International Inflation Dynamics and the New Keynesian Phillips Curve :
The Role of the Global Output Gap
1. Introduction
In the wake of the recent global financial crises, the behavior inflation has surprised many observers. Despite persistently weak growth and extraordinarily accommodative monetary policy on the one hand, and significant increases in global commodity prices on the other, worldwide inflation rates have remained remarkably stable, especially in advanced economies. Accompanying low and subdued inflation, there has also been an observed increase in the degree of inflation rate comovements worldwide.
The underlying forces responsible for such changes in inflation dynamics has been subject to much debate. Some attribute low and stable inflation to improved monetary policy, particularly the widespread adoption of inflation targeting in the 1990s that has helped to anchor long-run inflation expectations. However, given that such effects are likely to work primarily through lowering long-run inflation trends, it cannot explain the rise in the synchronicity of inflation rates over the short to medium term.
Another view that has gained prominence is based on the `good-luck' hypothesis, which attributes globalization with helping to mute inflationary pressures around the world through a series of favorable external shocks. For example, the rapid integration of lower cost economies into the global trading system is seen to have depressed international goods prices. This acts to moderate inflation directly by lowering import prices, as well as indirectly through enhanced price competition that restrain markups and producer prices in relevant industries. To the extent that greater integration of markets increases cross-border spillovers, it may also help account for observed co-movements of inflation rates across countries.
Consistent with the `good-luck' notion is the globalization hypothesis (GH), which orginated from the concerns of some policymakers. According to the GH, the internationalization of goods and financial markets should have altered the sensitivity of inflation to its key driving variables. In the Phillips curve relation, short-run inflation dynamics is driven primarily by the domestic output gap, which measures the degree of economic slack or resource utilization pressures at the country level. The GH applied to the Phillips curve thus implies that as markets become increasingly integrated, inflation dynamics should have become more sensitive to global rather than domestic measures of slack. Should this occurence take place, the observed synchronicity of inflation rates across countries may be attributed to the influence of the global output gap acting as a common driving variable for inflation.
According to estimates of closed economy Phillips curves, a number of studies have reported
a decline in the link between inflation and domestic measures of slack during recent decades. However, whether the role of domestic slack in the Phillips curve relation is supplanted by a global output gap remains a topic that is still subject to considerable debate. On the one hand, a widely cited study by Borio and Filardo shows that measures of the global output gap can add considerable explanatory power to reduced form Phillip curve equations for 17 OECD countries. They also show that the influence of the global output gap for national inflation rates have been increasing over time. However, Ihrig et al. (2010), among others, argue that these results are not robust to alternative specifications of the Phillips curve.
In particular, they show that the positive results of Borio and Filardo can be overturned once a more plausible specification for inflation expectations is taken into account.
This paper joins the existing literature and examines the GH for 17 advanced and emerging market countries. A principal component analysis (PCA) is employed to first examine whether globalization matters for inflation at all, by investigating whether there exists a common component responsible for the overall movements in international inflation rates. Then, the influence of the global output gap for domestic inflation is investigated within a New Keynesian Phillips curve (NKPC) framework. The response of inflation to global slack conditions is studied over different time dimensions and is also investigated in relation to a country's degree of openness in trade. Finally, the external influences of import and oil prices are allowed to enter the NKPC specification to explore whether the global output gap is still relevant for inflation beyond the import price channel.
The contribution of this paper are along at least two dimensions. First, a main shortcoming of past work is their limited focus on advanced countries. This paper extends the analysis to cover a broader number of countries, including emerging Asia Pacific economies. In doing so, it reduces potential specification problems in previous work where the measure of the global slack used in the Phillips curve estimation is too narrow because it does not account for resource utilization pressures from emerging market countries. Second, to limit potential model misspecification issues, the GH is explored by building upon the unobserved components (UC) modeling approach for the NKPC as outlined in Kim et al.
In the UC model, long-horizon inflation expectations as well as the output gap can be treated as unobserved state variables, which are estimated from the observed data. Therefore, compared to existing methods, less restrictive assumptions are needed when exploring the GH with the UC approach, as the empirical model `allows the data to speak' as much as possible.
As a preview of the empirical results, this paper finds strong evidence in support of the GH. Based on the PCA results, a sizeable common component drives worldwide inflation rates, and accounts for over half of the variation in its movements. The importance of this common factor increases over time, reaching a level of over 90% after the year 2000. The NKPC estimations for each country suggest that this common driving factor is most likely explained by the global output gap, as it enters the NKPC specification as a statistically significant explanatory variable for all countries. Furthemore, the link between inflation and the global output gap becomes more prominent since the year 2000, and is positively related to a country's degree of openness in trade. Finally, import and oil prices generally do not help account for the impact of global influences onto domestic inflation rates, suggesting that the global output gap is capturing the effects of global supply and demand pressures for inflation beyond the import price channel.
This paper is organized as follows. Section 2 discusses the channels through which globalization may have altered the behavior of inflation and explores via a PCA whether there is a common driving factor for international inflation rates. Section 3 outlines the empirical UC model based on the NKPC and Section 4 presents the empirical findings. Section 5 discusses the implications of the estimation results for monetary policy. Section 6 concludes.