As Kevin Drum
points out, the way you calculate inflation over the past 50 years has a big impact on what you think has happened to the wages of blue collar workers:
How have blue-collar workers done over the past half century? Well, if you measure inflation via the Consumer Price Index (brown line), they've gone nowhere. Literally. They're making exactly as much today as they did in 1970. But if you measure inflation via the Personal Consumption Expenditure Index (red line), their wages have gone up nearly 30 percent. That's not spectacular—real GDP per capita has increased 121 percent since 1970—but it's a lot better than zero.
Economists argue about which measure is better, and I have never been able to form an opinion on which I prefer. Here's a description of the difference, from the
Cleveland Federal Reserve:
What accounts for the difference between the two measures? Both indexes calculate the price level by pricing a basket of goods. If the price of the basket goes up, the price index goes up. But the baskets aren’t the same, and it turns out that the biggest differences between the CPI and PCE arise from the differences in their baskets.
The first difference is sometimes called the weight effect. In calculating an index number, which is a sort of average, some prices get a heavier weight than others. People spend more on some items than others, so they are a larger part of the basket and thus get more weight in the index. For example, spending is affected more if the price of gasoline rises than if the price of limes goes up. The two indexes have different estimates of the appropriate basket. The CPI is based on a survey of what households are buying; the PCE is based on surveys of what businesses are selling.
Another aspect of the baskets that leads to differences is referred to as coverage or scope. The CPI only covers out-of-pocket expenditures on goods and services purchased. It excludes other expenditures that are not paid for directly, for example, medical care paid for by employer-provided insurance, Medicare, and Medicaid. These are, however, included in the PCE.
Finally, the indexes differ in how they account for changes in the basket. This is referred to as the formula effect, because the indexes themselves are calculated using different formulae. The details can get quite complicated, but the gist of the matter is that the PCE tries to account for substitution between goods when one good gets more expensive. Thus, if the price of bread goes up, people buy less bread, and the PCE uses a new basket of goods that accounts for people buying less bread. The CPI uses the same basket as before (again, roughly; the details get complicated).
Looks like the three blue lines level off just around the median income.
ReplyDeleteWhat makes "life satisfaction" different?
"The CPI is based on a survey of what households are buying; the PCE is based on surveys of what businesses are selling."
ReplyDeleteWell then wouldn't the obvious preference be toward the CPI, since we're measuring from the point of the view of the buying power of a person's wages? Who cares about the business end?
"The CPI only covers out-of-pocket expenditures on goods and services purchased. It excludes other expenditures that are not paid for directly, for example, medical care paid for by employer-provided insurance, Medicare, and Medicaid."
Why would you include employer- or government-provided services in a measure of the buying power of a person's wages? If you compare two individuals who make the exact same wage, but one has employer benefits, it doesn't make any sense to suggest that one has a higher "buying power" - it would make far more sense to instead suggest that the one with benefits either makes a higher "effective wage", as modified by the approximated value of their additional benefits. And you certainly shouldn't ever be averaging the two different individuals together for a composite value of "buying power" for their single shared wage.
"The details can get quite complicated, but the gist of the matter is that the PCE tries to account for substitution between goods when one good gets more expensive. Thus, if the price of bread goes up, people buy less bread, and the PCE uses a new basket of goods that accounts for people buying less bread. The CPI uses the same basket as before (again, roughly; the details get complicated)."
Again, since we're trying to measure a person's capacity to buy what they want and need, is making allowances for substitutions all that appropriate? Sure, if bread gets too expensive, people will buy less bread - but that means if you compare wages between one point and another, they are effectively "higher" at the point where an individual can actually afford to buy the bread they want instead of being forced to go without due to high costs.
From what I've read here, the PCE feels like it maybe sets out to shoehorn in factors which really don't belong, in an attempt to suggest there has been bigger growth than has actually occured. But of course, I have only the most cursory understanding, so make of that what you will.