Inflation, Lies, and Statistics

When it comes to the subject of central bank money-printing and its effect on the prices of goods and service in the broad economy, there is a common misconception that the overall loss of purchasing power can be measured through the widely-reported Consumer Price Index, also known as the CPI. It is important to note, however, that the CPI does not measure the change in prices for all goods and services in the economy and is therefore not a total price index representing changes to the value of money over time. It is simply a basket of goods and services weighted in some combination as to attempt to reflect the typical purchasing patterns of the “average” consumer.

There is of course no such thing as an average consumer, only a representation of what some government-sponsored statistician believes average consumer spending patterns should look like. But even if it were possible to come up with an index that perfectly matched the typical spending tendencies of the consumers who are routinely surveyed by government agencies, it is fatally flawed logic to believe that whatever number is arrived at through a host of questionable assumptions and statistical data adjustments would also represent the general decline in the purchasing power of money.

It is often said that a housewife can tell you more about the rate of inflation than can the government statisticians who calculate the CPI. This widespread suspicion towards officially-reported government inflation statistics has led to many analysts accusing the government of “lying” about the rate of inflation or deliberately massaging the data to understate the true degree in which goods and services are increasing in price. While these detractors may be correct to some extent, the point of this article is not to argue some nefarious motive on the part of government, but to point out that the general decline in the purchasing power of money is virtually impossible to directly measure using a subjective basket of goods as the starting point. This difficulty is only accentuated through the additional flawed methodology used by government statisticians to adjust the CPI in every which way to arrive at a figure deemed more “realistic” of actual consumer spending patterns.

In a modern-day exchange economy, prices are constantly fluctuating in accordance with changes in individual consumption preferences, but this does not directly equate to a change in the aggregate purchasing power of money. If we consider a person, for example, who begins to prefer Product A to Product B and subsequently alters their spending patterns in accordance with this change in consumption preferences, the price of Product B will fall and the price of Product A will rise accordingly. But clearly it is absurd to say that anything has happened to the general purchasing power of money. All that has happened is that Product A has become more expensive for the consumers of Product A, and Product B has become cheaper for the consumers of Product B. If Product A is included in the CPI and Product B is not, then the government statistician will record an increase in the CPI and will say that prices have risen. But this is false. Prices have risen for some members of the population, perhaps even the majority of the population, but they have also decreased for other members of the population who are purchasers of Product B. The fact that many consumers will have seen prices rise does not negate the fact that prices must have correspondingly decreased in other areas of the economy that are not necessarily measured through the government CPI.

The initial problem, then, can be seen to be one inherent in the composition of the index itself, for any price index that is constructed based upon only a small subset of all possible goods and services in an economy must necessarily introduce tracking error between the index and the rate of general price increases incurred by society as a whole. While it is true that government statisticians make some effort to base the CPI on realistic consumer spending patterns, the decisions made even on which groups to include in the index is itself fraught with bitter controversy. Items purchased more frequently by society, for example, are naturally given a higher weighting within the CPI by government officials. But this leads to a resulting index composition that tends to be more inclusive of spending patterns of wealthier consumers who enjoy greater purchasing power than their poorer brethren. The CPI therefore contains biases right from the outset, marginalizing the spending patterns of those less fortunate consumers living on the financial fringes of society.

In a similar vein, there is often a divide in consumption preferences between urban and rural consumers, but since the CPI is made to myopically focus on the spending habits of urban citizens it will tend to under represent the spending patterns of rural consumers in favor of urban spenders. Even the decision as to whether or not various government taxes should be included in the CPI is subject to no small controversy. As currently constructed, the CPI includes sales taxes which, while certainly an additional cost incurred by consumers, is completely dependent on the whims of government and not a reflection of supply and demand characteristics of end consumer products or the quantity of money in circulation. It represents merely a surcharge by government, not a genuine change in the overall purchasing power of money.

And yet the list of problems grows longer still. Given all of the central bank money-pumping of the last decade, the casual observer is often left wondering as to why the CPI has yet to increase to the same degree as the economy-wide money supply. In response to the unprecedented global liquidity glut of the last decade, one of the primary outlets for this newly created money has been a significant increase in speculation in stocks, bonds, and real estate. The CPI, however, does not include these products in its calculation since they are classified as items purchased for the purpose of savings and investing, not consumption. This, however, ignores the fact that this river of newly-created money has bid up the prices of both financial and real assets, decreasing both future returns and income streams to would-be purchasers. The fact that a large portion of the population still purchases financial assets and real estate and is adversely impacted by an increase in its price once again leads to the under-reporting of general price increases when measured strictly by the CPI.

It is an unfortunate truth that any attempt to create a basket of goods representative of the spending patterns of consumers is doomed to failure since the goods and services in the economy that are available for purchase are not static. As new technologies are introduced they become ubiquitous in daily life and over time come to replace older technologies, but no price index can be manually measured and updated frequently enough to capture the rate of change of constant economic progress. For example, there is typically a time-lag of several years between consumer surveys meant to form the basis for the CPI and the inclusion of these updated spending patterns into the index itself. Even in this interim period, new items and technologies can come to instigate large shifts in societal consumption patterns while old ones are relegated to the dust-bin entirely.

Of course, even if we were to ignore these obvious flaws and inclusion biases inherent in any subjective price index, we would still be left with the additional fallacious statistical data adjustments introduced by government economists all on their own accord. Consider, for example, that when an item rises in price, the government statisticians who calculate the CPI attempt to account for something called “substitution bias”. They assume that when prices rise for specific goods and services, a certain segment of the population will begin to shift spending patterns towards cheaper “equivalent” alternatives. As a result, the composition of the CPI is altered to reduce the impact of those items that happen to increase in price but where statisticians deem an equivalent alternative exists.

If the price of beef rises, for example, government officials assume that some people will begin to substitute beef with the cheaper alternative of pork, and the CPI is subsequently adjusted to reflect the assumption that less people are now buying beef and more people are buying pork. Since the weighting of beef has now as a consequence been reduced in the CPI, this tends to under-report the extent by which the price of beef has risen. And yet there are still those who continue to purchase beef whose costs have now increased by a greater proportion than what is now reported by the CPI. Furthermore, there are countless others who have out of necessity switched to pork who will now be denied the enjoyment of eating beef. The act of reducing the impact of a particular item on the measured CPI precisely because it has increased in price is to introduce a perpetual understatement of the officially-reported rate of economy-wide inflation.

And still the list of questionable data adjustments marches on, for the statisticians who calculate the CPI are also known to “hedonically adjust” various components in the index to account for the fact that goods and services increase in quality over time. By way of example, if the price of a car remains the same from one year to the next, the statisticians deem that the cost of the car has actually decreased since consumers can now purchase a superior car for the same price. This process necessarily involves attempting to place a value on the various improvements that have been added to the car over the course of the year, but as the skeptics have continuously pointed out, exactly what is the value of better tires, improved safety features, or higher-quality engine performance?

By now it should be obvious that the majority of the problems inherent in the CPI stem from the fact that its composition is necessarily subjective. It is not, as some people are lead to believe, an economic variable; but merely a statistical measure subject to questionable data-adjustments that claims to measure price changes for a mythical “average” consumer rather than prices across the economy. Unfortunately, the task of measuring the change in prices for every single good and service in the entire economy is a herculean task which no government statistical department, no matter how large, could ever hope to accurately report. The CPI is the government’s best-guess answer to the question of how to go about measuring the immeasurable. As Albert Einstein once said: “Not everything that counts can be counted, and not everything that can be counted counts”. Inflation is a prime example of one of the most important concepts in modern-day economics where we find there is no precise tool to accurately measure its impact. In essence, by relying on a host of inaccurate government price indices such as the CPI, most of today’s economists are flying blind in this regard.

The fact that there is no definitive way to measure economy-wide prices may strike the reader as a prime case of someone offering a criticism without proposing a constructive solution. Fortunately, we can point to an existing framework for estimating changes in the general price level thanks to those few economists who actually understand both money and inflation. The methodology will be presented in a future article and does not purport to exactly and precisely reflect price levels in real-time, but over anything but the short run it is virtually guaranteed to be more accurate than the widely-reported CPI number that is fallaciously touted today as a valid metric on the state of price levels across the broad economy.