‘MILLENNIUM PLUS’ COMMUNITY DEVELOPMENT FOR SOUTH AFRICA

Don Ross

University of Cape Town and University of Alabama at Birmingham

dross@commerce.uct.ac.za

Of crucial importance to all parts of the transport services and materials sector in South Africa is the way in which the Government chooses to implement its ambitious plans to reinvest in the country’s basic infrastructure. How will it navigate competing demands from urban and rural environments, given the divergent economics that describe them? How will it balance the goals of poverty fighting, skills empowerment, and keeping SA internationally competitive, as it considers infrastructure project options?

Up to now, SA has not had a nationally coordinated set of guidelines for prioritizing infrastructure projects, or for prioritizing specific infrastructure elements such as roads with respect to others, such as irrigation. In the late 1990s it was hoped that provinces and regional municipalities would each do this for themselves through the Integrated Development Plans (IDPs) mandated from them, but this process has so far not provided much help. In part, this is because designers of IDPs have not been guided by any common project evaluation framework. Although participants in industries that supply and maintain infrastructure components are heartened by the Government’s ramped-up financial commitment to re-investment in the basic national plant, many are frustrated by the prevailing lack of clarity on which aspects will be targeted in what order.

Investment in public infrastructure is a topic in development economics. There are a variety of economic models of such investment that could be used to guide policy. In the following article, I will first report on the general current state of play in thinking among development economists about solutions to Africa’s problems. Then, in a second, longer part I will reflect on the implications of this thinking for South Africa, which faces special issues because of its greater aggregate wealth but more pervasive and urgent inequalities. One prevailing idea in the discussion to come is that infrastructure investment will be of greatest national benefit if it is implemented in comprehensive clusters – e.g., roads with irrigation, rather than a road here and a pipeline there. This in turns demands a level of coordination that is politically and administratively challenging. In light of this, it will be crucial that relevant sectors, such as transport and road-related industry players, understand both the economic arguments for and the political issues around the proposed policy approach, so that they can weigh in to help drive it along as partners.

1. The Millennium Village programme in sub-Saharan Africa

Here are two things it isn’t sensible to be: an Afro-pessimist and an Afro-optimist. By ‘Afro-optimism’ I mean an attitude which takes for granted that Africa will overcome its poverty and underdevelopment without devoted effort on anyone’s part. This might seem like an amazing view for anyone to take, but in fact it was the leading stance of development economists in the 1960s. Orthodox theory then held that capital would flow naturally to wherever it was in shortest supply and would attract the highest price – that is, to poorer countries. Thus it was thought that less developed and more developed countries would tend to converge unless actively prevented from doing so. When the flaws in this reasoning were exposed by devastating experience in (especially) Africa, development theory had to be re-examined. Clearly, something had to be added to poor countries’ mixes of circumstances to get them growing. At first this was thought to be financial capital, and donor nations poured cash into Africa. Results were disappointing. Then the missing element was thought to be infrastructure on grand national scales, and donors financed dams, factory towns, and modern concentrations of agriculture. Results were disappointing. The so-called Washington consensus of the 1990s concluded that what was needed were new policies that would facilitate integration with the world economy through increased trade and capital flows. Results, at least in Africa, have been disappointing. Thus development economics now finds itself with a profusion of theoretical models and a litany of policy failures. This is the main reason that Afro-optimism is rare and Afro-pessimism is common.

Certainly, we should learn from this that Afro-optimism is unsustainable: we cannot settle back and expect African poverty to vanish in the undisturbed course of events. Work is required. But Afro-pessimism, the conclusion that we might as well stop wasting resources on Africa because nothing will ever work, is neither a valid conclusion nor morally acceptable. If people always regarded a problem as insoluble when their first three or four attempts at solving it didn’t work, we really would all still be living as hunter-gatherers. Most people who conclude from the short, frustrating history of development economics in Africa that the continent should be left to its supposedly endemic misery would not reason so hastily if the first three doctors they visited for a serious health problem struggled to find the diagnosis. Thus Afro-pessimists suggest by the quick leap they take to their conclusion that they in some sense want to give up. To seek a reason to do nothing to help very poor people is moral failure and should be treated as such.

A few economists are Afro-pessimists, but fortunately most are not. (The former World Bank economist and recent book author William Easterly is sometimes cited as a leading voice of pessimism, but this is unfair – he clearly urges that we keep trying.) What it’s most reasonable to think we’ve learned over the past fifty years, in my view, is that there is no simple, overarching policy for development to be found at the scale of whole nations. Therefore we should stop looking for such a magic bullet and then being disappointed each time we fail to find it. Poverty recedes only when a combination of factors interact and amplify one another. This in turn suggests that it’s bound to be hard to find successful anti-poverty measures if one confines attention to too large a scale. At the level of whole countries, one is too far from the knots – individual communities – at which particular such combinations of factors can be anchored. Any policy implemented only at a national level will improve prospects for some communities and worsen prospects for others, so effects of such policies will usually be ambiguous by their nature. This doesn’t imply that we should never undertake national reforms. Quite the opposite: these will typically be necessary aspects of efforts to facilitate local improvements. The point is that there will be trade-offs, and we have to make them by reference to possibilities and efforts at local levels.

The idea that we best combat poverty through concerted, coordinated efforts at the level of communities is now being practically taken up in Africa by the UN’s Millennium Project and its head, Professor Jeffrey Sachs. Sachs is sometimes criticized for appearing to suggest in his recent book The End of Poverty that he has at last found the Big Solution where all others have failed. Worse, critics charge, his ‘solution’ is one that has already been unsuccessfully tried: give lots of money to national governments for national-level investment initiatives. I think that these criticisms have some justice but that there are points to be made in Sachs’s favour too. However, entering into this debate isn’t my purpose here. Instead, I want to describe one aspect of the Millennium Project to which the sort of criticism I’ve just mentioned clearly does not apply. Then I want to consider what we have to learn from consideration of this aspect for anti-poverty efforts in South Africa, where our development challenges are serious but more tractable than in most countries to our north.

The aspect of Sachs’s initiative I have in mind is its implementation through a set of ‘Millennium research villages’. So far 12 have been established, in Senegal, Mali, Ghana, Nigeria, Ethiopia, Kenya, Uganda, Rwanda, and Tanzania. 66 more have been identified for launch around the continent, and the plan is that there will be 1,000 by 2009, with this number to then multiply until it takes in all of sub-Saharan Africa. In the opening phases, each Millennium village (MV) receives $250 US per resident over 5 years from the G8’s Millennium Fund. As the number of MVs approaches the number of African villages this will obviously cease to be possible. But the idea is that it won’t need to be: as MVs expand this will mean that the continent will be climbing out of poverty, businesses and tax bases will be expanding, and MVs’ numbers will rise exponentially because their growth will be kicked off by the rising prosperity all around them.

In this exuberant hope we obviously see an element of the kind of Afro-optimism I criticized: we are again invited to imagine Africa growing all by its natural self. So let us set aside this starry-eyed part of the scheme for the moment – I will come back to it later – and focus on the core that is actually happening, namely, the attempt to select communities and apply ‘extreme make-over’ to turn them from poor and stagnant to growing and getting richer. How is this supposed to work?

The key economic concept behind the MV idea is that of the multiplier. If you start with a few numbers and then add more, the resulting total rises much slower, and at an exponentially decelerating rate, than if you introduce new numbers by multiplication. This applies just as surely if the numbers in question represent inputs to and predictors of wealth. But do factors that predict wealth actually tend to multiply in reality? The answer to this is: not always but far more often than not as long as critical thresholds are attained. If someone has access to financial capital they might be able to start a business. If there is decent and reliable infrastructure where they live, this means that their costs will be lower and each unit of capital is that much more likely to be the seed for a successful business. The availability of an educated workforce makes such success even more likely. Tax revenue from successful business ventures is what mainly pays for public infrastructure and education. This in turn begets more successful businesses. In addition, already educated people will tend to move to, rather than emigrate from, economically growing areas. Independently of monetary inputs, this in itself improves business prospects and the quality of education (because educated parents are as important an input to children’s education as good schools). One thus does not predict the wealth-creating impact of financial capital, infrastructure and human capital by adding them; since they naturally tend to reciprocally boost one another, one calculates their joint effects by multiplying them.

Most people understand this. Development economists in the past tended to recommend single-factor approaches not because they didn’t appreciate multiplication, but because they hoped that privileging one input – which is extremely helpful for policy coordination – might be sufficient to set off multiplying processes. Each main class of development inputs had its moment in the sun as the favorite. Another point that is often lost, and the key one for my purposes here, is this: multipliers work more reliably on scales smaller than whole medium-sized countries. At the level of a whole country it has turned out that coordination on simple policy objectives tends to impose costs in miscoordination in the way that people respond to overly sweeping incentive shifts from sector to sector and from region to region. As a result, virtuous multiplication often gets swamped by unexpected consequences and distortions. Hence the new focus in development economics on multi-factor policy emphases at community-level scales.

This is the rationale for the MV programme. Villages are selected on the basis of need (are they very poor?), potential (is there something in the area for capital infusion to work on, such as raw materials, good soil, or proximity to markets) and absence of predatory government (will the rule of law protect the fruits of growth so it doesn’t get stifled?). Because the MVs are so poor, investment emphasis is on ten fundamentals which are always critical in sub-Saharan Africa (at least outside of South Africa):

(1)  soil health, water management and preservation of crop quality

(2)  control of malaria (through, e.g., durable bednets)

(3)  clinical services (for AIDS, obstetrics and prevalent African diseases)

(4)  safe drinking water

(5)  roads

(6)  reliable, locally sourced nutrition for schoolchildren

(7)  reliable and affordable energy for households and enterprises

(8)  cellular telephony and internet access

(9)  transition from subsistence farming to surplus production

(10) access to credit and market mechanisms (e.g. cooperatives for warehousing and supply-chain management)

How is the programme working so far? It is only two years from inception, so these are early days. But there are clear signs of growth in all twelve of the pioneering MVs, and, most encouragingly, there is a correlation between rates of growth among the twelve and the chronological order in which implementation got under way in them. This is evidence that the accelerated growth can be attributed to the project. The most frequent criticism thus far is based on a consequence of success rather than failure: concern has been expressed that the MVs pull in resources from contiguous areas, making people in them worse off.

Consideration of this objection brings us back to what I called the ‘starry-eyed hope’. Every village in Africa cannot become an MV. A consequence of development must be that some communities economically absorb others. This is the process of urbanization, and urbanization is a constituent aspect of growth in all industrial societies. Emphasizing village development should not be associated with the romantic anti-economics of the ‘small is beautiful’ movement. This point will emerge as especially important when we consider the current approach to economic development in the context of South Africa.

2. Applications in South Africa?

There are areas in South Africa – for example, in the former Transkei and parts of rural KwaZulu-Natal – where levels of poverty and underdevelopment are comparable to those found generally in sub-Saharan Africa. This in itself helps to illustrate the point made earlier about the scales at which multiplier effects in development work. They have obviously been effective in the history of Gauteng and the Western Cape, where financial capital and infrastructure are plentiful, and human capital, though scarce by developed-world standards, is present in abundance compared to continental norms. Yet their multiplication cycles have not extended across the country as a whole. In part this is a legacy if the fact that the apartheid regime deliberately stifled such extension. However, racist attempts to block normal economic processes seldom actually worked. Furthermore, multiplication from the large urban economies has spread in only a patchy and limited way during the twelve years of democracy and open internal markets.

Thus there is motivation for the general MV concept to be extended, with important modifications to be discussed below, to SA. In fact, the concept is not new here. Since 2001 the Southern African Bitumen Association (SABITA) has funded a project to develop an Opportunity Value Assessment (OVA) tool for SA. This is an instrument designed to measure the local multiplier impact of coordinated infrastructure and capital deepening investments at the community level. It might be wondered why the economists working for SABITA thought we needed such a device. Why not just run standard cost-benefit analyses (CBAs), of the sort which the World Bank, among other agencies, has done valuable work in tailoring to specific classes of infrastructure in the developing world? The answer to this question takes us again, by another path of reasoning, to the importance of concentrating on integrated investment efforts at community-level scales.

'Development’ is, in the long term, all about moving resources, including human resources, from areas where they are inefficiently deployed to areas where they can be more efficiently used. ‘Efficiency’ is here understood as a measure of expected effect on medium-run wealth creation. One thing this typically implies in the current world is urbanization, the shifting of economic activity and population from rural to urban conditions as capital stock accumulates – a very 'classical’ picture, which generic CBA models encourage. However, because SA’s most efficiently placed capital is highly efficient by developing-world standards but also thin by first-world ones, there is constant high risk here of shifting populations from rural to urban areas too quickly. In some sense, all South African planners have always been aware of this. Apartheid-era planners certainly were, but their situation was hopelessly complicated by struggles between economic imperatives for moving labor, on the one hand, and long-term political goals that were both morally odious and rationally impossible, on the other hand. Unfortunately, we cannot escape from these complications by sheer willpower and better intentions. It is possible to depopulate rural areas at an unsustainably rapid rate through over-concentration of new capital reinvestment in already existing efficiencies. Furthermore, in SA, with its world-leading level of inequality, this tends to keep resources concentrated in the hands of people who are already relatively well off. Urbanization will and must continue in SA. But to avoid much – unnecessary – misery along the way, it needs to be accompanied by attempts to create economically growing communities outside of the large cities.

The focus of OVA design has been on community-level development from the beginning. To that extent we in SA were thinking along the same lines as the MV programme before the latter began. A team funded by SABITA began in 2001 by reviewing municipal Integrated Development Plans (IDPs) in northern KwaZulu-Natal. We found that almost none of the IDPs were attentive to potential multipliers, despite the widespread resort to use of private consultants in drafting them. In mid-2003 we estimated that were the existing KZN IDPs to be implemented just as written, this would represent a 70-80 % deadweight loss on the opportunity value of the capital.

We can properly be said to be developing a poor area only if what economists call its terms of trade are improving, that is, if we see increases in both outflowing goods and services from it and in financial inflows to it. In general, projects will promote such outcomes just in case they (i) yield improved ratios of locally generated production inputs to externally generated ones, and (ii) increase the local money supply and the velocity of its circulation. To put this point vividly: when we are trying to improve welfare within a small community, we want to see more vehicles carrying produce out, while we simultaneously count more vehicles carrying revenues in.

OVA is not a static instrument. It has been designed to its present point by adjusting standard assessment tools to make them sensitive to South Africa’s large inequality levels between communities. The approach from here will be to continuously refine the model by comparing its outputs with fine-grained survey data gathered as individual projects are rolled out. For each such project, an intended impact zone is pre-identified. The zone is studied with respect to both direct demand for the planned infrastructure resource in question, and with respect to well-established proxies for welfare – especially changes in consumption expenditures. This reference to welfare raises a conceptually tricky issue that is especially relevant in the context of anti-poverty policy in South Africa.

I do not think that there has been clear debate or, therefore, progress toward a national consensus on what the goal of anti-poverty policy in South Africa in fact is. It is clearly a part of that policy that welfare levels of poor households should increase. It is also a goal – one particularly pertinent to infrastructure provision – that, as described above, impoverished communities should be 'developed’, in the sense of seeing improving terms of trade. This second goal reflects attention to the following interrelated facts: (i) an important aspect of anti-poverty policy is rectification of past injustices with respect to distribution of resources amongst population groups; (ii) extreme poverty in South Africa is disproportionately concentrated in rural areas; and (iii) the overwhelming majority of poor rural South Africans are black.

These two goals – welfare improvement and local economic development (LED) – can often be pursued in tandem, because success on one goal will often, in the natural course of things, be correlated with success on the other. Thus, for example, it would be surprising if a poor rural community showed sustained improvement in its terms of trade over a meaningful period of time without corresponding improvement in its average household welfare. However, there are readily imaginable circumstances in which increasing welfare levels might counter-predict improved terms of trade. Suppose, for example, that a new road made it less expensive for people from a rural village to travel to the nearest city, where they could buy food and clothing more cheaply. Their household welfare would thereby increase, but the balance of resources flowing out of their community might worsen.

Considerations of this sort suggest that household welfare and community development need to be tracked as separate indicators. (Obviously, the same data will often be relevant to both.) Consumption expenditure is certainly the most reliable proxy for household welfare. But changes in terms of trade must be independently measured also. A third indicator that might serve as measure of the expected congruence of welfare and LED is women’s income. Evidence from developing countries around the world shows that women consistently invest substantially greater parts of new marginal income in the human capital of their children than men do. Such investment is itself a form of LED; but it also, equally clearly, an aspect of household welfare.

Where direct measurement of LED (i.e., terms-of-trade improvement) is concerned, OVA uses what is known as a ‘social accounting matrix’ (SAM) approach to construct an aggregate measure. The SAM methodology is directly derived from so-called factor-productivity models as originally developed for application to international trade. Rural South African local economies have much in common with poor countries: shortage of readily fungible capital, low diversification of business base, skill shortages, and under-developed links to wealthier economies. An especially important indicator in South African communities is improvement in output per unit of imported (especially fossil) fuel, since almost all returns on this significant production input are captured by cities.

OVA thus examines development investment impacts in terms of three aggregate policy targets: (1) sustainably rising household consumption levels; (2) increasing women’s income; and (3) improving terms of trade in communities. Different relative weightings amongst these targets will yield alternative project prioritization schedules. This is appropriate: anti-poverty policy in South Africa is to some extent about welfare improvement, and to some extent about community development. As the MV programme appreciates, both of these goals are best pursued starting from local scales.

I have so far concentrated on issues in rural community development. This is not the whole story, or even the biggest part of it, where LED in SA is concerned. As President Mbeki has emphasized in public remarks for years, SA is characterized by two large economies – a formal, wealthy one and an informal, much poorer, one – that are only shallowly linked to one another. This in itself is not unusual; indeed, it is typical of large middle-income developing countries. However, a property SA shares only with Brazil is that our two economies are on top of one another geographically, with the informal sector concentrated in the giant former townships in (especially) Gauteng. The potential multipliers available from fusion of the two Gauteng economies into one could drive an explosion of new wealth. This will not likely happen through a process of ‘colonization’ of the informal economy by the existing formal economy. The reason for this is that there are typically higher short-term returns to be obtained on a marginal unit of capital from combining it with another such existing unit in the formal sector than from trying to push it across the formal / informal divide with attendant high risk. A more plausible path is for the communities based on informal activity to formalize ‘from within’ as they grow wealthier, and then to grow links into the existing formal sector as their entrepreneurs seek new capital and (very importantly) new technology. As relative risk levels begin to even out, Gauteng, and on smaller scales the Western Cape and eThekwini, will each become one economy instead of two through a process of genuine fusion rather than one-way takeover.

How and why should we expect the former townships to become wealthier? The answer is the same as in the case of rural villages: if financial capital, infrastructure and human capital can all be driven over critical thresholds then they can be expected to multiply. Then, as (e,g.) Alexandria’s terms of trade with Johannesburg improve this will just amount to part of the process of fusion described above.

Once we reach the stage of talking about developing communities which are integral parts of a great city, we have come quite far from the ‘bare bones’ focus of the MV programme. That programme’s ten investment targets obviously need revision in transfer to the SA context. More even provision of clinics and safe drinking water are processes that the democratic regime already has relatively well in hand. Roads remain an urgent priority in SA just as in the rest of Africa. They constitute the basic infrastructure for multiplication, the arteries that bind communities into networks and are, literally, the avenues for improving terms of trade. In the rest of Africa, investment in roads typically consists of providing them were none previously existed. In SA, the appropriate intervention should generally take the form of upgrading gravel roads – which are expensive to maintain once traffic volumes are significant, which contribute to higher costs of product shipment, and which are much less environmentally friendly and good for people’s health than paved roads. In SA, LED should also attach greater relative emphasis to improved energy efficiency and reliability, to widening internet access, and to extension of credit than characterize the MV schemes elsewhere in Africa (at the present stage).

Note, however, that these are transformations of emphasis among the MV targets, not different goals altogether. The core economic idea remains the same: promoting development by selecting priority communities (considering both need and potential), studying their potential multipliers, targeting investments in infrastructure, human capital and market structures until multiplication begins driving itself, and then evaluating the results by reference both to improved household welfare and local terms of trade.

Note also that, as with the MV programme, this strategy implies transformation of population distribution, and hard political choices. Some communities must be selected for MV-style investment ahead of others. As a matter of logic, every community in SA can’t enjoy improved terms of trade, since net outflows must equal net inflows. This is just a way of re-stating the point made earlier, in answer to the leading criticism of the MV programme, that the objective of that programme obviously cannot be for every village in Africa to become a MV. The absorption of some community economies by others – urbanization – is the very nature of growth in a modern economy, not the appropriate basis of a policy objection. Perhaps it sounds paradoxical to say that development should focus on communities in order to ultimately submerge their identities. But sounding paradoxical isn’t equivalent to being paradoxical.

It is intended by the architects of the MV programme that it will eventually extend to SA. However, the early phase that depends on first-world donor funding will not include us. Here, MV promotion will depend on our own resources. It will therefore also depend on our own political leadership rather than that of international agencies. Both of these things are consistent with basic assumptions that have guided our transformation since 1994. As Alan Hirsch emphasizes in his recent book Season of Hope, SA has carefully avoided dependence on aid or off-market loans from development agencies for the sake of keeping our social destiny in our own hands. It is, I think, good to be proud in this way, so long as we recognize that this implies assumption of the hard choices – selecting this community ahead of that one as a development anchor – as ones we will not be able to slough off as dictated by external agents or circumstances.

In light of these differences in the way that MV-style development will be manifest in SA – more sophisticated investment targets, self-financed and self-led in all phases – I suggest that we designate our coming LED hubs as ‘Millennium Plus villages’. By this designation we will simultaneously stress the continuity of development in SA with what is happening in the rest of Africa, and the important respects in which our circumstances are different.

I have been concerned in this article to emphasize that the MV approach does not come to SA as a novelty dreamt up by others; we were already working along these lines. The idea that development should be driven from the municipal level was implicit in the IDP process launched at the end of the last decade. That process has so far disappointed because we haven’t been entirely clear in our own minds how we should select among investment alternatives. In this respect, our short history of self-guided development economics recapitulates the wider story of that effort across Africa. We can and should learn from that story while retaining leadership of its next steps to be applied at home.