Output Volatility and Openness to Trade: A Reassessment
In lieu of an abstract, here is a brief excerpt of the content:

Output Volatility and Openness to Trade:
A Reassessment

Output volatility has been shown to be negatively correlated with economic growth across countries.1 Recent studies document that the correlation is not only robust to alternative samples and estimation techniques, but that the direction of causality goes from volatility to growth.2 Given the robustness and the policy relevance of these results, it is surprising that so few attempts have been made to identify the causes of output volatility to date.3 This paper bridges that gap by studying the empirical determinants of volatility, with a particular focus on the role of openness to commercial trade.4

Output volatility naturally relates to the frequency and size of the shocks that affect an economy and to the manner in which the economy handles the shocks.5 Openness to trade is thus commonly associated with greater output volatility: presumably, the more exposed to trade a country is, the more [End Page 105] vulnerable it is to shocks coming from abroad.6 Nevertheless, economists believe that trade openness promotes economic growth.7 The combination of these results has led some observers to identify a general consensus on the interrelationship between openness to trade, output volatility, and growth. As Kose, Prasad, and Terrones state, “While there appears to be a general consensus that openness to trade flows stimulates domestic growth, it is also the case that such openness increases the vulnerability to external shocks.”8 To the extent that external and internal shocks are not negatively correlated, a greater vulnerability to external shocks implies more output volatility. If openness to trade increases output volatility and growth, but output volatility hurts growth, then either the direct effect of trade on growth outweighs the indirect effect, or there is something wrong with one of the presumed links. In this paper I present new evidence that the latter is likely to be the case. In particular, I show in a single cross-section of seventy-seven countries—twenty-one of which are members of the Organization for Economic Cooperation and Development (OECD)—that the effect of trade openness on output volatility is negative rather than positive.

In the paper, I do not deal with the direct link between trade openness and growth, and to the extent that I touch on the link between output volatility and growth, I rely on research by Hnatkovska and Loayza, who use the same dataset.9 Exploring these issues in depth is beyond the scope of this paper, however. Here, I present new evidence that points toward a negative causal link between trade openness and output volatility. This new result is consistent with research showing that openness to trade reduces vulnerability to some forms of financial crises and that openness to trade smooths the adjustment in the aftermath of external shocks.10 On both accounts, openness to [End Page 106] trade might reduce output volatility. This effect counteracts the effect that goes from trade openness to exposure to trade-transmitted volatility in world goods markets (that is, terms-of-trade shocks). This is also consistent with Calderón, Loayza, and Schmidt-Hebbel, who find evidence that trade openness provides the means for the domestic economy to diversify away some external sources of risk.11

The previous empirical attempts that either directly or indirectly assess the impact of openness to trade on income volatility do not test whether trade exerts an independent effect on volatility once the terms-of-trade risk is taken into account. I do so by introducing in all the regressions the de facto trade openness variable (the trade-to-GDP ratio), along with an interacted variable to account for the possibility that more open economies are naturally more prone to terms-of-trade risk. The underlying hypothesis is that, to the extent that the latter effectively controls for that risk, any other effect of trade on volatility should manifest itself through the point estimate of the openness coefficient.

Another relevant issue that is largely ignored in the related literature is the probable endogeneity of trade in this setting. If trade is endogenous to output levels (because, for example, richer countries tend to liberalize trade barriers, as...