Friday, February 8, 2008

Inflation - part 1/2

Many people will be surprised to learn that the dollar today is worth only 4 cents of a dollar in 1913 when the Federal Reserve, created for the stated purpose of fighting inflation and managing unemployment, was issued the charge of creating the nations currency. The process by which the dollar has gradually lost value (buys fewer and fewer goods) is known as inflation. Today the dollar can purchase less than one-twentieth of the real wealth (tangible physical goods) it could purchase in 1913 despite significant productivity gains registered over the last few decades. This fact alone should illustrate the importance of inflation with regards to the general standard of living, yet it is one of the most widely misunderstood concepts in economics today. This article is written in 2 parts. Part 1 attempts to deconstruct the mainstream view of inflation, and part 2 attempts to build a more meaningful understanding of the process.

Paasche or Laspeyres? False Choice!

Before we can understand what inflation is, it will be helpful to understand what inflation is not. Inflation is widely regarded as a general rise in prices. Economics by Parkin and Bade defines inflation as
an upward movement in the average level of prices. Its opposite is deflation, a downward movement in the average level of prices. The boundary between inflation and deflation is price stability.
This definition is quite vague and is disputed by Austrian economists, but, for argument, let us assume that Parkin and Bade are indeed correct and examine the consequences. So how precisely does one measure the "average level of prices"? The standard approach is to use a price index, a basket of goods and services considered representative of the economy as a whole. Tracking the change in this "representative" basket of goods, one has a proxy for changes in the "average level of prices". There are many problems with this approach:
  1. It is extremely hard to determine a "representative" basket of goods and services. Every basket must necessarily make simplifying assumptions in order to be tractable.
  2. Should one look at final consumer goods, intermediate goods, or a mixture of the two? In fact, we have the CPI and PPI precisely to make this distinction.
  3. Various sectors of the economy experience different changes in "average prices". For example, higher education has it's own "representative" basket, known as the HEPI.
  4. Most importantly, markets are constantly evolving to better satisfy human wants and needs. New products are introduced and obsolete products removed; existing products are improved and upgraded; how does the price index capture this? Price series either do not exist or may not be comparable.
As we can see, there are various practical problems related to constructing price indices. However, that is not the crux of the issue; there is a deeper philosophical problem: there is no objective manner in which to define a price index.
  1. There is no objective "representative" basket of goods and services. The inclusion (or exclusion) of any good or service from the price index may not be arbitrary, but it is certainly not objectively determined.
  2. With the market constantly upgrading and improving -- not to mention introducing and obsoleting -- various goods and services, how does one objectively compare price series on qualitatively different goods or when no such series exists. The approach here is known as hedonics, which is a quality adjustment made to the price in order to capture improvements and upgrades. Needless to say, this is a highly subjective affair.
  3. As the market is constantly changing, so too are consumer preferences. We do not live in a static world. Consumers, producers, economic actors of all stripes, are continuously evaluating alternatives and taking action based on their perceived self interest. People react to prices and alter their lifestyles accordingly. This cannot be objectively accounted for. It requires the subjective redefinition of the index.
Lies, Damned Lies and Statistics.

We have seen thus far that defining and measuring the "average level of prices" cannot be done in any meaningful objective manner. However, this does not imply that it is arbitrary. While there is much lee way, the "solutions" have a clear pattern: to systematically understate the true rise in prices. John Williams of Shadowstats demonstrates how the government alters economic reporting to paint a rosy picture. For example, looking at pre-Clinton, or pre-Carter methodologies for calculating the CPI shows that inflation is running much higher (7-10%) than the 3% it is currently advertised as. This is perhaps more in line with many peoples anecdotal experience as everyday costs appear completely divorced from government reporting.

To fully understand how the government manipulates CPI numbers, I suggest the reader start with the Boskin Commision. A detailed analysis is beyond the scope of this article, but very briefly, here are the major influencing factors:
  1. Hedonics: as described earlier are quality adjustments made to prices. A new car costs 20% more? Well, it is 30% "better" so the price has actually dropped 10%.
  2. Substitution bias: the price of steak has risen 20%? The price of hamburgers has declined 10%. Well, consumers must have "substituted" away from steak for hamburgers so prices are falling.
  3. Outlet substitution: that new jacket costs 20% more at the mall? Well, if you drive 50 miles to another mall, you can get it at the same price. Hence, no rise in prices.
  4. Redefining the index: housing prices rising? No problem, use rents instead (what is called owners equivalent rent). This was actually the scheme used to keep CPI numbers under control during the housing bubble that topped in 2005, allowing interest rates to remain at historic lows. Now that the bubble has burst expect a redefinition shortly that uses the Case-Shiller housing index (rapidly falling) instead of rents.
Although simplified and trivialized, this is an accurate picture of how easily CPI numbers can be doctored. But what is the motivation? For one, consider the obvious conflict of interest with the government having many liabilities (TIPS, Social Security, etc) linked to CPI "inflation". Even a rudimentary incentives analysis of how economic data impacts the perception and functioning of the government will demonstrate that government statistics are not to be trusted. A more detailed analysis of the methodology (eg shadowstats) confirms this initial impression.

Conclusion

While it is disconcerting that government statistics are so misleading, even more disconcerting is the definition of inflation as rising prices. As mentioned earlier, Austrian economists dispute this definition, and for good reason. First, as we demonstrated, a general rise in prices is a concept that cannot be objectively determined, let alone accurately measured. Second, and more importantly, it does not attempt to understand the root cause of inflation and serves to shift attention away from it by focusing the debate on completely meaningless red herrings related to the price index. To examine the true nature of inflation, we must shine a flashlight through this smoke screen.

One caveat related to our discussion here is that I fully recognize that rising prices are a phenomenon that is important to the general public. I do not mean to shift the discussion away from rising prices, nor do I mean to deride price indices in general. If our stated goal is to measure prices then there is no alternative to using a price index, despite the problems discussed above. However, our stated goal is not to measure prices, it is to understand the causes of inflation. We have not argued that price indices are useless, only that defining inflation as a general rise in prices is inaccurate. Certainly rising/falling prices are the crux of the issue, and we attempt to better understand them in part 2.

References

Inflation in one page, but Ludwig von Mises.

No comments: