Smallholder Agricultural Credit Administration (SACA) 8 was that loans largely went to the better off or elite borrowers instead of the targeted poor.Despite the better lending terms, SACA's recovery rate stood at 43.5%.PAP also carried out a periodic review of its activities using external independent firms, after which the recommended changes were subsequently implemented.Therefore, PAP's "good practices" as manifested in its effective programme design positively impacted on its clients.This however seemed not to be the case with ECS whose poorly designed programme did not translate into similar benefits to its clients regardless of extending sizeable loans with better terms and conditions.

STATE-SPONSORED CREDIT SCHEMES AND POVERTY REDUCTION
Having looked at the linkages between programme design and organizational effectiveness, this paper Muhumuza 127 proceeds to assess the impact of state-sponsored credit on poverty reduction.In other words, it seeks to establish whether state-sponsored credit created opportunities for employment creation and asset accumulation and consequently reduced poverty among the beneficiaries.The first achievement of both PAP and ECS was to avail credit to the rural poor who would otherwise not access it from formal financial institutions.Available data from ECS and PAP illustrate that there was substantial credit coverage in terms of outreach, which refers to the number of beneficiaries, gender categories served and loan amounts disbursed among others (Mcguire and Conroy, 2000).For example, PAP credit had by August 1999 been accessed by 25,380 clients in 29 districts out of which 15,228 were women and 10,152 men.The total loan portfolio was Ush.11.5 billion.In addition, PAP was able to mobilize Ush 1,028,683,030 (US$1,028,683) worth of clients' savings (Republic of Uganda, 1998).Likewise, ECS covered the entire country and extended credit worth Ush.9, 308, 400, 000 (US$9.92 million), which was accessed by 39,200 people as illustrated in Table 3.
The above illustration shows that ECS was accessed by the youths, disabled and women as well as urban and rural areas.In terms of outreach, 30% of ECS beneficiaries were constituted by the youths and women while 60% of PAP's participants were women.Therefore, there was ample evidence that the two credit schemes generally reached out to many and different people and areas that were historically neglected by formal financial agencies.Accordingly, if outreach is used as a yardstick for good performance, then there was no doubt that ECS and PAP performed well.In this respect, the indicators for good performance include credit utilization, high repayment rates, the ability to sustain and even expand their income-generating projects without further borrowing, improved household welfare, acquisition of assets, as well as the enhancement of savings.
In stark contrast to ECS, the clients of PAP generally performed better.This was regardless of the fact that its terms were more stringent than those of ECS.An example was the charging of market interest rates.The argument that market interest rates are not necessarily bad for small-scale borrowers had for long been proved by the resilience of moneylenders despite frequent government interventions with subsidized credit in developing countries (Stiglitz, 1990).A comparative experience on the interest rates issue can be drawn from Indonesia where prior to 1984 the government used to give subsidized agricultural credit to farmers through BIMAS 9 programme administered by Bank Rakyat Indonesia (BRI).This approach to rural financial intermediation had to be abandoned because of high losses and disincentive to saving.However, with the government's change from subsidized to commercial interest rates and the encouragement of savings mobilization, the new approach reversed BRI's decline and generated over two-thirds of its total profits in a period of only five years (Robinson, 1994;Mosley, 1996).Another comparable experience can be drawn from Kenya's Juhudi Credit Scheme.Its adoption of a financial systems approach that emphasized market interest rates made the programme achieve costeffectiveness and improved its impact (Mutua, 1994).Therefore, problem of rural credit markets is usually not about highinterests rates but the unavailability or inadequacy of access to credit opportunities.
There was evidence that the performance of beneficiaries' income-generating projects had a significant relationship with loan utilization and gender composition.Other variables such as age, education level, and household size did not have any serious influence.Participants who accessed and successfully utilized state-sponsored credit were those that never diverted loans to projects other than those specified in their application.
In regard to the reasons for borrowing, all responses indicated that it was for investing in income-generating activities.However, when it came to loan use and the achievements realized, responses given were mixed up.These included investment, paying for services, and consumption.Those who indicated that they invested their loans in income-generating activities were 60% while 37% diverted their loans into services such as health and consumption, and 3% had their money confiscated by their spouses.It was therefore not possible to make a clear distinction between the purpose for borrowing and credit utilization.Neither was it easy to separately categorize traders and smallholder farmers because all those involved in trade were also engaged in agriculture.
There was ample evidence that many of the recipients of ECS lied in their loan applications.This is why the general response to the reason for borrowing was to invest in an income-generating activity and yet many did not.This was done for the sake of qualifying for credit.Hence, most of the ECS recipients (mostly men) used loans for a variety of activities other than those stated in their application.These varied from paying for household services such as education, health to consumption.The diversion of loans can be attributed to organizational laxity and the use of local politicians as part of the implementing machinery.In contrast, PAP loans were overwhelmingly invested in income-generating activities and bore direct benefits to clients than was the case with ECS.Another observation is that it was difficult to make a clear distinction on how credit was utilized.This is because some clients invested part of the credit while the other part was spent on paying for pressing household needs such as services and consumption.The failure to separate loan proceeds from household demands which was common to all households made it difficult to sustain their income-generating projects without further borrowing.There was also no clear separation of income or benefits from the loans from those of other household activities.The benefits were mixed up and the beneficiaries had a lot of difficulty in pointing out with precision the accruing benefits.
It is these kinds of problems that make it difficult to assess the real impact of credit in rural contexts where there are no records of transactions, benefits and losses or specialization.This phenomenon is characteristic of peasant households and their transactions as noted by Rhyne and Otero (1994).They argue that many micro-enterprises are not autonomous units but are part of larger family or household units.Hence, the cash associated with one micro-enterprise is frequently mingled with that of other household activities.They, therefore, contend that the financial needs of households, or at least of individual entrepreneurs, are often not separable from the financial needs of the enterprise themselves.
Besides, it was evident that more women compared to men beneficiaries adequately used their loans and had plenty of achievements to show for such loans.However, men could not generally identify benefits accruing to credit.They gave a lot of excuses and put blame on the small amounts given, the delays involved, as well as natural hazards (in the case of agriculture) for the poor performance of loans.It was therefore confirmed that better credit utilization from ECS and PAP depended on both gender and organizational vigilance.In both credit schemes, women restricted their credit to those very projects that were indicated in their loan applications.In the particular case of PAP, all women implemented their loans in approved projects while 75% of all women in ECS also invested their loans in approved projects.In addition to better credit utilization, women also generally performed better (69%) than men (35%) in repayment performance of both credit schemes.In the case of PAP, 83% of women repaid compared to 50% for men while in the case of ECS, women still performed better with 67% compared to 30% of men who repaid the loans.The 23% of women who defaulted either had no experience in that specific project or were affected by genuine factors such as natural hazards or lack of markets in the case of agriculture.Other circumstances that led women to default include business failure, sickness and confiscation of loans by spouses.
Nonetheless, it is significant to observe that there is a general reluctance for people to honour repayment of loans from the state.Even in circumstances where they paid well as was the case with PAP, there had to be a clear demarcation from politics as well as strong provisions for enforcement.According to Hoff and Stiglitz (1990), funds from the state are viewed as grants and therefore there is hesitation when it comes to repayment.This perception tends to afflict state-sponsored credit schemes elsewhere as illustrated by the case of Malawi Mudzi Fund (MMF).Buckley (1996) observes that organizationally, the MMF failed to distance itself from the government and many rural people viewed it as a government welfare-based agency.In an interview with one administrative officer in Mpigi district on why Entandikwa beneficiaries were hesitant to repay, his response was that: As long as it is government money, it cannot be taken seriously.If someone steals from government, it is not considered stealing.People perceive funds from the state as a kind of repatriation from what was rightly theirs but had been lost to the state or government.It is taking back your things.
The perception about financial obligations to the state seemed to confirm Ekeh's (1975) proposition about the two publics in post-colonial Africa.He observed that the civic public in Africa is amoral and lacks the generalized moral imperatives operative in the private realm and in the primordial public.He further noted that the unwritten law of the dialectical relationship between the two publics is that it is legitimate to rob the civic public in order to strengthen the primordial public.
However, the criterion of using repayment rates as a measure of the effectiveness of credit schemes is highly contested.According to Snow and Buss (2001), there is widespread tendency to measure the effectiveness of credit schemes using outreach and repayment rates.They further contend that although the "persuasive" assumption has been that micro-credit increases the wellbeing of the poor, very little has been done to determine the extent to which it actually does.Rahman (1999) shares the same concern especially when he argues that in recent years the most important criterion for the success of micro-credit schemes is determined by their ability to achieve financial sustainability with less regard to the beneficiaries' wellbeing.Therefore, high repayment rates may give a wrong impression that credit schemes are benefiting participants.While repayment performance is a commonly used criterion for evaluating the effectiveness of credit schemes, it is equally significant to look at the material contribution of credit schemes to the beneficiaries.Available evidence, for instance, shows that whereas the much heralded Grameen Bank has been able to maintain high repayment rates of over 95%, its clients did not necessarily fare well contrary to the impression created.Rahman (1999) has pointed out how the Grameen Bank's policy to ensure economic viability has had crippling effect on the beneficiaries especially women.He argues that this approach has inevitably increased the indebtedness of borrowers, frustration, and led to acts of violence.Therefore, it is necessary to go beyond the repayment rates and look at their impact on the welfare of the beneficiaries as a measure of the success of a credit scheme.
Whereas the good performance of women can be attributed to PAP's vigilance in terms of strict supervision, the fact that a similar situation transpired in ECS where there was organizational laxity and inadequate supervision proves that women are generally exemplary in their credit performance.Not only did women utilize their credit better but also their repayment rates were higher compared to men, and benefits from credit had a positive impact on their household welfare.This phenomenon lends credence to the argument that women are good credit performers than men (Pitt and Khandkar, 1998).According to Rahman (1999), women's good credit repayment record is attributed to having limited physical mobility because of their culturally patterned behaviour.In other words, women are easier to trace, are more disciplined, fear legal action and humiliation.
In the case of absence of a grace period, it was evident that some clients of PAP borrowed from other sources in order to repay loans.Rahman's (1999) research on the micro-credit programme of Grameen reported a similar practice.He noted that women were forced to borrow from other sources to meet weekly payments due to stringent loan terms.Therefore, loan recycling can be Muhumuza 129 attributed to the harsh loan terms especially where agricultural projects require a longer gestation period before the proceeds are realized.For example, PAP did not have a grace period because loan applicants were required to engage in on-going economic ventures.Whereas the available information created an impression that PAP emphasized agriculture by allocating it 60%, the reality on the ground was different.Field officers were strictly sticking to the terms and conditions of PAP, favoured trade whose returns were immediate and allowed speedy repayments.Even where agricultural loans were mentioned, it was in the area of marketing.
Although the conditions of PAP enabled a good repayment performance, they discouraged agriculture, which is the backbone to Uganda's economy as well as the main source of employment and income to rural people.Therefore, if the strategy of using credit to reduce poverty was to be effective, different loan terms would have to be devised for those engaged in agricultural activities.
The importance of participation of beneficiaries in development programmes intended to benefit them has for long been acknowledged (Ascroft and Masilela, 1994).The participation of stakeholders is frequently emphasized as a prerequisite for the success of interventions designed to improve the welfare of the poor (World Bank, 1994).However, the major shortcoming of ECS was failure to involve key local level stakeholders right from inception to implementation.This omission adversely affected the supervision of ECS activities.District officials were not vigilant because they were not involved at the level of programme design and hence felt that they did not own ECS.The inception of ECS had no grassroots input but only involved the top technocrats at the central government level.It evolved from a presidential promise, which was later formalized.Even local government officials revealed that the policy framework on credit as an interventionist strategy against poverty did not include their input: The government disregarded the districts when initiating ECS and yet it was the districts that were supposed to oversee its implementation.The management of Entandikwa loans should have been fully decentralized to the districts because they are the ones accountable to the people and not the distant bureaucrats based in Kampala.This is why there is no serious follow up and the programme hasn't had a serious impact.As far as we are concerned at the local level, this programme is not owned by us.We have no power over it and therefore cannot have influence over its implementation.So, why be responsible for something we do not own? 10  Therefore, there is no doubt that state intervention with credit to reduce poverty did not involve participation of key stakeholders.It was instead an outcome of a topdown initiative.This vividly explains why its performance was dismal.
In contrast with ECS, PAP's approach allowed a limited degree of participation in its operations.It was a condition that applicants first organize into solidarity groups according to the prescriptions of PAP before they could access credit.This was irrespective of whether those credit applicants were already organized in self-help associations or not.In case of existence of such grassroots associations, they were restructured according to PAP's methodology before members could access credit.In regard to the participation of beneficiaries, PAP allowed them to organize monthly meetings to discuss their problems.These meetings were held between solidarity groups and IEs or sometimes between solidarity groups and APOs.Also, the beneficiaries of PAP sometimes made suggestions, which were accepted by PAP officials.An example was the suggestion that group sizes be enlarged from 7-15 to 15-30 members so as to lower individual contributions as well as group administrative costs.Another indicator of participation was attending to clients' complaints and sometimes rescheduling loans.While to a certain extent participation was allowed, it did not have a significant influence on the terms and conditions of PAP loans.
Although the participatory approach used by PAP was to some extent limited, it some how led to positive programme outcomes.The group mechanism to access credit enhanced the formation of social capital among the clients of PAP credit.Cases of improved relations and cooperation between members were cited.It was for instance reported that group members advised and sometimes assisted each other in various ways.There is truth in the argument that participation is positively correlated with improved performance as demonstrated by PAP's impressive record payment performance of 93% and improved welfare of its clients compared to ECS, which performed poorly with a repayment rate of 40% and did not positively impact on the welfare of its beneficiaries.Therefore, the failure of state-sponsored credit schemes to involve stakeholders such as the district leadership in the initiation and design partly explains why ECS' performance was futile.It is now recognized in development circles that the involvement of stakeholders in identifying the problem, designing the strategy, implementing, monitoring, and evaluating the programme outcome leads to success (Kleemeier, 2000).Participation is therefore positively correlated with the success of development programmes (Isham et al., 1995).

POLICY IMPLICATIONS FOR STATE CREDIT SCHEMES: A CONCLUSION
The different outcomes exhibited by the two statesponsored credit schemes show that the programme design is instrumental to performance.PAP had adopted various organizational innovations that were responsible for successful credit institutions such as the Grameen Bank.The other reason for enhanced performance of PAP was attributed to strict monitoring and supervision by IEs and credit officers to guard against diversion, which was reinforced by an inducement component in form of repeat loans.In addition, PAP clients benefited from pre-loan training in group dynamics, enterprise management, and book keeping.They were also provided with the services of an agricultural extension worker as well as market information.These organizational innovations enhanced the institutional performance of PAP to the extent that it was able to achieve a high credit repayment rate averaging 93%.On the other hand, ECS which gave big loans that ranged from Ush 100,000 to 1,500,000 and charged a small interest rate had recorded dismal performance.Not only was ECS unable to recover the loans, but it also had no tangible impact on clients' welfare.The Permanent Secretary for MGLSD conceded the crisis facing ECS when he appeared before the Parliamentary Public Accounts Committee.He revealed that government required Ush.766 million (US$383,000) to recover outstanding loans and to produce a status report.He also told the committee that his Ministry had no records to compile an up-to-date status report on the scheme (The New Vision, June 5, 2002).
Apart from organizational success, PAP also generated material and social benefits to its beneficiaries.It extended investment capital to those income groups that would not have qualified to access traditional financial institutions.There was also evidence that PAP credit had positively contributed to the beneficiaries' welfare.Besides, PAP's credit delivery mechanisms, which required applicants to organize into solidarity groups, enhanced the formation of social networks.Some of the accruing social benefits accruing to groups included increased co-operation between members, free advice and counselling, material assistance to one another, and sharing of information.The enhancement of social networks contributed to the good economic performance of PAP beneficiaries.While both material and social benefits were associated with PAP, the reverse was the case with ECS, which was a parallel state-sponsored credit scheme.Poor programme design, organizational laxity, weak credit delivery mechanisms, patronage tendencies, lack of or inadequate supervision, and loan recovery problems bedeviled the ECS.Likewise, the performance of the clients of ECS was poor.No serious material and social impact could be associated with Entandikwa credit.This was partly because Entandikwa loans were diverted to consumption rather than investment in income-generating ventures.
The different outcomes of the two credit schemes can largely be explained by two factors.First, the superior programme design and mode of operation of PAP compared to that of ECS was largely responsible for the variation in performance.Second, the operational autonomy given to PAP enabled it to undertake useful innovations while ECS was derailed by political interference.The issue of separating the credit scheme from politics also had a significant influence on the good performance of PAP compared to ECS.Whereas PAP was protected from political interference, ECS was not.Politicians discouraged repayment of ECS for purposes of cheap popularity.Both local and national politicians promised to write off the loans once they were elected during local and national election campaigns.A similar situation pertained to northern Uganda where area politicians discouraged people from repaying ECS loans because the government had failed to end insurgency in the area.Even President Museveni himself was alleged to have warned those enforcing repayment of Entandikwa during a pre-election rally (2001) in Wakiso not to harass his people (beneficiaries) since they would pay anyway.Moreover, the nation-wide local council elections of 1998 saw many members of the old committees being voted out and new ones coming, a change that affected the continuity of CSCs.Such phenomenon reaffirms Braverman and Guasch's (1986) argument that subsidized credit tends to be driven more by political considerations than development motives (Hoff and Stiglitz, 1990).Therefore, defaulting on Entandikwa loans can to some extent be attributed to the politics of patronage, which PAP was able to overcome because of the well defined enforcement mechanisms.Besides, its credit activities were contracted to experienced NGOs which helped to distance them from politics.The degree of autonomy given to PAP can also be owed to the positive role played by ADB which contributed 90% of its funding compared to ECS, which was fully funded by government.The involvement of ADB helped to inculcate good practices and consequently enhanced the capacity of PAP by building efficient management systems and enforcing strict financial controls.
The issues arising from the analysis of the two state credit schemes have two policy implications.First, state intervention in credit intermediation is still relevant and necessary because of continued neglect by formal financial institutions.Moreover, credit is increasingly being recognized as a strategy against poverty (World Bank, 2000).Credit is believed to empower poor people to participate in the economic growth of their countries through employment generation and income creation that are essential for the improvement of the welfare of the poor.State-sponsored credit schemes are therefore necessary to fill this gap so that the rural areas in general and the rural poor in particular can be activated to participate in the economy.Secondly, state-sponsored credit schemes can also be effective in reducing poverty provided their programme designs are well crafted and they are given operational autonomy as the case of PAP demonstrates.The experience with credit programmes managed by the private sector and NGOs has shown that these actors tend to have capacity limitations especially in outreach terms and their loan terms are stringent and exploitative.Besides, credit schemes targeted at poverty reduction need to be financially subsidized so as to give affordable credit to the absolute poor.This is necessary to avoid a situation where credit institutions are forced to charge high interest rates in order to ensure sustainability of their activities.This may have negative effects especially if such credit is targeted at reducing the poverty among the absolute poor, which may not be the case with the active poor.Hence, there is urgent need for policy makers to extend support to credit schemes aimed at helping the poor while those not associated with antipoverty policy initiatives can be left to charge market interest rates.Therefore, the state should, in addition to creating a supportive policy environment, directly intervene with state credit especially to those areas and sections of the population that are less likely to attract Muhumuza 131 non-state actors such as NGOs and the private sector.Alternatively, the state can also extend credit to independent, member-owned and managed savings and credit associations (SACAs) at the grassroots.This is important for sustainability purposes.In order for antipoverty strategies such as credit to be effective and sustainable they must be owned by the poor people through their grassroots organizations.In other words, state-sponsored micro-credit should be channelled through informal self-help associations (ISAs) so as to avoid problems of asymmetrical information and imperfect enforcement that frequently plague rural credit markets (Hoff and Stiglitz, 1990).Besides, Bolnick (1992) argues that using informal credit agencies as agents of traditional financial institutions can reap a lot of benefits.These include taking advantage of their low transaction costs, great flexibility, astute to knowledge of local conditions, and recourse to social sanctions.There is also ample evidence that ISAs promote grassroots participation in credit activities and subsequently improve their management.By allowing a high degree of beneficiary participation in the management of credit schemes is believed to be associated with better performance (Isham et al., 1995).
Accordingly, the state should not abdicate its key responsibility of combating extreme forms of poverty among its subjects.Besides, the state's anti-poverty strategies such as micro-credit schemes should be tailored to the different realities in order to have a meaningful impact.In the context of Uganda, the state not only needs to intervene to expand credit access and undercut the stringent credit conditions that characterize the non-state players but also give affordable credit to the very poor engaged in agricultural activities that are frequently discriminated against by the private sector.This is because the absolute poor people in Uganda mainly reside in rural areas where they mainly engage in off-farm activities such as agriculture.Therefore, for credit to be a major anti-poverty strategy then it must support agriculture in addition to other economic projects.

Table 1 .
Administrative costs of ECS as of July 1997 (Ush.million).