Recent Posts

Wednesday, February 15, 2017

U.S. Inflation Trends -- Not Really A Macro Theory Story

Chart: U.S. Core CPI Inflation

The January CPI inflation data were perkier than expected, causing some to dial up their forecasts for Fed rate hikes. I would not be surprised if the Fed becomes somewhat more hawkish over the coming months. Since they were hiking rates at a pace of 25 basis points per year, that bar was set fairly low. Although we should expect inflation to drift higher in the late stages of the economic cycle, macro theory is probably not going to give us much insights for near-run inflation forecasts.

As the above chart shows, the annual inflation rate for core (excluding food an energy) CPI inflation was 2.3%, which is marginally above the 2.16% compounded average over the 1993-2017 interval. This puts this inflation rate near the cyclical highs that we have seen since the beginning of this cycle (in fact, it almost exactly the same level as it was in January 2016,) It remains below the cyclical highs of previous cycles, which were in the 2.75-3.0% range.

The stability of core inflation in the post-1990 environment has meant that macro theory has not been particularly useful for forecasting it. There are at least a half dozen reasonable explanations why inflation has been quiescent in this period, which is completely unlike its behaviour in the 1970s. As a result, so long as you avoid boneheaded mistakes (such as trying to fit a regression model to inflation rates, or believing in the Quantity Theory of Money), you could have correctly forecast inflation outcomes using any number of theories.

The least common denominator of the various plausible inflation theories is that inflation rates will rise slowly during an expansion. Although this is somewhat useful to know, it is not giving us a lot of information. Instead, we need to take deep dives into the price index data, and try to understand what is happening at the sector level. My feeling is that a lot of the drivers for the inflation rate in the current environment have little to do with the output gap story that the Federal Reserve allegedly has power to control.

I recommend the article "Considerations on Cost Disease" by Scott Alexander. He discusses the some of the major drivers of inflation in recent decades, which are all in the services sector. These components of the price index have exploded relative to the rest of the index. The main culprits have been:
  • medical costs;
  • education; and
  • rent.

Alexander's article focuses on medical costs. The high costs for medical treatment in the United States is linked to the structure of its system. As a Canadian bystander, I am going to dodge the politically-charged debate around the reform of "Obamacare." However, as Alexander's article shows, the rising cost of treatment is not related to medical wages, which have been stagnant in real terms. In other words, they are unlikely to have any relationship to the output gap.
Chart: Rent Divided By All Items CPI

The chart above shows the trend in "real rents" (the rental component of CPI divided by All Items CPI). It has been on a broad secular up trend, interrupted by the real estate busts in the early 1990s and then a smaller reversal around the Financial Crisis. Since the beginning of the decade, rents have been on a tear.

Rent CPI Inflation Rate

The chart above focuses on the inflation rate for rent in the post-1990 era. It has been steadily marching upward, and is almost double core CPI. (Since rent is a major component of the core, this means that rent is heavily outperforming the rest of core CPI.)

It is quite unclear how Federal Reserve rate hikes are supposed to suppress rent increases. The natural competitor to renting is to buy a home, and mortgage interest costs are a major factor for affordability. Meanwhile, it raises the hurdle rate for investors who own rental properties.

About the only way the Fed can control rent increases is to emulate its policies of the last cycle. Not regulating the financial sector and letting it blow itself up seems to be the only reliable way to put downward pressure on rents. That said, there are some unwelcome side effects from that policy stance.

Chart: U.S. Tuition Component of CPI

Turning to education, it has not been a factor in the recent pick up in inflation. The inflation rate of Tuition and Childcare component of CPI  has dropped to roughly the same level as the overall index. This is somewhat welcome, but it does follow a period of continuous rising "real prices" since 1980.

Scott Alexander linked to the Delta Cost Project, which is examining the spending increases in the American college system. I did not get a chance to examine the data in detail, but the summary appears to be that "student services" are a major source of cost inflation.

This matches my experience with the higher education system. The amenities that are available to university students are incredible when compared to what was provided even in the early 1990s when I was a grad student. Living conditions still resembled those seen in The Young Ones.

One common complaint is that the cost overruns in education and medicine is the result of government interference. In a sense, that is true. If you tilt policy in a fashion designed to raise costs, the costs of a programme goes up. Although the Canadian medical system is far less dysfunctional than the American, medical cost containment is a political issue that will not go away. Meanwhile, the situation in the university system is probably the result of "market oriented" reforms.

Over the past few decades, the big idea was to bring in MBA's who would run universities "like a business." They managed to trample over centuries of inherited knowledge, and blow out the academic cost structure by turning universities into educational theme parks. The fans of modern management techniques failed to understand that the only reason that most modern corporations remain profitable is that their competitors are also run by MBA's who are even more incompetent. (I owe that insight to the best management theorist out there, Scott Adams.) Since universities are rarely culled as a result of competitive pressures, there is no way to counteract incompetent leadership.

In any event, such cost trends are unlikely to have any relationship to the business cycle, which is underlined by the fact that they are moderating while the expansion continues.

(c) Brian Romanchuk 2017


  1. My idle speculation on how Fed policy could affect rent inflation- Large rate hikes cause a drop in demand for purchases financed through borrowing. This leads to unemployment and recession. Unemployed people find they cant pay rent and get evicted or foreclosed on. This increases the supply of rental units available while decreasing the supply of people who have effective demand for rental units. Rents come down.

    So its really very simple- to lower rent inflation we just need to increase unemployment and homelessness.

    1. Yes, I exaggerated when I said the Fed needed to cause another financial crisis; just a recession would do the job. :-(

  2. Shouldnt we look at underlying reasons for each of these sectors inflation increases? You touch on medicine, since we dont see real wage increases in medicine some other component must be contributing.

    Looking at rent specifically, and im just purely speculating...but i believe we seen changes in renters preferences over the last two decades. Urbanization, increased preferences for people wanting to live in cities. Those sorts of consumers changes that cause distortions in long term inflation shouldnt seem to concern fed officials too much?

  3. These sectoral stories are enough to move the inflation rate 1/2% or so (my guess for the magnitude); if people spend more on rent (medicine...), they have less money to spend on other things, so there's compensating downward price pressure elsewhere. Since the overall inflation rate has been so stable post-1990, these sectoral shifts are comparable to the cyclical shifts in inflation, which is the only thing the Fed can hope to control.

    If we ever got wild cyclical swings in the overall inflation rate again (like in the 1970s), these sector stories matter a lot less.

    Each sectoral story is different. For rent, the long run up in house prices has got to be a major contributor. However, the conditions vary from region to region. I live in Montreal, and cost of renting and buying homes has exploded since 1998 or so (coincidently when the CMHC loosened lending standards to idiotic levels), whereas house prices and rents had been stagnant almost for a couple of decades before that (even in nominal terms, never mind inflation adjusted).

  4. Rents went up when rates went down because monthly payments were stable. Raising rates will stop price increases in housing, but it will also kill the market and probably bring about recession. It's not like there aren't plenty of people still on the edge. The jobs are in the cities, which gives rentiers power. But only so much as wage growth is falling. School debt in US is still out of control.

  5. Rents went up when rates went down because monthly payments were stable. Raising rates will stop price increases in housing, but it will also kill the market and probably bring about recession. It's not like there aren't plenty of people still on the edge. The jobs are in the cities, which gives rentiers power. But only so much as wage growth is falling. School debt in US is still out of control.

  6. This comment has been removed by a blog administrator.

  7. This comment has been removed by a blog administrator.


Note: Posts are manually moderated, with a varying delay. Some disappear.

The comment section here is largely dead. My Substack or Twitter are better places to have a conversation.

Given that this is largely a backup way to reach me, I am going to reject posts that annoy me. Please post lengthy essays elsewhere.