Public interest in inflation has certainly not peaked.
As economists debate whether inflation has peaked, among the public, Internet searches about inflation have continued to climb.
P eople who haven’t noticed the return of inflation are, like baby formula, growing scarce in America. According to a recent poll, 70 percent of Americans now regard inflation as “a very big problem” facing the country. Economists, for their part, are interrogating the data. The newest numbers are now prompting them to ask: Has inflation peaked?
Answers vary based on how you interpret the economic data. In the latest monthly trends, a murmur of a “yes” avails itself. In the latest release of consumer price data, for the month of April, at 8.3 percent, the year-over-year rate of inflation represented a decrease over March, when the barometer came in at 8.5 percent. Scrutiny of individual goods that comprise the consumer price index, however, tell a different story. A decline in gas prices that contributed to the deceleration in April, for instance, has since more than reversed.
Public interest in inflation has certainly not peaked. The chart above shows the Google Trend that tracks the frequency of searches for “inflation” along with the year-over-year change in the overall consumer price index. As you can see on the chart, if the typical relationship between the search data and inflation holds, then the release of May’s consumer price data on Friday will reveal no evidence of a peak in March.
To expect Google’s data on searches for “inflation” to confess at least some accurate information about forthcoming official consumer price data is only to expect it to behave now as it has in the past. Between January 2011 and April 2022, the monthly value of the Google Trend for “inflation” registered a correlation of 0.72 with the year-over-year change in that month’s overall consumer price index. For context, a correlation of between 0.7 and 0.8 puts the predictive relationship between the two on par with the predictive relationship between Body Mass Index and body-fat percentage. Having increased from 82 in April to 100 in May, if the previous data are any indication, the Google Trend now foreshadows that Friday’s release of official inflation data for May will show an uptick, perhaps around 9 percent year-over-year.
Google search data wasn’t available to those who lived through the inflationary eras of America’s past. And generating inferences about macroeconomic outcomes like inflation based on search engine data may, to some, seem like a stretch. But grounds for optimism about its utility as a new tool for tracking inflation do exist. In the early days of the Covid-19 pandemic, Google search data predicted Covid-19 cases with such a high level of precision that it even helped researchers in identifying symptoms of the disease.
Economists have also managed to extract answers to hard-to-answer questions from this type of data before. It improves predictions of variables, such as foreign-exchange rates and building permits, that have long furrowed the brows of economic forecasters. Others perceive even greater potential in it.
One author, trained as an economist, wields Internet search data to reach conclusions about everything from unemployment’s effect on pornography viewing to racism’s role in voting behavior. “Nietzsche, Freud, and Foucault would have drooled over” data like Google Trends, he writes. After all, in his view, this type of data measures human minds as they ruminate in the privacy of homes, expressing themselves anonymously – therefore shamelessly, and truthfully – into keyboards rather than to pollsters or friends.
If true, it seems, Nietzsche, Freud, and Foucault would all agree: The upcoming data on inflation are likely to be bad news.