The release Monday of the Federal Reserve Bank of New York's monthly Survey of Consumer Expectations has forced many economists to reconsider whether the Federal Reserve will raise interest rates again this year after Wednesday's meeting. This uncertainty is the result of falling consumers' inflation expectations (let's call it CIE) combined with lower inflation readings for the past three months. While the falling core inflation rate is worrisome for the pace of normalization, the Fed or any economic agency should not rely on surveys of consumer expectations of inflation in their projections.
In its latest statement, the FOMC noted that "survey-based measures of longer-term inflation expectations are little changed, on balance," indicating that it does look at consumers' (and businesses') expectations of inflation, but needs to see a dramatic shift for the measure to affect their decisions. Even this mentality, however, overvalues the importance of these surveys.
The premise behind estimating what consumers expect inflation to be is that if consumers expect high inflation, they will tend to increase consumption now and vice-versa. That is, inflation expectations are supposed to be a self-fulfillment prophecy and since consumption forms the majority of the American economy, CIE should determine future inflation.
In practice, however, this idea rarely pans out. In an economy like our current one, the inflation rate is stable between 1 and 2%. At annual differences of at most 1%, consumers don't care what exactly inflation is. While economists and policymakers worry about every tenth of a percentage point change in the inflation rate, households usually don't pay attention to inflation until the economy is in a deflationary period or in a period with inflation above 4% or so-that is when inflation begins to make news.
As a result, most households do not have any expectations for what inflation will be. Thus, measuring those expectations is a futile exercise. More importantly, even consumers who monitor inflation will not base any of their decisions on small shifts in the inflation rate. No one consumes more because inflation will be 2% rather than 1%.
This result was demonstrated by the Federal Reserve Bank of Boston. The economists in Boston, however, point to two other reasons for the disconnect between theory and reality. The first is the nominal rate illusion, which says that consumers view money in nominal terms without respecting the role of inflation in determining the value of money. Thus, higher inflation would hardly make a difference to consumers' spending habits. The second reason is that most consumers expect their wages to lag behind inflation. Thus, higher inflation expectations would mean lower wealth, which too would serve to lower spending. Whatever the particular reason, it is clear that inflation expectations don't impact consumption, indicating that the statistic is not helpful in predicting future inflation.
A better and more leading measure of CIE would simply be to look at the price changes of food and oil. Since they are the most volatile components of inflation, they are also the ones that most dramatically affect consumers' view of inflation. Of course, the Fed eliminates them from its consideration and instead focuses on core inflation for the exact same reason, further depleting the power of CIE.
In fact, these two unstable parts of inflation form the majority of how consumers view inflation. In creating an Everyday Price Index that emphasizes items whose prices consumers care most about, economists from the American Institute for Economic Research found that the cost of food and beverages accounted for 41.7% of consumers' view of inflation, while changes in fuel prices accounted for a further 11.4%. The heightened importance of these two items is surprising given that economists like those at the Fed don't usually consider them in their reading of inflation.
That consumers' perception is a product of volatile components of inflation is evident when one takes the correlation between the Federal Reserve Bank of New York's CIE and the actual inflation in the same month. The correlation coefficient between the CIE and headline PCE is a reasonable 0.35, but between the CIE and core PCE the correlation falls to -0.054. Thus, there appears to be no link between consumers' expectations of inflation and core inflation. Households are not predicting inflation; they are reacting to news of inflation when filling out the survey, news dominated by food and oil prices.
CIE's lack of predictive power can be visually seen by comparing it to core PCE:
Source: US Bureau of Economic Analysis and Survey of Consumer Expectations, © 2013-2015 Federal Reserve Bank of New York (FRBNY) through FRED.
At best, core PCE and CIE seem to be concurrent indicators and in fact, CIE appears to be a lagging indicator of core PCE, rendering the measure useless.
We can also directly measure whether CIE affects personal consumption and thus growth since consumers' inflation expectations are supposed to affect inflation through their effect on personal consumption. Looking at the graph between personal consumption and consumers' expectations of inflation in the future yields no discernible connection between the two:
Source: US Bureau of Economic Analysis and Survey of Consumer Expectations, © 2013-2015 Federal Reserve Bank of New York (FRBNY) through FRED
Indeed, the correlation between personal consumption and consumers' expectations of inflation is 0.008, or in other words, zero. This graph also points to a striking result of looking at CIE: consumers' expectations hardly change. Even as oil prices were plummeting and the Fed raising interest rates, consumers' expectations have budged by less than one percent. This is surprising considering the Federal Reserve Bank of New York surveys the same people repeatedly up till a maximum of twelve months. One would think such consumers would become more accurate, but it appears that they are merely guessing.
We find similar results by taking a closer look at the data for the New York Fed's CIE. In its roughly four-year-long history, the lowest the 75th percentile estimate of inflation one year in the future has been is 4.42%, far above what inflation has been during that time. In this light, the steady decline in the rate could be viewed as survey respondents reading what actual inflation is at some point while giving their estimates and adjusting their answers accordingly. More tellingly, in the past nineteen months, only twice have the 25th percentile for the estimates of inflation one and three years in the future not been exactly 1%, signaling that consumers choose round numbers for their estimates. Since inflation is a figure that usually changes minutely, such estimation is bound to be less accurate than market-based measures of future inflation.
While considering consumers' expectations of inflation is a sound idea in theory, it can only mislead policymakers looking to estimate future inflation since the statistic is merely a construct of economists that fails to account for how consumers actually think. If we were facing a supply shock such as the one in the 1970s or deflationary pressure like Japan has, it would make sense to ask consumers what they think inflation will be since in that case, they will take inflation into account when making consumption decisions. In the current American economic climate, however, consumers' expectations of inflation should not be in the economist's dashboard.
Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.
I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it. I have no business relationship with any company whose stock is mentioned in this article.