Business Development in Four Developing Countries: Institutional Barriers and Supports*

by

Paul G. Hare

* Paper prepared for the Development Studies Association Annual Conference, to be held at Reading University, November 11th 2006. The paper has been written as part of the DFID-funded Research Programme Consortium on Improving Institutions for Pro-Poor Growth (IPPG Programme, for short). The author is grateful to DFID for the financial support, and to colleagues within the Consortium for helpful advice in the course of preparing this paper. However, the author bears full responsibility for remaining errors.

Contact Information: School of Management and Languages Heriot-Watt University Riccarton Edinburgh EH14 4AS

DRAFT September 2006

E-mail: [email protected] Website: www.sml.hw.ac.uk/ecopgh IPPG Programme website: www.lse.ac.uk/collections/IPPG (due to move to a new site in October or November 2006).

DSA Annual Conference, November 11th 2006

Theme: 3. Business, finance and poverty reduction

Business Development in Four Developing Countries: Institutional Barriers and Supports Abstract Drawing on the findings from four country studies – covering Mali, Tanzania, Bangladesh and Bolivia – as well as lessons and ideas from extensive literature surveys carried out for the IPPG Research Programme, this paper will focus on business development issues and the role of institutions. The World Bank and other bodies already carry out regular surveys of the business environment in countries around the world, and these surveys commonly find that in many developing countries the business environment is deficient according to various qualitative indicators. Sometimes, as a result of such surveys, countries are advised to reform this or that aspect of their policy as it affects business conditions, but in practice the advice is either not taken, is poorly implemented, or fails for some other reason. Our view is that successes, where they occur, as well as the frequent failures, have a lot to do with aspects of the economic and political institutional environment that are often not picked up in general surveys. We propose to illustrate this through the four country studies referred to above, showing why, in some cases, apparently good policies fail, and why, unexpectedly, apparently less good ones succeed. It is only quite recently that the institutional conditions for successful development have started to receive the attention they deserve, and our initial findings not only offer some immediate lessons, but also suggest important directions for further research.

2

Business Development in Four Developing Countries: Institutional Barriers and Supports 1. Introduction The purpose of this paper is to report on some of the initial findings of the IPPG Research Programme. These findings have been put together in the form of eight papers written earlier in 2006, four of these being literature reviews, four being individual country studies. These papers are IPPG Working Papers 1-4, and IPPG Discussion Papers 1-4, all of which are fully cited in the references at the end of this paper. Six papers are already on the IPPG website, and the remaining two will be completed very shortly. In sum, they amount to over 250 pages of text, so the present paper will inevitably be a highly selective and condensed account of the first stages of our research. The IPPG Research Programme is based upon the following: We hypothesise that Pro Poor Growth (PPG) depends critically on the interactions of formal and informal political, social, and cultural institutions with economic institutions. Together, these interactions constitute an institutional matrix which may either enhance or constrain PPG. This programme seeks to identify, understand and promote institutional interactions which support the achievement of PPG. Given the central objective of this research programme to explain the circumstances and institutional settings most favourable to PPG, the overarching questions to be addressed are these: ‚

How are the institutions that affect economic growth and its distribution established, sustained and changed?

‚

What determines their effective functioning? How is this related to the social, cultural and political matrix from which they arise and in which they operate?

‚

How do institutional interactions influence economic growth, the pattern of growth and, specifically, the possibilities for pro-poor growth?

We interpret PPG broadly, to mean economic growth that enhances the capabilities of poorer people, which may be achieved both through the ways in which growth is brought about and also through more equitable distribution of the benefits of growth. Whether formal or informal, public or private, institutions are relatively stable social arrangements (embodying rules, norms or conventions) possessing a number of special features: ‚

They regulate behaviour in ways which, in the short run, often conflict with individual preferences;

‚

They are based on shared expectations and meanings, derived from custom and legal provisions that establish trust;

‚

They are best thought of as ‘repeated games’ in which most types of interaction occur many times . 3

Given these features, many institutions are likely to have the character of public goods. This implies that the ‘supply of institutions’ generated by the market mechanism left to itself is unlikely to correspond with social efficiency. By ‘supply’ in this context, we mean both the question whether a given institution or type of institution would be created at all by the market, and more detailed issues such as the scale and coverage of the institution concerned. For instance, wealthy people might well provide themselves with services to protect their property rights, but such services might not extend to the poor (indeed, in the worst case, the property rights of the poor might be neither recognised nor protected). There is evidently a potential role for the state both in creating institutions which the market does not provide and in regulating in the public interest those which it does. On the other hand, in considering the state we should avoid the error of supposing it to be an inherently beneficent agent, in some sense external to ‘the economy’ per se. Whatever specific country we care to examine, this naive and simple approach does not prove to be especially helpful. What turns out to be critical for successful growth is a state that is effective and competent within its spheres of authority, but whose discretion and power are subject to effective institutional constraints. In the present paper, we focus on a particular aspect of institutional development and its role in fostering economic growth in general, pro-poor growth in particular, namely institutions that facilitate business formation and development. In Section 2, we summarise what our literature reviews have to say on this subject, while Section 3 sketches what we learn about business development from our four country studies. Section 4 links the country discussions with the earlier hypotheses, and concludes with some pointers to further research.

2. Business Development - Literature Reviews From the politics literature review (Leftwich, 2006), three major research areas were highlighted, concerning: (a) the political ‘demand’ for PPG; (b) the structure and politics of economic decision-making and institutional design; and (c) general issues to do with ideas, interests and institutions. The second of these is most relevant to our theme of business development, since to understand the process well we need to know about the formal structures of state power (both the executive and the legislature), their relationships with the bureaucracy and the judiciary, and the details of the policy environment within which business operates. Likewise, it is important to understand how economic interests - from business, union, agrarian and other relevant organisations - are mediated with and influence the political processes. Likewise, implementing policies that affect business is partly about setting the framework and ‘rules of the game’, partly a question of incentives, either to cheat or to comply. Thinking along these lines, it can sometimes make sense to design simple policies that might be, in some sense, technically inefficient, if they are nevertheless less prone to corruption, cheaper and easier to administer, and hence more ‘incentive compatible.’ It would be a mistake, though, to imagine that we can ever sensibly design such policies by assuming we are starting from a tabula rasa. As Harriss (2006) argues, institutions in any society are necessarily highly path-dependent, implying that anything new builds on the past, one way or another. This simple but important observation also explains why policies and organisations that seem to work well in one place might fail elsewhere, simply because the history is different, along with the underlying cultural and social norms. The same paper also emphasises that most 4

developing countries can be characterised as ‘neo-patrimonial’ in the sense that much authority is highly personalised rather than rule-based and bureaucratic; indeed a transition to the latter type of modern state might be expected to accompany development, though one should not assume that one is a pre-condition for the other. For business, of course, this raises extremely tricky questions since the presumed need for a fairly competitive and open ‘playing field’ for business might come into conflict with the more personalised state-business relations that are more often found in the patrimonial state. This possible conflict also lies at the heart of some of the hypotheses about business put forward in Hare and Davis (2006). For the sake of brevity, we refer here to just three business-related hypotheses, though some others in the paper also have a bearing on our topic. The three hypotheses are: ‚

‚

‚

The costs and complexity of doing business are frequently too high, and discourage much potentially beneficial economic activity. The institutional arrangements that result in such high costs are therefore damaging to economic performance, but their persistence depends on aspects of the political-economic structure that are hard to change (and in any case, slow to change). The lack of modern institutions to manage risk results in firms that remain too small (and often informal), farms that are over-diversified and technically inefficient. These conditions often tend to result in low incomes, low rates of income growth. Whether privatization proves to be beneficial for the economy or not depends on postprivatization ownership structures (e.g. it is not usually particularly desirable for incumbent elites to end up as major shareholders) and on the quality of the regulatory and competitive environment in which the privatized firms then find themselves.

All three hypotheses point to aspects of the development process that are frequently neglected or under-estimated, which is partly why policy advice is often too similar between countries, and why it so often fails. Finally, what does the available empirical evidence tell us about the connections between institutional quality and economic growth, or more specifically, PPG? Sen et al. (2006) examines this topic very thoroughly, reviewing both macro-level multi-country studies and some more sectoral, microeconomic research. The available macroeconometric studies do appear to find a significant role for institutions as a factor explaining economic growth, though even the most carefully conducted recent studies suffer from some important weaknesses, so the debate is far from over. Thus even the direction of causality (i.e. from institutions to growth versus from growth to institutions) cannot be regarded as settled. Also, the institutional variables most often used are necessarily rather broad and general, so they do not allow much to be said about the very specific topic of this paper. Finally, there is scarcely any work done yet on the inter-relations between institutional quality (somehow measured) and pro-poor growth. So, we have a sense that institutions are probably important for growth, but the empirical evidence does not (yet?) allow us to feel great confidence in that conclusion. Accordingly, let us now turn to the four country studies to see whether they can help us to reach firmer conclusions.

5

3. Business Development - Country Studies 3.1 Bolivia Uniquely in Latin America, per capita income in Bolivia is virtually the same as it was back in 1950, whereas in every other country incomes have at least doubled (see Wiggins et al., 2006). With the wealthier parts of the economy based on large production units - hacienda-style agriculture operating under near-feudal conditions, and large firms monopolising much of the resource extraction - institutional development relevant for business has proceeded slowly, with many interruptions along the way. For the most part, the large production units have been able to survive and even flourish perfectly well in an under-developed institutional environment, simply by taking advantage of their relatively easy access to political élites. Such personalised support, however, is of no use to the numerous smaller firms that need more secure property rights, protection of business contracts, flexible conditions for entry to and exit from markets, and so on. In this sense, Bolivia serves as a negative example. For on the one hand, the élite groups were able to use their political power to ensure their own economic success, while the general failure of the economy to grow over such a long period illustrates the damaging impact of bad institutions. Moreover, the World Bank’s survey of world business conditions, Doing Business 2007 (published late 2006), shows that from an already poor ranking, Bolivia had even slipped down the rankings a little in 2005. Thus there is not yet much evidence of real progress in the business environment. 3.2 Bangladesh Our country study of Bangladesh, Chatterjee et al. (2006), focussed on two specific business sectors in order to illustrate both their institutional conditions and their likely impact on pro-poor growth. These were the ready-made garment (RMG) sector, and the poultry sector. Regarding the RMG sector, none of the interviewed factory owners cited any problems regarding the functioning of the supply-chain, or with the labour force. Moreover, Bangladesh has survived the initial stage of the new and emerging international order of trade in textiles and clothing, following the ending of the Multi-Fibre Agreement. The RMG sector of Bangladesh is a success story that has made two notable positive contributions. First, the sector has created a set of entrepreneurs who have established and run their businesses despite facing a number of constraints. Secondly, an alternative source of employment has been provided, which contributed to more pro-poor labour market dynamics in Bangladesh. However, the model under which this sector is operating is exclusively ‘buyer-driven’ (with a clear manifestation of patron-client relationship) – it has evolved in a manner such that RMG entrepreneurs follow the needs of the buyers. This model provides little or no incentive for diversifying products as well as markets, and there is similarly little incentive for developing the skills of the workers to enhance their productivity. Hence the key to pro-poor growth of the RMG sector lies in developing policies to improve access to skills and earnings by the poor garment workers (both the existing labour force, as well as those who are entering this labour market). An assessment of the poultry sector, especially in terms of its effect on pro-poor growth, is necessarily more ambiguous than our findings for the RMG sector. On the one hand, poultry 6

output has grown, and there is evidence that poultry farms are profitable, though profits per bird are significantly higher on commercial than on household farms. On the other hand, grain used for poultry feed might need to be imported, unless domestic production of maize can increase rapidly. Other key inputs are also imported, often at prices that are out of reach for the very poor. The poor producers operating household farms have limited access to credit, for instance. Also, weak and inconsistent regulation of the sector leads to problems with adulterated or even toxic ingredients in feed; chicks are not properly screened for health and vitality so mortality rates are high; and the disposal of waste from poultry farms in open spaces is environmentally very harmful. The poor mainly participate in the sector through agents/middlemen. However, the kind of agent involvement in the sector is often counter-productive. Due to a lack of formal contracts with agents, there is no scope of legal redress for poor and ample opportunity for the agents/middlemen to exploit producers. This suggests there might be scope for effective government and/or private sector regulation of the agents/middlemen, possibly even with agents replaced by cooperatives at village level. Such associations and cooperatives might be helpful in building social capital and networks, in lobbying for better terms and market rights, or in overcoming diseconomies of scale and agglomeration. Thus in both the studied sectors there is a clear need for reforms to improve conditions for the poor, and hence to strengthen the contributions of these sectors to pro-poor growth. However, finding how to insert the needed reforms into the political process and then embedding them into each sector’s practice requires further research. 3.3 Mali The Mali study focussed on two meso-level institutions, namely CMDT (Compagnie Malienne pour le Développement des Fibres Textiles) and ODN (Office du Niger). Both are currently public sector companies (though there is talk of possible privatisation, encouraged by the World Bank among others). The former works with numerous mostly small farms concerned with cotton growing, the latter is concerned with farms that grow rice, mostly on irrigated land. These are key economic activities in Mali, and vital for pro-poor growth in the country. Based on these studies of CMDT and ODN, our Mali paper’s core theme was ‘farms within firms’ (Olukoshi et al., 2006). CMDT is a vertically integrated research-production-marketing structure, still linked to (and partly owned by) the French organisation from which it originated in 1974. The system has been considered a success (especially as compared to poor performance of the cotton sector in Anglophone Africa) for many reasons, including the use of local extension agents to promote technology diffusion and the spread of best practice; and the use of pricing schedules that rewarded farmers for sorting their cotton by quality. The extension agents also provided credit from CMDT to farmers over the growing season. The CMDT system, working through a network of AVs (Associations Villageoises), has proved effective in providing access to modern inputs and research on cotton, and has served to mitigate the risks borne by farmers. But these gains come at a price, namely complaints from farmers that they were not receiving a high enough price for their cotton, a tendency for the CMDT central organisation to grow too fast and hence inflate overhead costs (as with parastatals elsewhere in Africa), and complaints that CMDT surpluses have not been applied to develop infrastructure in the cotton-growing areas. The way forward for CMDT remains highly controversial, politically very sensitive in Mali. 7

While the CMDT system has proved problematic enough, it is fairly straightforward compared to the complexities and controversies surrounding ODN. In principle, ODN seeks to do for the farmers in the irrigated rice-growing areas of the Niger Delta what CMDT does for cotton. The practice, however, has been rather different, and has changed a good deal over time. No longer a vertically integrated monopoly in its areas of operation, ODN maintains and develops the irrigation system and supplies water to farmers, but the latter (individually, or through their local AVs) no longer have to sell their rice to ODN for threshing, husking and marketing. Also, rice production is insufficient to meet all the needs of farming families, so they have had to diversify into other crops. Property rights in land have not been as clear as they might be, with ODN reluctant to cede individual farmers full rights. Experiments in leasing land to foreign companies to undertake agri-business have not proved successful (e.g. a Chinese project where, in the end, the Chinese just sub-let the land to Malian farmers), and in the ODN area many of the AVs are considered highly corrupt. Overall, then, the ODN is far from being a clear success story. So CMDT and ODN present some interesting contrasts. Both face difficulties and an uncertain future, but the former starts with a record of considerable success, the latter with a far more mixed record. What remains hard to discern is the precise institutional differences that account for this. 3.4 Tanzania Some decades after Independence, following a socialist period and various attempts at more market-oriented reform, Tanzania remains a country still struggling to find an effective development path. It has been growing in recent years, but not especially rapidly, and both agriculture and manufacturing have lagged behind overall GDP growth; also, the conditions for doing business are not particularly good, as our country paper explains in some detail (see Ajakaiye et al., 2006). Some measures of institutional quality that are particularly relevant to firms’ operations and growth were drawn from the World Bank’s Investment Climate Survey 2004. Comparing Tanzania with its East African neighbours, Kenya and Uganda, and the Sub-Saharan Africa average, we find that firms viewed access to electricity, tax administration, business licensing and permits and corruption as greater constraints on their operations and growth in Tanzania than the SSA average. Most interestingly, business licensing and permits were particularly viewed as a severe constraint by firms in Tanzania (27.4%), relative to firms in Kenya (15.2%), Uganda (10.1%) and the SSA average (13.7%). This finding is in line with others to do with Tanzania’s poor regulatory quality. Thus, the specific institutions that matter most for firms in their investment activities are particularly weak in the Tanzanian context. The paper ends by identifying a number of puzzles of the Tanzanian growth experience, formulated as hypotheses to guide our subsequent research. For the sake of brevity, only two are noted here. First, why has there not been a more labour-intensive growth strategy based on smallholder agricultural production and the growth of small and medium enterprises - especially the agro-processing industries with their strong backward and forward linkages - given Tanzania’s geographical advantages, and the recent institutional reforms? In other words, which institutions are blocking this process and what is the motivation for their reproduction? Second, and perhaps more to do with political than with economic institutions, in what way does 8

the country’s high aid dependence combined with a small and powerful executive impact on propoor growth? In other words, could the rules, patterns and conditions surrounding international aid and development assistance be constraining PPG? This could occur, for instance, if it suited elites to appear ‘needy’ in relation to foreign assistance, since they could then siphon off part of the foreign inflows for personal benefits; at the same time, to satisfy domestic constituencies they might have to ‘allow’ some PPG, but not too much since this would demand more radical institutional change.

4. Conclusions - Ideas for Further Research Much research is still to be done on all our countries, plus some others currently being added to the IPPG Programme (such as Peru, Ghana, West Bengal, possibly others in due course), but we can already pull together a few conclusions, albeit necessarily rather tentatively at this stage. From the hypotheses about business referred to in Section 2 we have not, in our first round of papers, had a great deal to say about privatisation, but on the other two, all of the country studies shed some light. First Hypothesis - Costs and Complexity of Doing Business The finding that doing business is difficult and costly came up in all four of our initial country studies. The nature of the institutional blockages varied somewhat across countries, but the effects were fairly similar across four very diverse countries. The informal sector, of course, helps to fill many of the resulting ‘gaps’, and that is good for jobs and family incomes, probably propoor on balance; on the other hand a large informal sector is bad for the public finances, since it is by definition untaxed; and it is likely to be bad in terms of employment protection and other aspects of social protection. The interesting research question, then, is to understand the interests that come together to make institutional reforms in this area of the business environment apparently so difficult. Second Hypothesis - Lack of Institutions to Manage Risk In this case, it was certainly found that the lack of such institutions contributed to firms and farms remaining small, often over-diversified, and so on. However, our studies did reveal a number of possible solutions, more or less successful. Thus the meso-level organisations in Mali, CMDT and ODN, for all their faults, were able to provide a degree of risk management on behalf of the local farmers within their respective areas, and this had proved generally beneficial. Similarly, the patrimonial relationships that were found to be so significant with the RMG sector of Bangladesh also helped to mitigate the risks experienced by individual, small-scale producers. Other types of solution might also be found in other sectors, other countries. For future research, it seems important to try to understand better how particular riskmanagement institutions come into being, and how and why they work when they do; conversely, of course, it would be useful to know more about why such institutions fail to emerge in some circumstances where one might expect them to be of great potential benefit.

9

References Ajakaiye, Olu; Kilindo, Ali; Nyoni, Timothy S.; Proctor, Felicity; and Sen, Kunal (2006), “Institutions & Pro-Poor Growth Inception Phase Country Case Study: Tanzania”, IPPG Working Paper No.3 Chatterjee, Bipul; Davis, Junior R.; Eusuf, M. Abu; Harriss, John; and Purohit, Purnima (2006), “Institutions & Pro-poor Growth in Bangladesh: IPPG Inception Phase Study”, IPPG Working Paper No.2, June Hare, Paul G. and Davis, Junior R. (2006), “Institutions & Development: What We (Think We) Know, What We Would Like to Know”, IPPG Discussion Paper No.3, April Harriss, John (2006), “Notes on a Historical Institutionalist Approach to the IPPG Agenda”, IPPG Discussion Paper No.1, April Leftwich, Adrian (2006), “The Political Approach to Institutional Formation, Maintenance and Change: A Literature Review Essay”, IPPG Discussion Paper No.4, forthcoming Olukoshi, Adebayo; Dougnon, Isaie; Sen, Kunal; Morton, John; Sall, Ebrima; Coulibaly, Lamissa; and Traore, Abdoulaye (2006), “Improving Institutions for Pro-Poor Growth in Mali: An Exploratory Study”, IPPG Working Paper No.4, forthcoming Sen, Kunal; te Velde, Dirk Willem; Wiggins, Steve; and Cali, Massimiliano (2006), “Institutions & Pro-Poor Growth: Towards a Framework for Quantitative Analysis”, IPPG Discussion Paper No.2, April Wiggins, Steve; Schejtman, Alexander; and Gray, George (2006), “Bolivia Case Study: An Interpretative Summary”, IPPG Working Paper No.1, April

PGH/20.09.06

10