Abstract
This paper quantifies the consequences of long-term changes in the terms of trade on the growth rate of real income in Guyana, Jamaica, and Trinidad and Tobago. We find that magnitudes of both the long-term changes in the terms of trade and the responses of income growth rates vary considerably across these three countries. We then develop a model that links a change in the terms of trade to a change in income through the degree of sectoral reallocation of labor. Using this model, we analyze whether specific fiscal and trade policy responses cushion or amplify the impact of terms of trade on income. In a calibrated version of the model, we demonstrate that the degree with which an economy reallocates its labor across sectors is not necessarily a good indicator of its vulnerability to changes in its terms of trade.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
Certainly, sectoral labor reallocation is not the only channel through which changes in the terms of trade may be transmitted to income. Yet, while economists tend to think favorably about policies that facilitate the reallocation of labor across sectors in response to disturbances, the concrete ways in which small open economies reallocate labor has received relatively little attention. In this context, p. 394 Little et al. (1993) argue that, “the policies and general characteristics of an economy—especially its flexibility in responding to shocks—are more important than the size of shocks.”
To put these numbers in perspective, using a sample of 66 developing countries, Bidarkota and Crucini (2000) find that the standard deviation in the country quartile with the most volatile terms of trade was 25% per year. It was 16, 12.5, and 8.5%, respectively, in the next three quartiles in their country distribution.
Trinidad and Tobago has natural gas reserves and is the leading exporter of liquefied natural gas to the United States. Part of its fuel imports are liquefied for exporting.
We study the impact of long-run changes in the terms of trade. See Prebisch (1950) and Singer (1950) for the economic consequences of long-term permanent changes in the terms of trade in developing countries. In the data, terms-of-trade shocks are highly persistent (e.g., Powell 1991), and that is sufficient for our purposes. Mendoza (1995), and Dennis and İşcan (2006) study the short-term consequences of terms-of-trade fluctuations. Mendoza (1997), on the other hand, studies the link between terms-of-trade uncertainty and long-run economic growth.
The difference is called the “trading gain or loss,” \(T=\left(\frac{X-M}{P}\right)-\left(\frac{X}{P_{x}}-\frac{M}{P_{m}}\right)\), where X is exports, M is imports, P is the GDP deflator, P x is the export price index, and P m is the import price index. From the national accounts relations, we have:
$$ \text{GDP (constant price GDP) $+$ T (trading gain or loss) $=$ GDI (constant price GDI).} $$A more comprehensive measure of consumption possibilities would be gross national income, which includes net income receipts from abroad, and unilateral transfers. The analysis here uses the balance on trade from the balance of payments accounts, rather than the entire current account. The former is sufficient to identify the direct impact of changes in the terms of trade on domestic resource reallocation.
Kohli (2004) shows that the potential divergence between real GDP and GDI can be quantitatively large, adding up to more than 10% of GDP in less than two decades in several of the 26 countries he analyzes. Using Kohli’s (2004) accounting framework, Macdonald (2007) shows that differential changes in the terms of trade yield a significant wedge between income growth rates in Canada and the United States during the period 2000–2006.
Of course, this accounting framework is not intended to explain precisely how shares might change over time. Section 3 addresses this more fundamental question using a stylized model with explicit production and preferences. The accounting framework does not address the possible causal links running from the terms of trade and international trade to GDP growth, as well. However, given the similarities in terms of their income per capita, and given that their international trade patters are driven largely by their natural endowments, we think that comparing the terms-of-trade effects across these countries is informative.
While the decomposition is relatively straightforward to implement, it nevertheless requires data on individual components of aggregate demand, some of which are not immediately available. Specifically, for 2006 and 2007, in several instances, there were missing observations that were extrapolated from recent trends. See Appendix for details on the data sources and the construction of the variables.
Recall from Eq. 7 that the terms-of-trade effect depends on both relative price changes and changes in the share of import in GDP. For instance, a terms-of-trade effect that is smaller than the percentage change in the terms of trade is indicative of the ability of an economy to substitute away from imported goods.
So, a change in the relative price does not lead to gross substitution and complementarity in domestic private consumption demand. We have no direct evidence on the Caribbean economies we study that suggests otherwise.
It is possible to allow for government consumption of traded goods as well. However, this would complicate the analysis without adding useful insights. In addition, in practice, government expenditures are overwhelmingly concentrated in the nontraded goods sector.
For simplicity, tariff revenues are not part of the government budget constraint, although it is straightforward to include them in the revenue side of government finances. In the model, Tariff revenues may be thought of as financing direct imports by the government sector.
There is no explicit production of natural resources in the model. However, the model allows for an exogenous stream of income from abroad possibly representing royalties on the production and export of natural resources. Therefore, either through the export sector or through income from abroad, the model can capture key features of the Dutch disease phenomenon, such as the appreciation of the real exchange rate in response to an increase in the export price of natural resource. However, in the case of Trinidad and Tobago, since F is negative (due to interest payments on outstanding debt), we do not emphasize the latter channel.
We do not have direct empirical evidence on sectoral production function for the economies we consider. However, we conjecture that a unitary elasticity of substitution between labor and intermediate inputs (i.e., the Cobb-Douglas case) is less likely to be appropriate for the nontraded sector.
Given the long-run analysis we undertake in this paper, this is plausible. For a model of short-run intersectoral labor mobility, see Dennis and İşcan (2006).
In relation to the notation used to illustrate the income decomposition in the previous section, we have q = 1/P T , and \(h=P_{T}/P=\left( \eta ^{\eta }(1-\eta )^{1-\eta }\right) \left( P_{T}/P_{N}\right) ^{1-\eta }\), which is a function of the relative price of non-tradable goods, and hence a key component of the real exchange rate. Also, we take foreign prices—both prices of domestic imports and foreign nontraded goods—as given. So, here, an increase in P corresponds to a real exchange rate appreciation.
In the benchmark model, the ratio between the percentage change in the term of trade and the terms-of-trade effect is slightly higher than one (1.4). However, a value of μ = 0.1 delivers a ratio between the percentage change in the term of trade and the terms-of-trade effect close to unitary for a range of parameter values. These parameter combinations we considered include (ω = 0.42, β = 0.4, η = 0.5), with different benchmark values for \(A_{_{G}},A_{_{X}},F\). We also experimented with values of μ = 0.9. We don’t report the results of these sensitivity analysis to conserve space. In the model, there is no growth and foreign trade is always in balance, so GDP and the trade-balance effects from Section 2.2 are zero.
See, for instance, Corden (1984) for a survey of the Dutch disease literature. In our context, an improvement in the terms of trade is associated with a real exchange rate appreciation, and as we discuss below, triggers supply-side and demand-side responses. The responses we discuss overlap with those noted by the Dutch disease literature, which associated these responses with distinct income and substitution effects. Relative strengths of these income and substitution effects and their implications for labor reallocation are ultimately quantitative and empirical matters. Our simulation analysis provides a quantitative assessment of these responses in the case of Trinidad and Tobago. A complementary approach, pursued by Egert and Leonard (2008) in the case of Kazakhstan, is to assess these responses using an econometric model.
Our results depend on the specific parameter values we use in the numerical examples. The general point we stress, however, is the economically significant influence of policy variables on the transmission of changes in the terms of trade to income through sectoral reallocation of labor.
By contrast, when in the nontraded sector labor and imported inputs are strongly gross substitutes (the elasticity of substitution between labor and imported inputs is relatively high), nontraded producers can switch toward imported inputs with more ease, which translates into higher employment in the traded sector. Thus, the resource reallocation effect can dominate the income effect, resulting in a reallocation of labor toward the traded sector. For instance, in our simulations this happens when μ = 0.9.
References
Bidarkota P, Crucini MJ (2000) Commodity prices and the terms of trade. Rev Int Econ 8:647–666
Corden WM (1984) Booming sector and Dutch disease economics: survey and consolidation. Oxf Econ Pap 36:359–380
Dennis BN, İşcan TB (2005) On the terms of trade and sectoral reallocations. Rev Int Econ 13:892–903
Dennis BN, İşcan TB (2006) Terms of trade risk with partial labor mobility. J Int Econ 68:92–114
Diewert WE, Morrison CJ (1986) Adjusting output and productivity indexes for changes in the terms of trade. Econ J 96:659–679
Egert B, Leonard CS (2008) Dutch disease scare in Kazakhstan: is it real? Open Econ Rev 19:147–165
Gavin M (1990) Structural adjustment to a terms-of-trade disturbance: the role of relative prices. J Int Econ 28:217–243
Hope KR (1987) Development finance and the development process: a case study of selected Caribbean countries. Greenwood Press, New York
Kehoe TJ, Ruhl KJ (2007) Sudden stops, sectoral reallocations, and the real exchange rate. Unpublished working paper, University of Texas at Austin, Texas
Kohli U (2004) Real GDP, real domestic income, and terms-of-trade changes. J Int Econ 62:83–106
Little IMD, Cooper R, Corden WM, Rajapatirana S (1993) Boom, crisis, and adjustment: the macroeconomic experience of developing countries. Oxford University Press, New York
Macdonald R (2007) Canadian and U.S. real income growth pre and post 2000: a reversal of fortunes. Statistics Canada Micro-economic Analysis Research Paper Series No. 48, Ottawa
Mendoza EG (1995) The terms of trade, the real exchange rate, and economic fluctuations. Int Econ Rev 36:101–137
Mendoza EG (1997) Terms-of-trade uncertainty and economic growth. J Dev Econ 54:323–356
Mottley W (2008) Trinidad and Tobago industrial policy 1959–2008. Ian Randle, Kingston
Prebisch R (1950) The economic development of Latin America and its principle problems. United Nations, New York
Powell A (1991) Commodity and developing country terms of trade: what does the long run show? Econ J 101:1485–1496
Singer HW (1950) The distribution of gains between borrowing and investing countries. Am Econ Rev 40:473–485
Acknowledgements
I would like to thank an anonymous referee, the editor, A. Rebucci, and several conference and seminar participants for constructive comments and suggestions, Laura O’Hearn for excellent research assistance, and Natàlia Díaz-Insensé for editorial suggestions.
Author information
Authors and Affiliations
Corresponding author
Appendix: Data
Appendix: Data
Data for Kohli’s (2004) decomposition come from the Inter-American Development Bank (IADB) database, with specific accounts given in parentheses below.
-
GDP, current prices, in millions of domestic currency (from National Accounts)
-
GDP, constant prices, in millions of domestic currency (from National Accounts)
-
Total Exports: percentage of GDP (from External Accounts)
-
Total Imports: percentage of GDP (from External Accounts)
-
Domestic Absorption: percentage of GDP (from National Accounts)
-
Export Prices: Index (from Prices and Wages)
-
Import Prices: Index (from Prices and Wages).
Data for the calibration of Trinidad and Tobago are based on the following methodology and data sources.
-
G/Y: Government expenditures and nominal GDP in domestic currency units are from IMF, International Financial Statistics.
-
F/Y: Financial inflows as a percent of GDEP flows are from IADB, External Accounts database.
-
τ M : Tariff rates are from IADB database on tariffs.
-
ω: Share of labour in nontraded sector. We provided a benchmark value for this parameter by using the estimates of persons employed in the nontraded sector (in thousands) and value of imports of goods (imports in millions of USD). The data source underlying the employment shares is ILO, Laborsta database.
-
μ: Elasticity of substitution between labour and imports in the nontraded CES production function. We provide a benchmark estimate for this parameter based on relation between the change in (log) employment in nontraded and the change in (log) imports.
-
η: Expenditure share of traded good. We provide a benchmark estimate for this parameter by using export, import, and domestic production data for all available sectors. Specifically, we calculated the exports and imports for each sector. We deducted exports from and added the imports to the value of domestic production to obtain domestic consumption, and expressed the final figure as a percent of GDP. The data source underlying sectoral exports and imports is WTO, Trade Profiles, and sectoral output is World Bank, World Development Indicators.
Rights and permissions
About this article
Cite this article
İşcan, T.B. Changes in the Terms of Trade and Sectoral Reallocation of Labor: The Case of Guyana, Jamaica, and Trinidad and Tobago. Open Econ Rev 23, 473–500 (2012). https://doi.org/10.1007/s11079-011-9201-9
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11079-011-9201-9