One the most fruitful demand-oriented approaches to the study of economic growth is the balance-of-payments (BOP) constrained growth theory developed by Thirlwall (1979) based on the Harrod foreign trade multiplier. This theory, which analyses economic growth in open economies, basically states that since relative prices do not play an important role in international trade in the long run and the BOP must be in equilibrium, then a country’s growth rate is determined by other nations’ growth and by its export and import income elasticities. Broadly speaking this model and its extensions have shown a good performance forecasting long-run growth rates (see McCombie, 2011, Setterfield, 2011, and Thirlwall, 2011, or Soukiazis and Cerqueira, 2012, for a review). However, evidence of its remarkable implications for short-run dynamics is much scarcer, putting aside the augmented versions of the original model that incorporate capital inflows to build a less-constrained version of the model, as seen in Thirlwall and Hussain (1982) or Meyrelles Filho et al. (2013).
Deviations from long-run growth rates are implicitly considered in most of these works, but business cycles are not explicitly investigated. To our knowledge, Soukiazis et al., 2012, Soukiazis et al., 2013, Soukiazis et al., 2014, Soukiazis et al., 2018 in their different explorations of what they call the twin-deficit approach, Garcimartin et al. (2016) in their analysis of business cycles in Portugal and Spain, Dávila-Fernández and Sordi, 2019a, Dávila-Fernández and Sordi, 2019b in their application of Goodwin’s distributive cycle model (Goodwin, 1967) to the BOP-constrained model, and Nishi (2019) in the consideration of the stability and nature of the transitional dynamics, are the main exceptions. In particular and apart from sample differences, Soukiazis et al., 2012, Soukiazis et al., 2013, Soukiazis et al., 2014, Soukiazis et al., 2018 analyze the joint effect of internal and external imbalances on growth rates, while Dávila-Fernández and Sordi, 2019a, Dávila-Fernández and Sordi, 2019b and Garcimartin et al. (2016) investigate the potential determinants of business cycles in the BOP framework (capital accumulation and initial conditions with capacity constraints in the former and trade shocks and capital inflows in the latter). In contrast, the main contribution of this paper is not related to the factors that can generate temporary imbalances, but to the adjustment process towards equilibrium and how this process affects future growth rates. Although this issue has received much less attention in the literature, to investigate if the BOP constraint governs not only long-run but also short and mid-run dynamics has, in our view, remarkable implications for economic policy and forecasting.
In particular, the primary goal of this study is to contribute to the analysis of the dynamics of the BOP-constrained growth model by investigating its capacity to carry out unconditional forecasting of economic growth rates.1 Namely, based on a stylized BOP-constrained growth model with non-zero net foreign debt, we forecast future growth rates of the GDP by decomposing past growth rates into a long-run growth component consistent with a balanced current account and a cyclical growth term caused by net capital inflows. Whenever the cyclical component, driven by current account deficits, is substantial, its reversion towards equilibrium is expected, thus conditioning future growth rates. We will try to show that the proposed decomposition is informative, and that the derived components are significant in predicting future economic growth rates within and beyond a decade.
This is in sharp contrast with the predictions of forward-looking rational expectations models, where current account deficits (behind our cyclical component) emerge because of faster future growth expectations. Current account deficits are driven by consumption smoothing, thus implying a positive correlation with future economic growth rates, if any. It has been argued that the persistence of current account deficits is proof of well-functioning markets for contingent securities (see Obstfeld, 2011, for an exposition and criticism of this argumentation), but the results of our study contradict this forward-looking rational expectations model’s view.
In order to reveal how the BOP-constrained growth model compares with other (more supply-side-linked) predictors used regularly in empirical growth equations, we make also a methodological contribution avoiding the need to use pre-estimation-based values in the BOP-constrained growth forecasting framework. This allows for an explicit evaluation of the significance of the discussed demand-side and supply-side variables in growth equations. To our knowledge, this evaluation has not been done before in the BOP-constrained growth literature.
The rest of the paper is organized as follows. In section 2 we decompose growth rates into a long-run and a cyclical component, showing that both are proportional to directly observed variables/terms that can be used as proxies in the estimations of predictive equations. In section 3 we present the main results of this investigation. According to our findings, both the long-run and the cyclical components are significant in forecasting future economic growth rates. The former dominates in predicting the average growth rates in shorter periods, whereas the latter prevails in longer periods. In other words, the imbalances accumulated through the cyclical component can take some years, but in the mid-term current account deficits tend to adjust, thus conditioning output dynamics. Additionally, we control for some variables usually employed in the neoclassical convergence literature, showing that the BOP-approach factors retain their significance irrespective of the inclusion of supply-side factors. Our main conclusions are considered in section 4.
KVEDARAS Virmantas;
GARCIMARTÍN Carlos;
ASTUDILLO Jhonatan;
2020-05-29
ELSEVIER SCIENCE BV
JRC117605
0264-9993 (online),
https://www.sciencedirect.com/science/article/pii/S0264999319307370?via%3Dihub,
https://publications.jrc.ec.europa.eu/repository/handle/JRC117605,
10.1016/j.econmod.2019.10.026 (online),
