Finance and formalisation as mechanisms for poverty reduction in Africa

If you ask any informal enterprise in Africa what constraints they face, then the overwhelming majority will answer that they need credit. If you ask any finance organisation why it does not finance such businesses it will answer that they lack records or any legally enforceable collateral. Faced with such findings the response of development agencies has been twofold. The first is to promote institutions that provide such funds to small-scale businesses or, in the case of microcredit, to individuals who are self-employed. The second is to promote the formalisation of businesses. Microfinance programmes are seen as one of the most promising means for relaxing these credit constraints. Not only do they put money directly in the hands of the poor, but they also build on local knowledge and social networks to target and monitor its distribution. In parallel there have been interventions that have sought to make businesses in some sense more formal. The premise here is that it will then be possible for the business to raise funds and expand faster than was the case as an informal enterprise.

In this project we are addressing both these questions for Ghana and Tanzania. Is it true that firms are credit-constrained, and do they grow faster if they are more formal? The question is important as small-scale and informal businesses have been expanding in both countries. In Ghana we have particularly clear evidence of what might be termed a trend towards informality.

Table 1            The distribution of manufacturing firms in Ghana
Table 1	The distribution of manufacturing firms in Ghana

Table 1 shows the results of an analysis of two censuses carried out by the CSAE with the collaboration of the Ghana Statistical Office. Over this 16-year period the number of firms in Ghana’s manufacturing sector increased three times from, in round numbers, 8000 to 26,000. All this increase was among the relatively small firms. Microfirms, i.e. those with less than five employees increased five times, whereas the number of firms with more than 100 employees remained practically unchanged. A DPhil thesis completed at the CSAE in 2008 (Sandefur 2008) proves an analysis of what might be termed the growth of the microfirm to dominate Ghana’s manufacturing sector. Although the growth of microfirms has been dramatic, another implication derived from Table 1 is for employment. Although the number of firms has increased three times, employment has only grown by 55 per cent. This growth process produces firms not jobs. Why this might be is discussed in Teal 2008. Does it matter?

Figure 1 Trends in employment in Africa

Figure 1 Trends in employment in Africa

Figure 1, taken from Kingdon, Sandefur and Teal (2006), provides evidence for a trend in employment in Africa. There is a big divide within sub-Saharan Africa between countries like Ghana, Tanzania and Uganda, which have seen an explosion in the growth of self-employment, and urban Ethiopia and South Africa, which have seen a rise in unemployment. In all cases there appears to have been little, if any, growth in wage jobs. Does that matter? 

Figure 2 Earnings in self employment, small firms and large firms


Figure 2 Earnings in self employment, small firms and large firms - Ghana Levels] Figure 2 Earnings in self employment, small firms and large firms - Tanzania Levels

The answer implicit in Figure 2, taken from Sandefur, Serneels and Teal (2006) is possibly yes. Earnings do differ quite a lot across the small- and large-scale sector and research has shown this is not due to the differing skills across the sectors.

The research project is extending existing urban panel labour market datasets in Ghana and Tanzania, on which Figure 2 is based, by carrying out a re-survey that will build a five-year panel of individuals for these countries. The objective is to use these panel datasets to investigate two interventions related to finance and formalisation in determining access to credit in Ghana and Tanzania. We have sought to design the data so that they can be comparative across the two countries, Ghana and Tanzania, and links with earlier surveys, so that a longer run panel dataset can be built. With this in mind we have built on a survey conducted by the CSAE in 2004 and 2005 of urban workers in both Tanzania and Ghana. The fieldwork was undertaken from July to the end of 2008.

To address the role of credit we are working in association with an NGO in Ghana, which is currently running microfinance programmes targeted at the productive poor in the low-income bracket of the population who are engaged in microenterprises and income-generation activities. These are mostly women (75%) who do not have access to credit from formal financial institutions.
During the 2006 wave of data collection we identified 48 clusters from our survey sample (approximately 10 households each), which were located in the regions where the NGO was working (Accra, Tema and Kumasi). These clusters were randomly divided into ‘treatment’ and ‘control’ groups.
As a means of evaluating its programme activities, the NGO has agreed to form new microcredit groups in each of the 24 randomly selected treatment areas. The NGO credit officers accompanied the survey enumerators to the respondent households and all qualifying survey respondents in these clusters (self employed and over 18 years) were personally invited to join the programme. Those who chose to join are currently going through the NGO’s standard loan-screening process and, if approved, will receive loans between US$50 and $200 for an initial term of four months at a 4 per cent nominal monthly interest rate. Pending approval of this project, we plan to return to all of the households in the survey in both the treatment and control groups in the summer of 2007. In the context of this randomized design, a differences-in-differences approach – testing whether the mean changes in, say, income observed within the two groups are significantly different – provides a simple way to establish the causal impact of receiving microcredit and overcomes many of the biases that confound traditional programme evaluations.
To address the role of formalisation we propose to link our data collection to an evaluation of a programme of formalisation that is currently under way in Tanzania. We hope to be in a position to study the causal impacts of changing financial flows and formalisation on such economic outcomes as consumption and poverty levels, vulnerability to income shocks and the insurance role of such flows and investment in small- and medium-scale enterprises (SMEs).
This research initiative will have implications for a broad range of policy questions in both developing and donor countries. We hope to inform the following policy-relevant questions:

  • What degree and type of development impact can policymakers hope to achieve by funding microfinance projects? We propose to quantify causal impacts not only on income and employment, but also on other dimensions such as insurance, health outcomes, education enrolment and fertility rates.
  • How can microfinance projects be designed to maximize impact? We consider the effect of project features such as group versus individual lending, loan size, and collateral requirements on outreach and sustainability.
  • Will dramatically scaling up microcredit services in Africa diminish their current sustainability and impact? Here, we investigate project effectiveness across different types of borrowers and lenders to assess the potential for expansion.

The answers to the questions outlined above will have a broad audience, from academics to policymakers to the NGO community. We intend to provide rigorous microeconomic evidence to answer the question: how can increased financial flows most effectively generate pro-poor growth?

In addition to obtaining ESRC funding the CSAE has entered into a collaboration with the University of the Witwatersrand in South Africa, funded by the IDRC in Canada, to participate in a programme of work entitled: ‘Improving labour market outcomes for the poor in sub-Saharan Africa’. This programme runs until June 2009. Funding from this project is being used as part of our engagement strategy and the links within this project enable us to work not only with other researchers in Ghana and Tanzania but also to link to the other projects included in this wider study. These include both Madagascar and South Africa.


  • Kingdon, G., J. Sandefur and F. Teal, ‘Labour market flexibility, wages and incomes in sub-Saharan Africa in the 1990s’, African Development Review, vol. 18 (3), pp. 392–427, 2006.
  • Sandefur, J., ‘Explaining the trend toward informal employment in Africa: Evidence from Ghanaian manufacturing’, paper presented at the IZA/World Bank Conference: Employment and Development, Berlin 25–27 May 2006, CSAE mimeo, 2006.
  • Sandefur, J., Essays on Labour and Credit Markets in Africa, DPhil, Oxford, 2008.
  • Sandefur, J., P. Serneels and F. Teal, ‘African poverty through the lens of labor economics: Earnings and mobility in three countries’, CSAE/GPRG Working paper,, 2006.
  • Teal, F., ‘Using firm surveys for policy analysis in low income countries’, Journal of Development Perspectives, vol. 4 (1), 2008.