Globalization has changed inflation dynamics, but domestic developments still matter

A worker checks avocados at a packaging warehouse of Hoja Redonda plantation in Chincha, Peru, September 3, 2015. The eighth World Avocado Congress, an event dedicated to production, export and marketing of Hass avocados, will be held in Peru from September 13 to September 18. Peru is the second largest exporter of Hass avocados in the world, according to local media. REUTERS/Mariana Bazo - GF10000192546
Editor's note:

This blog post summarizes the findings of the paper, “Inflation dynamics: Dead, dormant, or determined abroad?” written by Kristin J. Forbes and part of the Fall 2019 edition of the Brookings Papers on Economic Activity. Read summaries of all six papers from the journal here.

In the Fall 2019 Brookings Papers on Economic Activity paper, “Inflation dynamics: Dead, dormant, or determined abroad?” Kristin J. Forbes of the MIT-Sloan School of Management tested whether growing globalization has played a role in inflation puzzles over the last decade. The issue is taking on increased urgency as central banks evaluate their ability to continue loose monetary policies in the presence of extremely tight labor markets. If inflation is largely determined abroad, a central bank could be less concerned about inflation exceeding its target and be more able to pursue a “high-pressure” economy that prioritizes job creation. If inflation is largely determined globally, however, central banks may also have more limited ability to stabilize inflation in the future. In the extreme, global factors that increase inflationary pressures–such as trade wars that cause companies to move away from global supply chains–could put more pressure on inflation than widely expected.

The core of the many models utilized to understand inflation usually focus on domestic developments, including variables for domestic slack, inflationary expectations, lagged inflation and import prices (as a control for all international developments). After the recession, inflation was persistently higher than predicted in these models, while more recently it has been lower. But Forbes highlights four areas of globalization which are not captured in this framework and could help explain these puzzles. Global variables that could also affect inflation are: increased trade flows, greater use of supply chains to optimize production costs, greater role of emerging markets and their impact on commodities, and a reduction in the bargaining power of workers. The author incorporates indicators for these metrics to model core inflation, CPI inflation, producer price inflation (PPI), and wage inflation.

Forbes’s research suggests that a single, common principle component may account for a large portion of the variance in PPI and CPI inflation, but the role of a shared global variable is noticeably smaller for wage and core inflation. This suggests that the dynamics of PPI and CPI inflation are more determined by global developments, but the dynamics of wage and core inflation more by domestic factors.

To better understand what is driving these patterns, Forbes shifts to more formal analysis, using models based on the Phillips Curve and a less structured trend-cycle decomposition. Analyzing a sample of 31 countries from 1995 to 2017, Forbes finds that CPI inflation can be better explained by adding additional controls for global factors (in addition to just import prices).

Furthermore, as shown in figure 2a, these additional controls for global factors have become more significant in inflation dynamics over the last decade, while the relationship between domestic slack and CPI inflation has become weaker when compared with a pre-crisis window of 1995-2007. The biggest difference in the models arises when trying to explain inflation during the global financial crisis, at which time adding global variables reduces the prediction errors from 1.51 percentage points to 0.90. And over the last decade, the model with global variables has smaller errors mainly due to the increased role of global commodity prices (and to a lesser extent global supply chains). Forbes notes that the global variables used together, and not individually, are responsible for the explanatory power of the model.

Figure 2a. Gap between actual and predicted CPI inflation

In contrast to these results of a more important role of global variables for CPI inflation, there is less evidence that global variables have become more important for explaining core inflation or wage inflation. Global variables can still play a role, but they are generally less important, and their role has not increased over the last decade. These measures of price pressures are still largely determined by domestic slack and other local variables, although there also remains a significant portion that cannot be explained.

The full set of results in the paper suggest that simple frameworks for understanding inflation dynamics are not “dead”, and even though inflation has been “dormant” recently, some of the puzzling patterns in CPI inflation can be explained by CPI inflation being more “determined abroad”. This does not mean, however, that there is no longer a role for central banks or domestic developments in inflation dynamics. Even though CPI inflation is increasingly affected by globalization, and most inflation measures move less tightly with domestic slack, domestic variables are still important determinants of inflation dynamics.