An unusual good news inflation story

Consumer spending has been surprisingly resilient in the face of a sharp rise in taxes and, more recently, higher petrol prices. It rose 0.3% (after inflation) in both January and February, the government reported this morning, and looks likely to rise at a 3% annual rate in the first quarter, a major reason many first quarter GDP growth estimates now top 3%.

Theories for this resilience range from the countervailing positive wealth impact of higher stock and home prices to the tendency of consumers to smooth through changes in disposable income. But a little-noticed factor has been lower inflation. I say little-noticed because inflation isn't lower based on its most widely followed measure, the consumer price index. Both total and core CPI (which excludes food and energy) are up 2% in the last 12 months, in line with its long-term trend and the Federal Reserve's 2% target.

However, based on the lesser-known price index of personal consumption expenditures (PCE), headline and core inflation are only 1.3% (see nearby chart). This is more than a technical curiosity. The PCE index is used to calculate real consumer spending. Using the CPI, real spending would be up only 0.4%, instead of 0.6%, in the last two months. Over the last year, using the PCE index instead of the CPI adds 0.7% to the growth in both real consumption and real incomes.

The two indexes frequently diverge because they are constructed differently. While the weights in the CPI basket change only every few years, the PCE's change each month, better capturing consumers' tendency to shift from more expensive commodities and outlets to cheaper ones. The CPI's weights are also determined by what consumers say they spend, whereas the PCE index is based on what they actually spend, or what is spent on their behalf, such as the employer's portion of health insurance, and what the federal government spends on Medicare. As a result the CPI assigns much more weight to rent and housing and much less to health care. PCE inflation over time typically runs about 0.3% below CPI inflation, but the current divergence, at 0.7%, is the largest in more than a decade, according to Goldman Sachs.

The reason appears to be the divergent behavior of rent and health care. Rent inflation has been relatively firm lately, reflecting strong demand by households who no longer qualify for mortgages or lost their home to foreclosure. Rents have an outsized impact on the CPI because they are used to determine the cost of owner-occupied as well as rental housing. Meanwhile, as has been widely discussed, medical inflation has eased notably in recent years. No one is sure why; the Obama Administration credits cost controls implemented under the Affordable Care Act. The weak economy, which has forced employers and employees to curb consumption and forced providers to control costs better, is almost certainly a factor, too.

This means that to the extent CPI inflation overstates what consumers spend on housing and understates what they spend on health, it is overstating the cost of living. To be sure, workers do not see all the benefit of lower health inflation since their employers and the federal government pay most of the cost . But over time this ought to translate into higher take-home pay as money that would have otherwise gone to benefits goes to wages instead. It's a helpful reminder of how much consumers stand to benefit from any reforms that reduce health care costs. (Mechanically, this is only true if price inflation falls by more than wage inflation, i.e. if it's due to narrower profit margins or increased efficiency, not reduced volume of services, or lower wages to health care workers). Similarly, higher rents only affect the small portion of the population now trying to rent an apartment. The majority of people living in their own home, or about to buy one, are still benefiting from low home prices in most markets and rock bottom mortgage rates.

This poses an interesting dilemma for the Federal Reserve. It considers the PCE index superior to the CPI as a measure of the cost of living, and so focuses on it in forecasts and policy decisions. That is why for some months it has said inflation is running "somewhat" below its 2% target. That's an understatement; Eric Rosengren, president of the Boston Fed, called it "far below" target in a speech Wednesday.

Indeed, judging by the PCE index the Fed is missing its mandate almost as badly on inflation as it is on unemployment, an even more compelling reason to maintain a fairly stimulative monetary policy. The main problem with below-target inflation is that it translates into higher real interest rates. But real rates are determined by expected, not actual, inflation. Because consumers and investors focus more on the CPI, their expectations haven't responded at all to the pronounced slowing in PCE inflation. If anything, they may have edged up a bit in the wake of the Fed's stated willingness to let inflation top 2% in pursuit of lower unemployment. If CPI inflation were to converge to PCE inflation, that would be a concern. Goldman expects CPI inflation to drop to 1.8% in coming years and PCE inflation to rise to 1.5%. It would be preferable for both to converge to 2%; but so long as inflation expectations remain where they are, it is of little consequence for monetary policy – and a tangible plus for incomes and spending.