Testing the Balassa Hypothesis in Low- and Middle-Income Countries

Abstract

This study analyses the long-run relationship between economic growth and real exchange rate for a group of 15 low- and middle-income countries for the period 1950–2011. Co-integration between growth and exchange rate is established by means of an augmented pooled mean group estimation method (which controls for heterogeneity and cross-sectional dependence ). Unlike previous studies, cross-sectional dependence is accounted for which implies that the productivity effect of the Balassa term is expected to be estimated consistently and without bias. Moreover, our results indicate that the effect of the Balassa term depends more on the income group (level of per capita income) than the rate of economic growth. In general, the power of the effect is stronger for higher income countries in the long run. The study clearly indicates that the Balassa hypothesis holds for middle-income countries, while this is not the case for low-income countries. However, fiscal policy and exchange rate volatility rather clearly explain the variations in the real exchange rate.

Appendices

Annexure 1

1.1 1.1 Model Derivation (Scott Hacker’s Contribution)

Suppose that the growth rate of the real exchange rate is defined as a function of productivity differential between the non-tradable and tradable sectors as in:

$$ \hat{Q} = {\delta }\left( {\left[ {\left( {\frac{\beta }{\alpha }} \right)\hat{A}_{T} - \hat{A}_{N} } \right] - \left( {\frac{\beta }{\alpha }} \right)^{*} \hat{A}_{T}^{*} - \hat{A}_{N}^{*} } \right) $$

(1)

with $\hat{Q} \equiv \hat{p} - \hat{p}^{*}$.

This is the same as:

$$ \hat{Q} = {\delta }\left( {\left[ {\left( {\frac{\beta }{\alpha }} \right)\hat{A}_{T} - A_{T}^{*} } \right] - \hat{A}_{N} - \hat{A}_{N}^{*} } \right) $$

(1.1)

If we let $\hat{M}_{0} = - {\delta }\left( {\hat{A}_{N} - \hat{A}_{N}^{*} } \right)$ and $m_{1} = {\delta }\left( {\frac{\beta }{\alpha }} \right),$ then:

$$ \hat{Q} = \hat{A}_{0} + m_{1} \left( {\hat{A}_{T} - \hat{A}_{T}^{*} } \right) $$

(1.2)

In levels form, this is equivalent to:

$$ Q = M_{0} \left( {A_{T} /A_{T}^{*} } \right)^{{m_{1} }} $$

(1.3)

and in log-levels, it is

$$ q = m_{0} + m_{1} ln\left( {A_{T} /A_{T}^{*} } \right) $$

(1.4)

where $q \equiv \, \ln Q\,and\,m_{0} \equiv \, \ln M_{0}$

We proxy $A_{T} /A_{T}^{*}$ with $Y/Y^{*}$ where Y is the home real GDP per capita and Y* is the foreign (US) real GDP per capita, so we get Eq. (2.23).

$$ q = m_{0} + m_{1} (lnY/Y^{*} ) $$

(1.5)

Annexure 2

See Tables 12.5, 12.6, 12.7 , 12.8, 12.9, and 12.10

Table 12.5 Descriptive statistics (all countries)

Table 12.6 Descriptive statistics (by country)

Table 12.7 Cross-sectional dependence tests

Table 12.8 Panel unit root tests (IPS and Fisher-type tests)

Table 12.9 Results of co-integration tests

Table 12.10 Model selection criteria

Annexure 3

See Tables 12.11 and 12.12

Table 12.11 Panel co-integration estimation: pooled mean group estimator (MICs = 5 countries)

Table 12.12 Panel co-integration estimation: pooled mean group estimator (LICs = 10 countries)