Comments

Christian Broda:

Economists usually agree that high oil prices are a reason behind the high inflationary episodes in the late 1970s and early 1980s around the world. In the last few years we have seen oil prices return to similar levels (in real terms) to those experienced in previous decades, but inflation has remained muted. The paper by José De Gregorio, Oscar Landerretche, and Christopher Neilson guides our thinking about the potential reasons for this muted response of the consumer price index.

In this brief discussion I focus on the statistical significance of the stylized fact uncovered in the paper (that is, that the pass-through from oil prices to consumer prices has decreased over time) and provide a simple back-of-the-envelope calculation to quantify how much of the decline in the influence of oil prices is due to the reduction in oil intensities around the world and how much is coming from a different interaction of oil prices and exchange rates in recent years, as opposed to the situation in the 1970s and 1980s.

While the paper documents the decline in pass-throughs over time, it provides no test of whether the decline is statistically significant. To determine the statistical significance of the results, I use the data on the United States and the United Kingdom kindly provided by the authors to replicate their results.

Table 2 shows pass-through coefficients calculated in the same way as was done in equation 2 of the paper for two different periods (the same breakpoints as in the paper are used). The new information of table 2 is that it includes standard errors, test statistics, significance levels, and confidence intervals for the nonlinear combination of parameters in equation 2. Calculations are based on the delta method.

The table shows that in the case of the United States the decline in pass-through coefficients is similar to that reported in appendix table A5. For the period before the break (1981Q4 [fourth quarter]), the pass-through coefficient is close to 0.07 and is significantly different from zero. For the post-break period, the coefficient falls to 0.04.¹ However, the test of equality over time is not rejected at standard levels of confidence. That is, in the case of the United States, the pass-through coefficient has not significantly decreased over time.

Table 2.

Pass-Through Coefficients for the United States and the United Kingdom^a

The case for the United Kingdom is similar in that the pass-through coefficient decreased from 0.14 to almost 0.04 and that this decrease is not statistically significant.² While this should not be interpreted as conclusive evidence that the decline in the pass-through is not significant in the entire sample (after all, this example is only two countries out of a sample of more than 30), it is suggestive that further evidence is needed to assert that this is a strong fact. More powerful pooled tests can potentially provide this evidence.

Independently of their statistical significance, pass-through coefficients have been declining in most countries. I use the main results in the paper to assess the quantitative contribution of the two main explanations behind the decline in pass-through coefficients. I will first focus on the link between oil prices expressed in U.S. dollars, oil prices expressed in local currency, and consumer prices. The average pass-through coefficient between oil prices in dollars and consumer prices in the period pre-1975 is around 0.13 and declines to 0.03 after this period (figure 2). This implies a gap of about 0.10. This gap can be partially decomposed into two parts: the impact that changes in oil prices in U.S. dollars have on the price of oil in local currency and the changes in the pass-through coefficient between the local currency price of oil and consumer prices. Figure 5 suggests that the majority of the gap is due to the differences over time in the pass-through between oil prices in local currency and consumer prices. The...