THE ‘INFLATION MYSTERY” EXPLAINED

An important article from the San Francisco Fed to explain why inflation has been so low for so long in spite of the Fed’s efforts to push it up.

Since the end of the recession, inflation has been persistently below the Federal Reserve’s 2% target, with core personal consumption expenditures (PCE) prices rising on average about 1.5%. (…) that the labor market has fully recovered, the dollar has appreciated, and oil prices are no longer declining as rapidly, Chair Janet Yellen recently said that the reason why inflation remains low is a “mystery” (Yellen 2017).

In this Economic Letter, we examine the factors keeping inflation low by drilling down into inflation rates by spending category. We distinguish between categories where inflation has historically exhibited a procyclical relationship with overall economic conditions, moving in tandem with the economic cycle, and those categories where inflation has been acyclical, that is, driven by category-specific developments that are independent of the state of the overall economy.

We show that procyclical inflation has steadily returned to its pre-recession level, in line with improvements in economic conditions and a tightening labor market. However, acyclical inflation has been persistently low, suggesting that idiosyncratic factors have helped hold down PCE inflation.

We show that the key driver holding down acyclical inflation, and hence core PCE inflation, over the past few years has been persistent changes to the health-care sector that began after the end of the recession. Specifically, cuts to Medicare payment growth rates—which can affect prices throughout the health-care sector—have restrained health-care services inflation (Clemens, Gottlieb, and Shapiro 2016). Because health care makes up a large share of PCE, price changes within this sector can have sizable effects on overall PCE inflation. We estimate that low inflation from this sector is currently subtracting about 0.3 percentage point from core PCE inflation (…). While health-care services inflation is expected to pick up in the coming years, it appears unlikely to return to its pre-recession level, which could restrain core PCE inflation for the foreseeable future. (…)

We find that the categories exhibiting a procyclical relationship make up 42% of the PCE and include housing, recreational services, food services, and some nondurable goods. The acyclical categories, which make up the remaining 58%, include health-care services, financial services, clothing, transportation, and some other smaller categories. Based on this categorization, we create two distinct aggregate inflation series.

Figure 1 shows that the two inflation series occasionally move together. However, there are periods when they have diverged—in the late 1990s through early 2000s and again starting in 2014. In both of these periods, overall inflation was low—1.7% and 1.6%, respectively. In line with the historical relationship, procyclical inflation has steadily increased over the past few years as the recovery continued to gain steam and economic slack diminished further. However, acyclical inflation has gradually declined. The patterns of these two groups suggest that core PCE inflation has been persistently low due to weak acyclical inflation.

To get a better sense of how each series affects core PCE inflation we plot their contributions in Figure 2. (… )

Procyclical inflation currently contributes about the same amount as it did in 2002–07. However, acyclical inflation is contributing about 0.6 percentage point less than it was during these years. In other words, core PCE inflation would be about 2% if acyclical inflation were behaving as it did in the mid-2000s. (…)

In Figure 3 we measure the extent to which health-care services has caused the decline in the acyclical group’s contribution to core PCE inflation in recent years. The black line is the deviation of the contribution of acyclical inflation from its 2002–07 average—that is, from Figure 2, panel B, the blue line minus the dashed horizontal line. In Figure 3, the blue bars represent how much of the acyclical deviation (black line) is attributable to health-care services, and the red bars reflect the contribution from other non-health acyclical categories.

(…) The very recent dip during 2017 in the acyclical group’s inflation contribution—and hence core PCE inflation in general—can be traced to a relative decline in the contribution from the acyclical categories that are not related to health care. We found this mainly reflects a steep decline in the prices of cellular phone services. (…)

What has been causing this persistent decline in health-care services inflation? A major factor has been legislated changes to Medicare payments. Through various pieces of legislation, including the Affordable Care Act (ACA), the federal government has slowed the growth of Medicare payments to physicians and hospitals. These payments are directly included in the BEA’s health-care services price index, which by construction translates into a slower rate of health-care services inflation. As shown in Clemens and Gottlieb (2016) these legislated changes to Medicare payment growth can also affect the payment growth for private insurers and thus pervade throughout the health-care services sector.

The Centers for Medicare and Medicaid Services (2017) recently published its inpatient Medicare payment schedule for fiscal year 2018. Payments are set to grow at 2.0% in this fiscal year, up from a rate of 0.9% in fiscal year 2016 and 0.6% in fiscal year 2017. Based on past estimates of how Medicare passes through to affect prices of other healthcare services, this projected increase could translate into as much as a 0.3 percentage point increase in overall health-care services inflation, albeit only a 0.05 percentage point increase in core PCE inflation. Furthermore, some of the legislated payment growth cuts mandated by the ACA will expire after fiscal year 2019. (…)

So:

• Inflation has indeed reacted to the economic stimulus as “procyclical” inflation has accelerated from 0.7% in 2009 to its current level of about 2.5% with a clear cyclical uptrend.
• Desinflation in “acyclical” goods and services is independent of fiscal or monetary intervention. It has been particularly strong in the past 18 months to totally mask the cyclical reflation trend.
• The Fed is wrong to focus on getting inflation up to its +2.0% target. The Fed’s policies have little, if any, impact on the desinflating half of the PCE price index.
• However, it is right to start normalizing interest rates because “procyclical” inflation is on its normal cyclical uptrend and is already above the Fed’s target.
• Total inflation seems set to accelerate if the “acyclical” component is indeed about to strengthen.

Other inflation facts:

• Core CPI troughed at +1.7% YoY in August. It was +1.8% in October. Aug-Oct. annualized: +2.4%.
• Total CPI troughed at +1.6%  last June. It was +2.0% in October. Aug-Oct. annualized: +4.2%.
• Core PCE inflation troughed at +1.3%. It was +1.4% in October. Aug-Oct. annualized: +1.9%.
• Total PCE inflation troughed at +1.4% last July. It was +1.6% in October. Aug-Oct. annualized: +3.0%.
• The Cleveland Fed median CPI troughed at +2.1% in July. It was +2.3% in October. Aug-Oct. annualized: +2.8%.
• The Cleveland Fed 16% Trimmed-mean CPI troughed at +1.8% in August. It was +1.8% in October. Aug-Oct. annualized: +2.0%.
• The Cleveland Fed Inflation Nowcast suggests Q4/17 annualized CPI of +3.6%, PCE of +2.7%, core CPI of +2.2% and core PCE of +1.9%.
• The NY Fed Underlying Inflation Gauges (UIG) “currently estimate trend CPI inflation to be in the 2.25% to 3.00% range, with both registering above the actual twelve-month change in the CPI.”