RDP 9705: The Response of the Current Account to Terms of Trade Shocks: A Panel-data Study 5. Conclusion

I consider two alternative explanations for the findings before providing a summary of the main results.

5.1 Tornell and Lane – Voracity Effect

Tornell and Lane (1996a, 1996b) present a model of competing fiscal claimants which under certain circumstances implies that a large positive shock to the terms of trade will lead to an increase in government expenditure sufficient to reduce the current account balance (the argument is symmetric for negative shocks). This ‘voracity effect’ is an alternative explanation of the finding of a negative relationship between the terms of trade and the current account for some countries.

To control for this potential effect, I re-estimated the panel regression of Equation (10) with the government fiscal balance (as a per cent to GDP) included as a right-hand-side variable:

where: ΔGVT is the demeaned first difference of the government fiscal balance (as a per cent of GDP); see Appendix B for more details. The government fiscal balance will have a direct effect on the current account balance according to the following identity:

where, S and I are private savings and investment, respectively, T is government revenue, and G is government expenditure. If private savings and investment decisions were independent of the government fiscal position, then φ0 = 1 (and φ1 = 0) in Equation (10′). The lag of ΔGVT in Equation (10′) accounts for the problem of government statistics being reported on a fiscal year basis which is often different from calendar years.[34]

An indirect effect of changes in the government fiscal position will exist if private savings and investment adjust to changes in the government fiscal position (this possibility is ignored by Tornell and Lane). Government expenditure may crowd out some private investment. An increase in the government deficit may lead to increased private savings in order to pay for expected future tax increases. These indirect effects work in the opposite direction to the direct effect of the government fiscal balance on the current account. If there is full crowding out and agents are perfectly Ricardian in their behaviour, then φ0 = 0, and φ1 = 0 (if agents are slow to adjust, then φ0 + φ1 = 0).

I ignored the possible endogeneity problem by asserting that any response of the government fiscal balance to current account ‘imbalances’ was likely to be delayed in all but crisis situations. Such episodes were not likely to be the dominant effect across a longer time period and a large sample of countries, as used in this paper.[35]

As the results in Table A1 (in Appendix A) show, there was a negative correlation between the terms of trade and the current account for countries with persistent terms of trade even after controlling for the government fiscal balance. Therefore, the voracity effect is not a convincing alternative explanation of the results of this paper. The effect of the government fiscal balance had the expected sign (although it was significant only for the country group with less persistent terms of trade shocks). The sum of coefficients on government fiscal balance was closer to zero than to one, which suggests a strong indirect effect (that is, private agents offsetting the direct effect of a change in the government fiscal position).

5.2 Credit Constraints

If consumers in a country faced significant credit constraints, the consumption-smoothing effect would be dampened in the case of negative terms of trade shocks. For such countries it is conceivable that the investment effect will dominate the consumption-smoothing effect.[36] However, I argue that the two country groupings (based on terms of trade persistence) face similar degrees of credit constraints for consumers and, therefore, it is not likely that credit constraints are driving the results that I have shown in Section 4.

I used the level of real GDP per capita of a country to proxy the existence of credit constrained consumers. Table 6 below shows that the two country groups have very similar distributions of real GDP per capita (see Appendix B for the exact description of the data). The significance of possible credit constraints seems constant across both high and low terms of trade persistence country groups.

Table 6: Distribution of Real GDP per Capita
1990 US dollars
Average across group(a) Min. within group Max. within group No. in
< 1,000
No. in
< 4,000
No. in group >10,000
11 countries with highest terms of trade persistence 6,236 485 15,631 2 6 3
10 countries with lowest terms of trade persistence 6,424 497 10,959 1 5 4
Note: (a) Indicates country averages taken over the period 1960 to 1992.

5.3 Final Remarks

Theory implies that the response of the current account balance to shocks depends on the degree of persistence of these shocks. The consumption-smoothing effect and the investment effect work in opposite directions. The greater the persistence of a shock, the more the investment effect will dominate. Therefore, in the years following a persistent shock, the current account will move in the opposite direction to the shock. The opposite is true for more temporary shocks. Terms of trade shocks were used to test this theory because they display a wide range of persistence across different countries. This paper has shown strong evidence that the persistence of shocks is an important determinant of the current account in accordance with this simple theory.

An episodic approach using a two-year window to examine large terms of trade changes showed that many countries experience both positive and negative correlations between the terms of trade and the current account. This is not surprising given that it is likely that shocks were neither entirely temporary nor entirely permanent. A stronger finding was developed by using two subsets of countries that lay at extreme ends of the spectrum of terms of trade persistence. Countries with the lowest terms of trade persistence had more episodes of positive correlation and less of negative correlation between the current account and the terms of trade than countries with the highest terms of trade persistence.

Panel-data regressions showed that the current account response is positively related to the terms of trade for countries with predominantly temporary terms of trade shocks and negatively related for countries with predominantly permanent terms of trade shocks. Moreover, the response of the current account for these two smaller country groups was shown to be significantly different from the response estimated across the full sample of countries.

These results were robust to the inclusion of changes in the government fiscal balance to control for possible ‘voracity effects’ as in Tornell and Lane (1996a, 1996b). Also, the findings do not appear to be related to the existence of credit constraints because both of the country groupings based on terms of trade persistence have very similar distributions of per capita income (across countries).


Other lags of the government fiscal balance variable were not included in the regression because they were never significant. [34]

Endogeneity is not an issue in cases where the government deficit is so high as to imply a balance of payments crisis and, therefore, is reduced either because of the crisis or to avoid a potential crisis. [35]

The existence of credit constrained consumers does not imply that large firms are also credit constrained, so the response of investment will still be the same. For example, many investment projects could be funded by foreign direct investment. [36]