Chart of the Day: Inflation Based on Actual Spending

How prices change could be the most important things that economists consider. Indeed, the Federal Reserve decides whether to flip the switch on monetary stimulus based in part on whether inflation is rising or falling. If prices are rising too quickly, they negatively impact the purchasing power of Americans, spending falls, and the economy begins to contract. That's what we saw over the past few months as energy prices rose.

Because inflation is such an important measure, calculating it accurately is vital. Yet some people complain that the government's methodologies distort reality and should be reformed. How would inflation look if its calculation was adjusted to answer one of those criticisms?

There are three major complaints about how the Bureau of Labor Statistics calculates the Consumer Price Index, which is the generally accepted measure for inflation in the U.S. First, BLS utilizes a practice called hedonics. This effectively lowers prices on goods that benefit from technological progress. Second, its estimation method assumes some substitution occurs when prices rise. For example, the calculation might assume that consumers switch to a lower grade of ground beef when prices rise. Finally, some people quibble about the product weights that the government uses for aggregate CPI, claiming that it doesn't accurately reflect actual spending.

That final problem can be easily solved, however. Another government agency, the Bureau of Economic Analysis, produces a price index based on its personal consumption expenditures calculation each month. It utilizes CPI data on a granular product category basis. So it doesn't escape hedonics or substitution. But for its aggregate index, product weights are based on actual spending -- not based on hypothetical weights that have been in place for several years. This makes BEA's price index particularly useful, because it shows how much inflation is actually affecting Americans in real time.

So let's compare the two. Below you'll find a chart with core CPI and the core PCE price index (core inflation excludes the often volatile items of food and energy and is the one most economists pay attention to in order to understand how inflation expectations are changing).

The first takeaway here is clear: these curves aren't that far apart through 2009. But you begin to see greater variance in 2010.

Interestingly, core CPI appears to actually underestimate inflation in 2010 and overestimate it in 2011. These variances have a few implications.

First, back in late 2010 when the Fed engaged in a second round of monetary stimulus, inflation doesn't appear to have fallen that low after all. According to the core PCE price index, inflation was fairly steady at 1% in November 2010 -- when the program began. At that time core CPI was just 0.6%.

However, the situation has shifted recently. This summer, we're seeing core CPI float above the PCE price index. Still, at 1.3% in June, the core PCE price index isn't low enough to warrant significant concern from the Fed that would push the central bank to provide additional stimulus to raise prices.