On the Transmission Mechanism of Monetary Constraints to the Real Side of the Economy

Abstract

Contrary to the predictions of the theory underlying international finance, inflows of capital triggered by financial liberalisation have neither equalised real interest rates nor increased income growth in many emerging economies. We explain this puzzle by developing a model that combines the balance‐of‐payments constraint approach to economic growth with a less stringent version of the real interest rate parity hypothesis. The model’s foundations are based on robust empirical findings or well‐established macroeconomic models. We show that a perverse combination of income elasticities of demand for imports and exports generates slow income growth and high real interest rates. As domestic income grows and imports rise faster than exports, the real exchange rate is expected to depreciate in order to clear the balance of payments (or the foreign exchange rate market). An incipient capital outflow arises and interest rates increase. Faster adjustment in capital rather than in the goods market therefore generates a higher real interest rate differential between the domestic small open‐economy and the rest of the world. The long run analysis shows that a constant degree of risk aversion implies a positive equilibrium real interest rate differential that affects economic growth. A permanent increase in default risk driven by persistent current account imbalances thus impacts on long run growth. The model’s results are illustrated with evidence from the three major Latin America economies: Argentina, Brazil and Mexico.

Key Words:

• Real interest rate differentials
• balance of payments constraint

Jel Classification:

• F43
• F15
• E43

Acknowledgements

I would like to thank Tony Thirlwall for invaluable insights and comments on an earlier version of this paper. I am also grateful to the Editor and two anonymous referees for important remarks. I finally acknowledge CNPq of Brazil for financial support. The usual disclaimers apply.

Notes

1. By equilibrium in the BoP we mean that the trade account is in balance as in Thirlwall’s (1979) original formulation. We are aware of the modifications in this initial specification. For example, Thirlwall & Hussain (1982) relax the constraint on demand through the inclusion of capital inflows in the (dis)equilibrium condition. We, however, cast some doubt on the sustainability of a capital account surplus, at least in the case of emerging economies. For example, Ferreira & Canuto (2003) present a model in which the day of reckoning comes and inflows are reversed into outflows through the payment of the principal in addition to profits, dividends and interest expenses on the outstanding debt. The paper shows that the model based on current account equilibrium (trade balance plus interest, dividends and profits) is able to fit the Brazilian experience from the 1950s to 1990s. It also demonstrates that capital inflows were extremely volatile during the aforementioned period. In spite of the higher predictive power of Ferreira and Canuto’s (2003) model, the simpler specification of Thirlwall (1979) has greater analytical advantage owing to its simplicity. For this reason, our analysis uses the trade balance as the overall condition for BoP equilibrium. In this paper as in Thirlwall’s (1979), the trade and current account balance are interchangeably used to broadly refer to the BoP equilibrium. Nonetheless, we recognise that the incorporation of the dynamic effects of capital inflows on the current account and, consequentially, on the constraint on demand can be a valid modification of the model that we further propose.

2. This stylised fact was later recognised by Krugman (1989). He noticed a strong statistical correspondence between an economy’s growth and the income elasticities of demand for imports and exports. However, his presumption was that income elasticities were not parameters, but the variables that changed in order to guarantee equilibrium in the BoP.

3. Authors, such as Wu (2005), have also found evidence of the correlation between export and income growth as predicted by the BoP model, however, the findings do not point to structural elasticities and, hence, do not lend support to the BoP constraint approach to economic growth.

4. The BoP constraint literature, which develops from Thirlwall (1979), is not built from microeconomic foundations. However, small open‐economy models based on utility maximisation show that a current account deficit in an initial period needs to be financed through a decrease in aggregate demand in a subsequent period, if any intertemporal budget constraint is to be respected (see, for example, Obstfeld & Rogoff, 1996; Blanchard & Fischer, 1989). The assumption of these models is that a country benefits from the intertemporal allocation of resources if the utility they obtain from consumption in the present is higher than consumption in the future, for a given subjective discount rate. The main difference is that relative prices rather than income growth correct imbalances in the BoP.

5. This is also known as the Fisher (1930) open‐economy’s hypothesis.

6. The implicit assumption that we made in this step is that t + 1 = t + k, hence, k = 1.

7. McCombie & Thirlwall (1994) justify this assumption by imperfect competition, instead of PPP. For a review on recent tests of the real exchange rate see Taylor (2001) and Sarno et al. (2004).