If not manufacturing, then what?

Botswana’s industrial development policy has long had manufacturing at its centre. There were many reasons for this. The “traditional” development model holds that most countries follow a natural progression, starting by producing and exporting mainly primary commodities, then moving on to manufacturing as the main source of exports and job creation, and eventually moving on to services.

There are, or were, good reasons for pursuing this approach. The long-term trend in terms of trade – the price a country’s exports, relative to its imports – has been negative for commodity exporters over a long period, as the price of primary commodities declined relative to imports of manufactured goods. This provides an argument for diversifying away from dependence on primary commodities to manufacturing. A second reason is that manufacturing is often thought to provide great potential for productivity gains, through “learning by doing” and moving from low-tech to high-tech manufacturing as skills and capacity improve. A third reason is that manufactured goods are exportable, and successful export-led manufacturing has great job creation potential.

Botswana’s experience with manufacturing has not been particularly successful, however. Between 1994 and 2009, manufacturing value added increased at an average rate of 4.1% a year - not bad, but not very impressive either, and slower than the growth of many other sectors so that manufacturing’s share of total GDP fell over this period. Employment in manufacturing increased by around 3.7% a year – also not very spectacular. Clearly, years of special support for manufacturing, through tax incentives, subsidies and institutional support, have not yielded the anticipated rewards.

But there is reason to believe that changing circumstances have made a focus less relevant than it was historically. Trends in terms of trade have switched – whereas the price of primary commodities declined for a long period relative to the price of manufactures, the opposite is now occurring. Rising prices for commodities – perhaps even a commodity supercycle that could persist for many years – and declining prices for manufactured items, remove one of the planks of the convention rationale for the move into manufacturing. And the intense competition from cheap manufactured goods from Asia makes it very difficult for African countries – with their relatively high labour costs and low productivity – to compete in global markets, especially landlocked ones with high transport costs, such as Botswana.

A second important change is that services are increasingly being traded internationally. Conventionally, many services were seen as non-tradeables, to be produced and consumed domestically but not traded on international markets. However, a great deal of the growth of world trade in recent years is due to increased trade in services, rather than manufactures. This includes some conventional services such as tourism, but also newly traded services such as business and IT services. Increasingly, as India is showing, services can provide a channel for export-led growth.

So what does this mean for Botswana? There are several implications. First, there is no particular reason to diversify “away” from minerals – instead we should aggressively develop the minerals sector, especially if the minerals sector itself can be diversified by exploiting a range of minerals (rather than our current extreme dependence upon diamonds). If we have a comparative advantage in an activity whose economic value is rising, then all well and good. Developing the minerals sector will in some cases include the beneficiation of those minerals (such as generating electricity from coal).

Second, there is no particular reason to focus on manufacturing, where our potential for comparative advantage is limited. While manufacturing has a role to play in the economy, there is no reason to give it special treatment – noting that an incentive in favour of manufacturing (such as tax concessions) is an incentive against other economic activities, which may have more long run potential.

Third, we need to explore the potential for developing services exports, especially those services that have potential for large-scale employment creation and productivity gains, such as business and financial services.

This will require a change of thinking on the part of government, moving beyond the “manufacturing is something special” approach. It requires investment in appropriate infrastructure (internet connections rather than rural roads). Most importantly, it requires engagement by government with the private sector in order to identify the bottlenecks to investment, and speedy action to resolve those bottlenecks.

How many flat-screen TVs can you buy for a shipload of iron ore?

One of the best-known ideas in development economics is the Prebisch-Singer hypothesis, which states that the terms of trade between primary products and manufactured goods tends to deteriorate over time. The hypothesis, which dates back to the early 1950s, asserts that countries exporting primary commodities (minerals and/or agricultural products) would be able to import fewer and fewer manufactured goods for a given amount of exports, and hence would become progressively worse off, or fall further behind developed countries. A common policy recommendation based on this idea is that developing countries should strive to diversify their economies and reduce dependence on primary commodity exports by developing manufacturing activities.

The hypothesis is closely linked to the idea that most countries follow a natural progression from producing mainly primary commodities, moving on to manufacturing and eventually to services.

The progression from primary commodities (minerals) to manufacturing has underpinned much of Botswana’s thinking on industrialisation and diversification, and the manufacturing sector has benefitted from incentives and special treatment by government. And yet the policy has been largely unsuccessful, in that manufacturing has shown only sluggish growth rather than dynamism. There are only a few examples of manufacturing firms that have demonstrated long-term sustainability, while many more have collapsed once subsidies or special treatment was withdrawn.

But perhaps the Prebisch-Singer hypothesis is wrong, or at least not as applicable as it once was. Two major developments suggest that it may be less relevant today.

First, the commodity “supercycle”, whereby an increase in the demand for primary commodities, driven by higher global economic growth, leads to sustained higher commodity prices;

Second, the steady fall in the prices of many manufactured goods, associated with a shift in the locus of global manufacturing production from developed to developing countries.

These two trends, if sustained, suggest that the reverse of the original hypothesis may now be true, and that the terms of trade of primary commodity producers are now improving, rather than deteriorating, and that a given level of primary commodity exports can now buy an increasing amount of imports of manufactured goods.

The importance of this was recently illustrated by the Governor of the Reserve Bank of Australia, a country that mainly exports primary commodities and imports manufactured goods. Five years ago, Governor Stevens noted, a shipload of iron ore bought 2,200 flatscreen TVs; now it buys 22,000. A long term trend of declining terms of trade seems to have been reversed, with Australia’s terms of trade improving by 42% since 2004.

The fact that commodity prices – for both minerals and foodstuffs – have resumed their upward trend after a temporary interruption in 2008-9 following the global financial crisis, suggests that this may well be a long-term development. Rising mineral prices are being driven, to a large extent, by demand from emerging market economic, especially India and China, which is likely to persist for some time, even if they eventually move to a less resource-intensive growth path. And while the decline in the prices of manufactured goods may be slowed as wages start to rise in China, there is still scope for manufacturing to move to other low-wage Asian economies.

What does this mean for Botswana? Certainly the focus of industrial development and diversification policy needs to change, with manufacturing no longer the centre of attention. But it also means that continued reliance on minerals, especially if the mining sector can itself become diversified, as seems to be happening, is not necessarily a bad thing.