Our Great Conundrum: An Unbroken String Of Inflation
In February 2000, the Federal Reserve reported to Congress that it was changing its preferred measure for inflation. When we think of inflation the CPI springs to mind without much effort. It has been ingrained in the mainstream economic psyche for decades, and thus has become synonymous with the concept. Even the federal government statutorily uses the CPI to calculate cost of living and other monetary changes.
The move away from the CPI was not addressed up front, in fact. The Fed actually buried it in the footnotes to a table on its economic projections for calendar year 2000.
"In past Monetary Policy Reports to the Congress, the FOMC has framed its inflation forecasts in terms of the consumer price index. The chain-type price index for PCE draws extensively on data from the consumer price index but, while not entirely free of measurement problems, has several advantages relative to the CPI. The PCE chain-type index is constructed from a formula that reflects the changing composition of spending and thereby avoids some of the upward bias associated with the fixed-weight nature of the CPI. In addition, the weights are based on a more comprehensive measure of expenditures."
The conspiracy theorists saw/see this as an academic invention to intentionally undercount monetary-driven inflation, particularly based on the elimination of the "upward bias" of the CPI. But that upward bias is, in fact, real in the academic setting, and therefore may not at all be appropriate. This is known in the statistical community as a "formula effect."
The Bureau of Labor Statistics (BLS), the agency responsible for the CPI, is well aware of these limitations. In fact, it acknowledges them by producing several different inflation statistics beyond the CPI-U (which is the primary indicator most people refer to). The CPI-U is calculated using a Laspeyres formula - essentially creating a basket of goods that represent a "typical" urban dweller's living expenses, and then calculating aggregate prices of that same basket of goods a year later (or whatever period is being measured).
That leads to complications and problems where the prices of certain goods rise while others fall, and we are left hoping that the end result of the mess gives us some indication about the general change in consumer prices. The primary problem with this approach, leading to that upward bias, is the lack of inclusion of newly created goods. This is Bill Dudley's infamous iPad analogy.
To overcome this limitation, which is very real, the PCE deflator uses a chain-type approach, calculated via a Fisher-Ideal formula. The BLS considers it a "superlative" (technical term, not adjective) index that reflects actual consumers because it is derived mostly from spending data.
The PCE deflator, then, incorporates the "basket of goods" so that they reflect the changing composition of modern life and economy. The 1974 basket of goods in the CPI is not going to contain an iPad, so it would, in fact, be difficult to calculate inflation without trying to obtain some measurement of changes in goods as they come and go through time.
As you might surmise from the inclusion of new technology, inflation would tend lower due to the process of goods being adopted in a broader marketplace. That is the essence of tech goods and consumer electronics, and does indeed mark a beneficial increase in living standards. But does that relate to inflation? There are numerous problems with incorporating productivity-driven gains in a calculation of price changes that may or may not be driven by other factors (primarily monetary).
But that is a separate discussion all its own. For now, the inclusion of new goods in the chain-type measurement is at least more consistent with the intent of the exercise - to gauge overall changes in prices. But there is an additional factor that leaves the CPI inferior, in the minds of the economic academy, namely the fuller concept of "substitution."
Substitution is very intuitive, bolstering the case against the CPI. In the Laspeyres, fixed basket of goods approach, as prices of certain goods change the weighting inside the basket does not. In the real economy, changes in prices are in no way benign like that. If prices rise, particularly if they rise a great deal, people respond by purchasing less of that good or searching for a substitute that may represent equal or better "value." Therefore measuring the same basket of goods from year to year does not fully capture what the real economy actually might look like or how price changes affect function.
Even economists have realized that changes in prices bring about changes in behavior. And there it is; the true idea of inflation, buried under layers and multitudes of mathematics and philosophical debates. As prices change, so does our behavior.
The entire enterprise of condensing the concept of inflation into a single, mathematic measure is a fool's errand in itself. Inflation is not some monolithic factor that can be quantified like some ballistic trajectory through known space; despite the gloss, there is no hard science here. The disagreements among the CPI and PCE deflator highlight exactly this idea, namely that price increases impact people and their behavior differently. We can hope that the aggregate measure captures the net effects across the spectrum, but it really never does.
This goes beyond using representative sampling or forcing commonality and uniformity into, or out of, the basket of goods, it really relates to the fact that people are influenced by price changes in very different ways that cannot easily fit into an all-consuming singularity. I mean to say that there is more here than just the difference between how "poor" people and "rich" people might absorb rising commodity prices - that is the obvious problem using a singular measure.
In looking back on the housing bubble and its disastrous aftermath, I think we are all pretty well acquainted with the scope of it all. There are several ways of quantifying the disaster, but in raw, construction terms it is a pretty simple task. In January 2006 there were 2.3 million (annual rate) new housing projects started, including 1.8 million new single-family homes. The 891,000 starts in August 2013, then, shows just how far the real estate sector has collapsed - and it has taken a "mini" bubble just to get back to a little over a third of the activity at the peak of the mania.
Given those figures, it is somewhat curious that nobody refers to the 1970's real estate market in the same way. In November 1978, housing starts were a little under 2.1 million, with 1.5 million new single-family homes. Given the smaller population back then, these numbers are easily comparable to 2006, if not even more extreme relative to the population size. By January 1982, the level of total new starts had fallen all the way back to 843,000, surely what might be called an epic bust.
Despite that, there is no common reference to the great housing collapse of the late 1970's. The reason, no doubt, relates to prices. There was no price retrenchment then. Between 1971 and 1990, the Census Bureau only records one quarterly decline in the median sales price of all homes. The average annual rate of change between 1971 and 1980 was 11.6%, close to double the 6.8% we saw between 2000 and 2006.
The disparity here is due to simple economics. There was actual demand for new residential dwellings in the 1970's, particularly as baby boomers came of age. Despite the ups and downs in the economy during that insecure decade, people were buying houses to live in them - a lot of them. They were most assuredly not buying them, in the aggregate, as a vehicle for savings. There was no frenzy to earnestly borrow an equivalent amount to their life savings in the hope of scalping some of the expected rapid price appreciation (beyond the fees and interest that is scalped by FIRE).
That sets up an interesting dichotomy that is highlighted by the characteristics of each "bust" period. In the 1979-82 period, demand for new houses fell off a cliff, but prices did not because there was no wave of selling. Since those that had bought previously were purchasing the economic good "shelter", prices were more than firm regardless of marginal new activity. Thus the quantity of construction was the only variable that changed once demand fell sharply in the high double-digit mortgage rates of the last years of the Great Inflation.
In some respects, that is contrary to how we think about behavior changes in inflationary conditions. Conventional wisdom posits that "real" goods are favored over paper securities as protection from such environments. Yet, here we have the most spectacular episode of avowed runaway inflation in our history, but savings tended toward paper securities. Household savings were proportionally far more favoring bonds and even vanilla deposit accounts because interest rates followed inflation rates. There were no disco-era flippers.
The housing bust after 2006 was such an epic event precisely because so many homes were built in excess of actual demand for shelter. The economic understanding of housing changed from a place to live to a place to invest. And that was true of circumstances beyond just construction. The entire refi boom and houses-as-ATM's was a sea-change in economic behavior toward real estate. Perhaps there was a generational difference between the early baby boomers of the 1970's who were much closer to the age of their far more conservative parents that had economically "survived" the Great Depression, and those that followed into the nascent housing bubble of the late 1990's. I don't believe, however, any such difference, if it actually existed, would offer a satisfying explanation for such a drastic shift in behavior.
At the May 2006 FOMC meeting, Fed Governor Susan Bies, in light of the burgeoning collapse in housing, stated openly,
"But the fact that inflation continues to be above 2 percent in the forecast period is something that does concern me, and I think part of my concern relates to the tremendous amount of liquidity that sits out there in the banking sector, in the U.S. financial markets, and clearly globally. The presence of this liquidity is something that we really need to think about. It's not back to where it was in my money supply days, when I started my career at the St. Louis Fed; but I do worry that liquidity is, as some of you have said, causing a lot of transactions to occur that economically perhaps wouldn't otherwise occur."
That last sentence just begins to get at the crux of what inflation really is, "causing a lot of transactions to occur that economically perhaps wouldn't otherwise occur." While not directly referring to the kind of behavioral change I describe above, it does not take a leap of faith to convert that sentiment over time into an entirely new behavioral paradigm. A lot of "transactions" that would not make sense economically of their own year after year after year would slowly and imperceptibly lead to the very conditions that created housing mania - and the stock mania that preceded it.
It's more than just a monetary imbalance, as the seeping of this kind of behavioral distortion into the economic fabric across decades would lead to an entire array of economic dyspepsia. Not the least of which is the growing acceptance and even dependence on financialization for everything. That is particularly true as marginal growth shifts from more productive endeavors, those transactions that might occur absent "that liquidity", to those encompassed by this behavioral transformation.
After several decades, the constant appeal to finance as a panacea becomes as second nature as believing inflation can be properly measured by the CPI or PCE deflator. That leads me to two conclusions that are utterly disdained by the mainstream. First, asset inflation does exist as evidenced by the clear change in behavior across the decades. It is not measured because it does not fit into the narrow box of accepted definitions. Further, there is absolutely no methodology or even conceptualization of how it might be measured and incorporated into the accepted model of monetary intrusions onto the real economy. The mainstream accepts the premise that monetary imbalances might lead to "consumer" inflation, but it ends there. Clearly that model of explanation is not sufficient to explain how houses went from a place to raise a family to a place to gamble the family's long-term future.
Second, and far more important, such asset inflation carried over time is even more insidious and harmful than consumer inflation. The current model of monetary mechanics posits that the answer to every economic problem is financial. Either through "liquidity" or interest rate psychology, there is always a monetary/debt rejoinder to lead the way into economic robustness. And in each episode, behavior responds by expecting financial results when, in fact, the real economy needs productive behavior - far less home flippers and day traders in favor of far more entrepreneurs looking to harness some industry that actually creates wealth rather than seeking to transfer someone else's.
Stocks are at, or are close, to an all-time high despite economic circumstances that are almost diametrically opposed to such an occurrence. There are certainly circumstances where share prices might move in opposition to fundamental factors (earnings have largely stagnated since the middle of 2012; prices have risen solely on multiple expansion), but in every single case where that occurs it leads inevitably to the "liquidity causing transactions that might not otherwise occur" explanation.
Yet, it can all be distilled back to behavior. This is not to say that stock rallies are never appropriate, even if they are not attached in commensurate proportion to "fundamentals", but more often than not there is a monetary-behavior scenario behind it. The best performing stock market in the world right now operates in Caracas. The Bolsa de Valores de Caracas is up 412% in the past year. Also in South America, the Buenos Aires Merval Index is up 119% since early October 2012. Yet, neither of those nominal increases represents anything other than a devaluing currency. Investors in those countries are seeking nothing more than some kind of shelter from more obvious "inflation", but such is still consistent with everything we know about how "prices" impact behavior.
The most absurd example is the Zimbabwe experience. While statistics are a bit hard to come by, we know that the Zimbabwe stock index was actually doing quite well, relatively, up to 2007. In April 2007, the Zimbabwe Industrial Index (I didn't know it existed, either) had risen 595% already for the year. In the prior 12 months, the nominal return was 12,000%. Inflation for the period was estimated at only 1,729% per year, so it appeared like stocks were "winning" the race. It did not matter that unemployment had reached about 80%, or that official stats estimated GDP had more than halved from 2000.
By the middle of 2008, it all fell apart, nominal or not. Inflation could not be "reliably" calculated, though it was estimated by outside sources (the Cato Institute, for example) in November 2008 at 516 quintillion percent per year; the official figure was only 231 million percent per year. No stock market gains in the world can keep pace when prices double every fifteen hours or so (at least there would never be margin calls). When a currency becomes worthless, there really are no places to hide, but you can easily sympathize with the attempt.
Again, admittedly those examples are extreme to the point of being reductio ad absurdum. But the larger point, I think, stands. In the United States, and other countries, particularly Japan, we have been conditioned into the financial framework where all that makes sense is a financial appeal. Everything is a potential investment because everything has been financialized. Hard work has been devalued by the chase of "easy money", so the economy reflects as much, as do financial indications.
None of this can fit into the current concept of inflation, even when it recognizes the monetary causation. However, even the idea of the CPI or PCE deflator comes down to an attempt to measure behavioral changes across the economic spectrum. At some point, we can hope the mainstream begins to broaden that scope to include all monetary imbalances, all "transactions that wouldn't otherwise occur." Our great conundrum is an unbroken string of inflation, spanning more than four decades now.