When the Bureau of Labor Statistics (BLS) released the August Consumer Price Index (CPI) figures, it provided social media economists another opportunity to cherry pick the data to “prove” almost any conclusion.
Consider that each of the following statements is true:
Consumer prices increased 2.9% year-over-year. But also, at an annualized rate, they were up 4.7% in August. What about energy? The August increase in energy prices was more than twice the July increase. But they are essentially unchanged in the past year. Food? Food price increases were 66% greater in August than in July. Compared to a year ago, however, food prices are only 3.2% higher.
It’s also true that Volunteer head football coach Josh Heupel is winless at Tennessee—based on the most recent week’s data.
As 19th century British statesman Benjamin Disraeli famously observed, “there are three kinds of lies: lies, damn lies, and statistics.” One power of statistics is that the same data can be massaged to support nearly any narrative, while still remaining technically correct.
This is why relying on the political bent of whomever writes the headlines in your Facebook feed is not a great way to understand a complex, multiperiod data series. My recommendation is to be your own economist–save the monthly BLS data in a big spreadsheet so you can calculate the first and second derivatives of each price category each period.
Well, that’s my ideal recommendation, but I might be more of a data nerd than some people. If you don’t want to dust off your differential calculus textbook, here are some suggestions.
Never rely on a headline to inform you about an economic statistic. This is especially true for any economic data or conclusion presented in a brightly colored box on a Facebook post. If you aren’t going to read the complete article, don’t let the headline influence you. (This is good advice about most things.) It’s okay if you don’t read the article; it probably makes little difference in your decisions if the BLS tells you energy prices changed a tenth of a percent more last month than the previous month.
Your decisions should be guided by broader, long-term trends not monthly minutia. The next time you see a passionate post about inflation, maybe pause before sharing it in outrage and ask yourself: what specific part of the CPI report are they using? Year-over-year? Month-over-month? Just eggs? While the number themselves may be accurate, the context is usually what prevents you from being misled. And context is always optional, both for the writer and the reader. Once you realize that, you start noticing the same trick everywhere.