Introduction
In the absence of the definite answer to the Ultimate Question of Life, the Universe, and Everything (not counting the number 42 from Douglas Adams’ “Hitchhiker’s Guide to the Galaxy”), we’re left with partial answers to more specific questions of socio-economic development. Economic science has offered several “laws” that attempt to describe long-term (secular) development trends by postulating certain relationships between various variables.
The honor of starting this process belongs to the founding fathers of economics. Both Adam Smith and David Ricardo noted the tendency of the rate of profit to fall (TRPF) as an empirical phenomenon but it was Karl Marx who offered a theoretical explanation. In his view, the tendency is due to a change in the “organic composition” of capital: as the fixed capital increases and the variable capital used, among other things. To pay workers’ wages, decreases, the rate of profit is bound to fall since surplus value can be extracted from labor only.
The founding fathers also believed in a secular improvement of the terms of trade of primary products. This belief follows from two classical laws – the operation of the law of diminishing returns in the primary producing sectors and the law of increasing returns in manufactures. The policy implication of this classical proposition is that an agricultural country need not industrialize to enjoy the fruits of technical progress in manufactures; free play of international market forces will distribute the gains from technical progress of the industrial countries to the agricultural countries by turning the terms of trade to the favor of primary products and primary product exporting countries.
This is not at all what the Prebisch-Singer theory, which is yet another attempt to describe a secular trend, suggests. Named after two famous development economists who first explored its implications in the 1950s, the theory argues that there was and would continue to be a secular decline in the terms of trade of primary commodity exporters due to a combination of low income and price elasticities of demand. This decline would result in an ongoing transfer of income from poor to rich countries that could be combated only by efforts to protect domestic manufacturing industries through a process that came to be known as import substitution.
TRPF and the Prebisch-Singer theory are just two examples of long-run relationships between economic variables conceptualized in economics. The reason these and some other long-term trends (e.g. a positive correlation between the savings ratio and the rate of growth in the Harrod-Domar growth model) generate so much interest and heated discussions is their theoretical importance and practical consequences. TFPF, in Marx’s view was an iron-clad proof of the eventual death of capitalism. Progressive exploitation could not but lead to a social explosion and revolution. On the other hand, capitalists could slow down this trend by expanding existing and opening up new labor-intensive sectors.
Likewise, the Prebisch-Singer theory has important implications for countries’ development strategies and, in particular, for the identification of the relative virtues of encouraging a country to specialize in primary commodities versus manufactures. What is more, because of the “fallacy of composition,” the sum of the countries’ individual efforts to promote commodity exports intensifies the downward trend of commodity prices. Consequently, these strategies are less effective for developing countries as a group than they are for any one country on its own. This fact has not been taken into account in the structural adjustment programs implemented in developing countries during recent decades. Thus, if this is the case, multilateral intervention in the market may be necessary in order to prevent this trend from hurting the interests of developing countries.
Decline in ToT: myth or reality
Paul Bairoch (1993) believed that the secular decline in the rate of profit was just a myth, alone with many others he analyzed in his Mythes and paradoxes de l’histoire économique. According to Bairoch, this myth is grounded in the wrong selection of the original data and the wrong use of the price indices, which served as the empirical evidence for Prebisch and Singer to formulate their theory. If a different time period had been selected. The original study used the data for a period between 1867/1880 and 1936/1938 and the price index based on the UK export and import prices. Bairoch argued that had a different period and different price indices been used, the result would have been different (even though he recognized that even in this case the decline would have been about 20%).
But Bairoch’s main point was different. Writing in the early 1990s, he claimed (and supported it with trade statistics) that the terms of trade for developing countries actually improved between 1867/1880 and 1989/1991. He recognized though that the improvement was uniquely due to a sharp rise in the price of oil in the early 1970s, which improved three times the terms of trade for oil exporting countries. The situation for other developing countries was not so good, marking a 20% decline between 1950/1954 and 1989/1991, in part because higher oil prices translated into higher prices for manufactured goods.
More recent research by David Harvey and colleagues (2010) found that overall, 11 major commodities show new and robust evidence of a long run decline in their relative price. These commodities are aluminum, coffee, heights, jute, silver, sugar, tea, tobacco, wheat, wool, and zinc. For example, the relative price of an important commodity like coffee has been declining at an annual rate of 0.77% for approximately 300 years.
Both because of this theory and because of the unfavorable terms of trade trends, developing countries have been trying over the past several decades to diversify into manufacturing exports. Some countries (South Africa, Taiwan, Hong Kong, Singapore) have transformed themselves from commodity exporters to manufacture goods and high-quality services exporters. However, for many developing countries this has proven to be unachievable. Relative prices within manufacturers have also diverged: over the past few decades, the prices of the basic manufactured goods exported by developing countries fell relative to the advanced products exported by rich countries. The price of textiles fell especially precipitously, and low skilled electronic goods are not far behind. In particular, using alternative methods, the United Nations found that the real decline in developing country export prices of manufactures in the 1980s was about 3.5% per year, or about 30% for the decade.
The tricky part of identifying a long-term trend is its long nature. Assuming the perspective of long cycles, it becomes clear that terms of trade develop in cycles and differ in different time periods due to a combination of factors, such as productivity growth in manufacturing and agriculture, invention of new technologies, transition to new sources of energy, conflicts and crises, etc. A very long trend (such as the one presented by Ocampo and Parra-Lancourt for 150 years starting from 1860) would show boosts and busts in global terms of trades reflecting the commodity price index and holding everything else equal. This long period does show a declining trend in the non-fuel commodity price index, indicating possibly a declining trend in the terms of trade for commodity exporting countries.
Over the past few decades, economies of many developing countries have structurally changed, with the share of agriculture shrinking and shares of manufacturing increasing. But this structural change has not brought as many benefits to most developing countries as they had hoped. Whereas the sectors outside agriculture have grown, in many countries agriculture remains the key employment sector. Manufacturing and services despite their higher share in the GDP offer few highly paid jobs and are characterized by low productivity. As a result, many developing countries rely on a small number of primary commodities to generate the majority of their export earnings. Overall, for the least developed countries, approximately 60% of export earnings are derived from primary commodities. However, for 40 countries, the production of three or fewer commodities explains all export earnings.
Uganda and its neighbors, close and distant
Uganda is a typical example. It relies on just two major commodities for export, gold and base metals (36%) and coffee (22%). Informal cross-border trade (ICBT) consists mostly of agricultural produce and fish. This export structure where gold dominates is relatively recent.
Until 2016, Uganda relied mostly of coffee as the major export earning commodity, supported by fish and tea exports. The country has experienced a sharp increase in gold exports since 2016. However, domestic gold production is low, with re-exports from DR Congo taking the largest share of the commodity’s export volumes. With the significant drop in gold export volumes in 2022 to US$200 million down from US$1,819 million in 2020, it is not clear if gold will retain its status as the foreign exchange earner.
So, how Uganda’s terms of trade changed over the past 22 years? On the face of it, there was a slight upward trend even though Uganda finished 2022 at the same level as in 2000 (about 100%). Unsurprisingly though the decline started in March 2020, which coincided with COVID-19 restrictive measures but despite a recovery in the GDP growth (which grew from 3.0% in 2020 to 3.5% in 2021 and about 6% in 2022), the terms of trade haven’t been able to recover to date.
If we take a longer period starting from 1990 and add other East African countries, a somewhat different picture emerges. Compared to Kenya, Tanzania and Rwanda, Uganda is the only country that did see a decline in its terms of trade from 145% in 1990 to 125% in 2020. The other three countries started this journey at a lower level, Kenya’s terms of trade index back then was 80% and Rwanda’s just 40%. But both countries have improved their position over 30 years, Kenya to lukewarm 106% and Rwanda to very impressive 180% whereas Tanzania reached 200% (starting from 107%).
The picture for low and low middle income countries of Southern and Eastern Africa is even more diverse. There are countries with a relatively high level of terms of trade (160 and above), such as Tanzania, Rwanda, Zambia, and countries with low terms of trade (120 and below), such as Uganda, Zimbabwe, Kenya, Malawi and Mozambique. Ethiopia is placed between these two groups with a score of 130.
Winners and losers
As far as the trend in terms of trade is concerned, one can distinguish three groups.
- Group 1 includes countries that show an increase in terms of trade: Rwanda, Tanzania, Zambia.
- Group 2 includes countries that show a decline in terms of trade: Mozambique, Malawi, Uganda.
- Group 3 includes countries that don’t show a significant change one way or another: Ethiopia, Kenya, Zimbabwe.
What is the difference between the “winners” and the “losers”? Can such an analysis elicit useful insights about development politics that lead to better trade position of developing countries? An improvement or increase in a country’s TOT generally indicates that export prices have gone up as import prices have either maintained or dropped. Conversely, export prices might have dropped but not as significantly as import prices. Export prices might remain steady while import prices have decreased or they might have simply increased at a faster pace than import prices. All these scenarios can result in an improved TOT. We will not discuss the possible impact of changes in export prices because such changes would have had the same impact across the countries under analysis due to the similar structure of exports.
As can be expected, there are many idiosyncratic factors that impacted each country’s terms of trade. For example, Zambia, whose commodity export is made up 80% of copper and other metals, benefitted from copper prices, which have been steadily growing over the past two decades from $0.84 to $4.08 per ton. Rwanda had the benefit of a very low base of 40, catching up and even surpassing Uganda eventually while increasing the share of metals in its export structure. Tanzania, also starting from a relatively low base, has diversified its exports from agricultural products to manufactures and – to a lesser extent – metals, reducing the share of agricultural products which now make up 60% of its exports.
On the other hand, Mozambique suffered from stagnating aluminum prices (raw aluminum accounts for about 30% of total commodity exports). These prices hovered around $1500 per ton between 2000 and 2020. Uganda, it appears, have never been able to recover from the precipitous drop in world coffee prices by more than 40% after the suspension of international quotas in October 1989. This shock was followed by an IMF Structural Adjustment Program with its emphasis on privatization, liberalization and abolition of monopoly of marketing boards as well as by the coffee wilt disease in the 1990s, which wiped out nearly half the country’s crop. The recent gold export boom has improved the situation somewhat but could not change the long-term trend. Malawi’s continued reliance on unprocessed agricultural exports (52% raw tobacco, 10% raw sugar and 9% tea) appears to confirm the Prebisch-Singer theory’s prediction that agricultural exports are doomed to lose in competition with industrial imports.
However, regression analysis of panel data for the nine countries reveals more general correlations. Some of them are expected while others are somewhat surprising. This regression analysis used four variables to test their significance for terms of trade. As expected, such an analysis returns positive correlations between terms of trade and three other variables: mineral rents, GDP per person employed and export diversification.
The strong positive correlation between ToT and mineral rents (i.e. the difference between the value of production for a stock of minerals at world prices and their total costs of production) is unsurprising. The higher is the rent, the better the terms of trade, particularly when international commodity prices are on the rise (which unfortunately is not always the case). The downside of relying on mineral trends is volatility of international prices as the example of Mozambique demonstrates.
GDP per person employed is a measure of the economy’s productivity. The more productive is an economy, the cheaper production and hence the exports can fetch higher prices on international markets. Again, productivity may differ significantly by sector and not affect export sectors. This is why the correlation is positive but weak. Agricultural productivity in the region has been characterized by slow growth (or even stagnation) and hence the countries relying predominantly on agricultural exports have seen their terms of trade deteriorating.
Export diversification measures, for each country, the degree of concentration of goods exported (it does not include services). It tells us if a large share of a country’s exports is accounted for by a small number of commodities or, on the contrary, if its exports are well distributed among many products. Countries whose exports are more diversified reduce the risks of a drop in export earnings because of the fluctuations in international prices and, therefore, expect to perform better.
Less expected is a weak negative correlation between ToT and the share of services as a percentage of total exports. This correlation is weak and statistically not very significant but it is somewhat counterintuitive because in modern economies services make up a sizeable share of exports (trade in services). This is also true for some countries in the region: for example, services account about 45% in Kenyan and Tanzanian exports. The likeliest explanation is that the bulk of the services exported by the countries in the region have low added value. African services exports are largely dominated by travel (42%). In comparison, high-income countries rely mostly on high-knowledge intensive services, such as financial, business, insurance or intellectual property services.
Conclusions
Obviously, a lasting improvement in terms of trade requires economic transformation of the regional economies. Increasing the share of manufactures, processed metals, high-knowledge intensive services and processed agricultural products is the way that countries with improving ToT have shown. In addition, regional economies need to work on diversification of exports and diversification of trading partners (both of which are highly concentrated, focusing in many cases on 2-3 export products and 2-3 major trade partners.
But what about the Prebisch-Singer theory and the secular decline in the terms of trade? Because of the long time horizon for the trend of shape, the is how long is the long term. Otherwise, it can be claimed that any period of time is not long enough and the findings reflects a temporary fluctuation from the trend rather than the trend itself. The theory cannot be proved (whether we’re talking about TFPF or a secular decline in the terms of trade of primary commodity exporters) until the very end (of the capitalist system, I assume).
However, this logic did not help Russian economist Nikolai Kondratieff who was the first to suggest a theory of long-term economic cycles experienced by capitalist economies and lasting about 40 to 60 years. His views were not welcomed in Stalin’s Russia because they suggested that capitalist nations were not on an inevitable path to destruction but, rather, that they experienced ups and downs. As a result, he ended up in a concentration camp in Siberia and was shot by a firing squad in 1938.
One is tempted to agree with Paul Bairoch who wrote that economic history teaches us that there exist no “laws” or rules in economics that would be valid for all historic periods or each different economic system. Or we simply can observe only a short period of history, which doesn’t allow us to understand economic laws and rules.