HARNESSING TRADE FOR ACCELERATED DEVELOPMENT IN AFRICA
By Prof. L Alan Winters - Professor of Economics, University of Sussex, also of CEPR, IZA and GDN
Mobility would seem to be the very essence of trade: if things don’t move, there is no trade. This is true, and it informs the parts of this article that talk about efforts to reduce the costs of doing international trade, including trade facilitation and aid for trade. But there are three other aspects of mobility that I want to stress: mobility between sectors – notably structural transformation - and mobility between areas – internal migration and urbanization. Third, I shall argue that mobility lies not only at the heart of generating more output and income, but also at the heart of sharing that income in a more equitable and sustainable ways. In a sense the last is about social mobility and I recognize this as one of the most important elements of achieving a sustainable society; however, as an economist I do not have the skills to move beyond the analysis of incomes, so I will concentrate on that.
Trade matters. This is most true within a country – a village that had to be wholly self-sufficient would be grim, or perhaps impossible, to live in. But in the modern economy, it is also true between countries, especially small ones. Trade allows specialization and the reaping of economies of scale, and it generates competition and, via that and by allowing larger markets, incentivizes innovation. All of these dimensions allow us raise efficiency and live better. Thus reducing the costs of international trade has become a major objective of some governments and most donors over the last couple of decades. Achieving such reductions requires both hard and soft infrastructure as well as effective current operations, and is no easy task. However, there are large reductions to be made and the benefits of making them are large.
The seminal contribution on the effects of transportation costs in the modern world is Limao and Venables (2000) who used the quoted prices for transporting a 40-foot container to capture international transportation costs. They found that 50% of the variation in transportation costs between different routes could be explained by differences in infrastructure (a combination of roads, rail and telephones) and only 10% by distance.
Raising the quality of a country’s infrastructure from the 75th to the 50th percentile was equivalent to taking approximately 3,500 kilometres off the average sea trip or 400 km off the average land trip that its exports and imports had to face! More recently, Portugal-Perez and Wilson (2012) have examined bilateral trade between a large number of partners and find that infrastructure quality is a key constraint on developing countries’ exports, along with ‘soft’ infrastructure (essentially trade regulations). They use this observation to argue passionately for trade facilitation – which potentially addresses both aspects - as a policy priority for development. Vijil and Wagner (2012) carry this line of reasoning one step further and show that infrastructure constrains developing country exports and that aid for trade devoted to infrastructure has a material effect on the latter’s quality. From this they argue that a 10 percent increase in aid for trade flows to a developing country could raise the recipient’s (export/GDP) ratio by 2.3 percent, which they calculate is equivalent to a cut of approximately 2.7 percentage points in the tariff and non-tariff barriers that those export face.
It is obvious from this glance at the research literature that trade facilitation and aid for trade could make a great contribution towards higher incomes in Africa. I would, however, offer three cautions. First, aid has to be well spent – both technically and in terms of carefully identifying priority projects – either hard or soft. The latter requires a close engagement with the private sector – they are the people who trade – coupled with rigorous attention to ensure that they do not capture the process merely for private gain. Second, investing in infrastructure reduces the costs of trade but to translate that into reductions in the prices of traded goods and services one needs competition. Teravaninthorn, and Raballand (2009) looked at the likely returns to investment in each of five road corridors in Africa and found that they vary significantly with the institutional situation.
In particular, they found that in west and central Africa transportation costs could be significantly reduced by reducing fuel costs, improving road quality and reducing border-crossing- times. However, they argue that the trucking sectors are so cartelized and heavily regulated in these regions that any cost savings so produced will just go into truckers’ pockets rather than stimulate trade and enterprise.
The moral is that until regulation ensures sufficient competition to ensure that cost savings are passed on, road and border-crossing improvement schemes will tend to be unequalising because they will increase rents. These results refer to the mid-2000s and things are changing fast in Africa, but the moral is general: fixing the kit is one thing, fixing the problem is another and it takes careful analysis to identify exactly what is going wrong. The third caution is that trade costs also include policy barriers to trade. For sure in most cases the burden that transportation costs impose on imports is greater than the burden imposed by tariffs. However, the critical difference is that transport costs cannot be reduced to zero (ever!) and in fact can usually be reduced only at considerable expense, while tariffs can be reduced to zero at the stroke of a pen: I may have to pay hundreds of dollars to have a critical input delivered to my factory, but that is not as reason to increase its costs even further by taxing it. While we work hard on trade facilitation, let’s not forget the (technically) easy and cheap stuff – trade policy.
For two hundred years or more economists have understood that opening up to international trade allows specialization – concentrating on producing what you produce best, selling some (lots) of that to other countries and buying the rest of what you need on the world market. The key insight is that an economy that does not trade is obliged to produce all that it consumes and consume all that it produces.
International trade breaks that tyranny and allows producers to break free of the shackles of the domestic market and so to produce far greater quantities of certain goods (the ones the economy is good at) than consumers could possibly absorb. But taking advantage of trade in this way is not necessarily a painless process: it starts by cutting back on some activities in order to concentrate on others; in other words, on inter-sectoral mobility.
Anything that hinders this process is likely to delay or even prevent the gains from trade emerging. Some hindrances are natural, like geographical barriers, but many are man-made in terms of regulations and restrictions.
While the issue is far from closed among academic economists, there is certainly some evidence that countries that have restrictive labour market regulations, restrictions on firm entry and exit or serious capital market failures benefit less from international trade than more liberal societies – for example, Chang, Kaltani and Loyoaza (2009).
Economic development is similarly seen as hinging around the idea of switching sectors, at least since Nobel Laureate Arthur Lewis’s classic work in 1954. Lewis characterised development as the reallocation of factors of production and employment from low productivity “traditional” sectors, such as agriculture, to high productivity “modern” sectors, such as manufacturing.
This increases average productivity and hence average incomes. These two processes go together – no country has developed in the modern era without engaging increasingly and eventually heavily in international trade. However, there is a potential disconnect in the medium term: what if international trade encourages countries to specialise in primary products, so that factors of production move towards lower productivity sectors (e.g. agriculture) or sectors in which there are very few jobs (minerals)? While this may generate high income growth while primary prices are high, and indeed can provide a basis for long-run progress (think of the USA, which counts as a primaries exporter on most measures), it is not sufficient for development. For development, one looks also to see movement towards ‘modern’ sectors, not only manufacturing, but also in terms of services.
The arguments about the need for structural change provide the rationale for the Aid for Trade initiative which was launched at the Hong Kong Ministerial Conference in December 2005. Although its origins owed at least as much to the desire to keep developing countries inside the tent of the Doha Development Agenda as to long-run economic reasoning, it was aimed at addressing whatever governments thought might be preventing developing countries from developing their exports and particularly their export bundles (the goods and services which they actually able to export). This was an obvious first step towards inducing positive structural change.
Last year the Commonwealth Secretariat asked Xavi Cirera and me whether Aid for Trade (AfT) had actually stimulated structural change in Africa – Cirera and Winters (2014). Looking across Sub-Saharan African countries since 1995, we first sought to relate AfT to the costs of conducting international trade and to trade performance and then we looked for traces of the effects of AfT on the structure of the African economies. With the minor exception that maybe policy-related AfT reduces the clearance times of goods in customs, we found no relationship between AfT and exports or economic structure at all. One should not necessarily conclude from this that AfT is a waste, however. First, our analysis faced formidable data problems which could easily have clouded the results. For example, it is the donors who label an aid flow ‘Aid for Trade’, rather than the recipients who could do so in the light of what the money is actually spent on. Second, given the huge differences across countries and the array of forces that influence economic structure (including the primaries price boom), our tests based on relatively small amounts of relatively poor data may just be too weak to find an effect. On the other hand, our results do suggest that recipients should make greater efforts to make sure that AfT is well directed and well spent and then well evaluated. And the argument of this article is that this ought to include thinking seriously about what it is that allows people and resources to move between sectors.
The key insight is that an economy that does not trade is obliged to produce all that it consumes and consume all that it produces.
The Mobility of People
Arthur Lewis’s model of development depends on moving people between sectors, but this also depends to a large extent on moving them between places. If we consider Britain’s economic history as the first industrial nation, rural-urban migration lay at the very heart of the industrialisation process. It contributed an increase of about 1 percent per annum to the urban population every year over a century from around 1776, and reduced the rural population by about 0.8 percent per annum in 1776 rising to about 2 percent in 1866 partly because the numbers of migrants increased but mainly because the rural population actually fell – Williamson (1990). Likewise, the remarkable growth of China has been accompanied by a dramatic increase in urbanisation, and now over half of China’s population lives in urban areas.
The role of cities in development is not a surprise. Even a little reflection suggests that the provision of many of the services that characterise development such as health, education and law enforcement are much easier to provide and likely to be of higher quality where population density is higher.
Moreover, there are very strong reasons why production is more efficient there too – cities allow for deeper and more frequent interaction, better coordination, a quick spread of ideas and large local markets to encourage economies of scale. Bettancourt and West (2011) summarise the evidence drawing on both developed and developing country experience to suggest that a doubling of city size is associated with an average increase of around 15 percent in measures such as wages and patents produced per capita. Their data also reveal that as city-size doubles, ‘its material infrastructure -- anything from the number of gas stations to the total length of its pipes, roads or electrical wires – does not’. And even better, perhaps, cities in developed countries are greener than other areas because people living closer together require less energy to get about. Of course, cities can be unpleasant – lawless, dirty, crowded – but this is not inevitable if they are suitably planned and governed.
Africa has low population densities on average, which makes industrialization an uphill struggle, and most countries have low rates of urbanization, which suggests that they are not doing the best they can to overcome this. Moreover, in many African countries the prevailing attitude is that rural urban migration is a problem which needs to be minimized. For sure, the difficulties of managing large inflows of migrants are challenging and there may sometimes be a case for temporarily slowing the flow; however, over the medium run deciding to facilitate and then manage this form of mobility can only be beneficial. In the University of Sussex we have a Research Consortium (Migrating Out of Poverty Research Program Consortium, http://migratingoutofpoverty.dfid.gov.uk/index ) that studies internal (as opposed to international) migration. Work in several African countries has revealed that even if migration to the town or city is uncomfortable and often dangerous for the poor and less skilled (especially for women), a large majority of the people we have interviewed see it as a way of improving their economic fortunes. Plenty also see it as a way of improving their lives in other, social dimensions, and, again, especially young women who see migration as a route to greater independence than traditional rural societies provide.
Mobility and Sharing the Gains from Trade
My final paean to mobility is to observe that mobility is key to sharing the gains from trade – or, indeed, the pain and thereby mitigating it. If a sector receives a boost in demand from a trade reform, the demand for labour will go up. If the labour force cannot be expanded, the existing workers will experience a strong rise in wages but nobody else will see any benefit: you get a narrow but large impact on welfare. If, on the other hand, the sector can bid workers away from other sectors, more people will benefit but wages will increase by less; the effect will be broader but smaller – i.e. less unequal. Similarly – and this is important – when a sector suffers from a trade reform because import competition causes it to decline, if the workers have no alternative employment they will probably suffer a large wage decline, whereas if there are other sectors they can go to, the wage decline will be mitigated. International trade typically raises living standards by reducing the prices of the goods that people consume and by stimulating economic growth overall, but nonetheless there are bound to be sectors that have to contract. Hence the fact that mobility reduces the hit on workers in such sectors, is important both in reducing the burden of adjustment (because both theory and experience suggests that halving the size of a shock more than halves the welfare costs of bearing it) and in increasing the likelihood that the net effect of the negative specific shock plus the positive general benefit is positive. I have spent some time studying the effects of international trade on extreme poverty in developing countries, and have concluded that mobility is a key consideration not only in reaping aggregate benefits from trade reform but in achieving a moderately equitable – and hence politically more sustainable – distribution of those gains. A survey of the recent literature – Winters and Martuscelli (2014) – uncovered several instances in which mobility played a key role in spreading (or otherwise) the benefits of trade reform. For example, in Vietnam (as in most places) unskilled workers are less mobile that skilled ones and so the export benefits that Vietnam reaped from its 2001 trade agreement with the USA, which greatly improved access to the US market, were geographically more concentrated for unskilled than for skilled workers. In Indonesia, workers are not very mobile across firms and so trade liberalisation strongly benefits people in export industries and does little for those in import-competing firms.
Similarly, the evidence shows that the impact of India’s trade liberalisation in 1991 was most pronounced among the least geographically mobile sections of the population - those at the bottom of the income distribution - and in Indian states where inflexible labour laws impeded the reallocation of factors of production across sectors.
Conclusion Mobility is by definition the sine qua non of international trade – you have to move goods or services across borders before you see any international trade.
Policies that increase such mobility by reducing its costs have a strong effect on trade and, through this, welfare. But mobility is also key within the trading country: the main gains from trade arise from changing your production bundle – you have to be able increase activity in certain sectors and reduce it in others. If this is thwarted, you do not get great gains and may even record losses from a trade liberalisation. But moving sectors often entails moving places, and if such movement results in agglomeration into large groups – cities or large towns – you get not only gains from trade but also gains to productive efficiency, innovation and the effectiveness of services.
Moving places means migration and it is important that African governments start to see internal migration in a more favourable light. Finally, mobility helps to share the benefits of trade (and other economic activity) more equitably and so both fosters development and increases its sustainability. Thus, as the title of this article says: if you are thinking about trade, think also about mobility.
- Bettancourt, L. A., and West, G. B. (2011). ‘Bigger Cities Do More with Less’, Scientific American, 305(3), 52- 53.
- Chang R, Kaltani L and Loayza NV. 2009. Openness can be good for growth: The role of policy complementarities. Journal of Development Economics, 90(1):33-49
- Cirera X and L A Winters (2014) Aid for Trade and Structural Transformation in Sub Saharan Africa, Commonwealth Trade Policy Discussion Papers, No 2015/01, London, Commonwealth Secretariat.
- Limao, N. and Venables, A.J. 2001, “Infrastructure, Geographical Disadvantage, Transport Costs, and Trade”, World Bank Economic Review, vol. 15, no. 3, pp. 451-479.
- Portugal-Perez, Alberto, and John S. Wilson. “Export performance and trade facilitation reform: hard and soft infrastructure.”World Development 40.7 (2012): 1295-1307.
- Teravaninthorn, Supee, and Gaël Raballand. Transport prices and costs in Africa: a review of the main international corridors. World Bank, 2009.
- Vijil, Mariana, and Laurent Wagner.“Does Aid for Trade Enhance Export Performance? Investigating the Infrastructure Channel.” The World Economy 35.7 (2012): 838-868.
- Winters, L Alan and Antonio Martuscelli (2014) ‘Trade Liberalisation and Poverty: What have we learned in a decade?’ Annual Review of Resource Economics, vol 6, 2014, DOI: 10.1146/annurev-resource-110713-105054