Monetary Policy

How Official Statistics Underestimate Inflation

(Igor Kutyaev/Getty Images)
The CPI shows low inflation, but it doesn't tell the full story.

In the wake of unprecedented federal deficits and central-bank accommodation, market watchers are growing wary of inflation. The Fed projects no long-term inflationary pressure above its 2 percent target rate, but investors are split. While the consensus resides with the Fed, a nontrivial minority of investors are forecasting 3-4 percent inflation and a sharp rise in interest rates.

Lost in this debate is a deceptively simple question: What is inflation and how is it measured?

Start with the issue of changing behavior. Within the consumer-price index (CPI), the standard measure of inflation, much of the so-called New Economy is missing or miscalculated. Today’s economy is driven by data, a good invisible to the CPI. Google, for instance, generates $182 billion in consumption revenue, but because its products are essentially free to consumers, that spending does not show up in inflation data.

The more serious problem is the government’s persistent bias to underestimate inflation. Social-security payments are indexed to the CPI, with tax brackets automatically adjusting for inflation. An increase in inflation would raise already-ballooning entitlement costs.

Then there is the cost of rising interest payments on public debt. A permanent 2 percent increase in measured inflation, if translated directly into rates, would more than double the government’s interest expense from $350 billion to $750 billion, consuming almost 60 percent of all federal discretionary outlays and exceeding total spending on national defense.

So the government has an undeniable incentive to underestimate inflation. But an incentive to misbehave is not the same as proof of misbehavior. For that discussion, multiple sources of underestimation bias bear attention.

Statisticians within the Bureau of Labor Statistics (BLS) continuously adjust the basket and methodology for comparing cost, quality, and substitutability of goods and services in the CPI. Sounds fine, except for the inherent opportunity to manage the process. An independent calculation of inflation based on prior methodologies reveals the potential for abuse, equal on one measure to a persistent 2 percent annual inflation undercount.

Even if the methodology changes are truly neutral, a look inside the index reveals imbedded distortions. The problem is most evident in housing, which also is the single largest component of the CPI, accounting for 42.3 percent of total measured costs. According to the BLS, rental housing accounts for 7.8 percent of total index costs. It is the imputed cost of owned homes, which constitutes a massive 24.2 percent of the index, where the error is manifest.

BLS statisticians want to know the monthly equivalent rent a homeowner would charge to rent their home. Literally. One quarter of the CPI is calculated based on the following question posed monthly to homeowners, “If someone were to rent your home today, how much do you think it would rent for monthly, unfurnished and without utilities?

Even as the method seems absurdly subjective and unscientific, how would one independently test the result for reasonableness? Rent-equivalent expense is necessarily correlated with the price of housing, which is measured monthly by the Case-Schiller index.

In the past twelve months, Case-Shiller housing prices appreciated 10.4 percent. By contrast, CPI owned-home rent-equivalent costs are calculated to have increased a mere 2 percent. The 8.4 percent difference would, of its own, have more than doubled the total reported CPI, from 1.7 percent to 3.7 percent.

The same story is evident in medical costs. In the CPI, medical care accounts for 8.9 percent of the total index. Yet within GDP, health-care expenditures total 17.7 percent of the economy. Why the difference? The CPI excludes medical costs paid through employer insurance premiums, even though those costs eventually are passed on to consumers. Nor does the CPI include any tax-funded medical care, including Medicare Part A and all of Medicaid.

Independently, private insurance premiums have soared in the past few years, doubling or more for many consumers. Yet the government index claims health-insurance premiums since 2013 have increased by only 45 percent. The difference is due to method: The CPI indirectly estimates insurance premiums “based on retained earnings method,” which the BLS explains as “leftover premiums income after paying out benefits.” Nor does the complex methodology used to estimate medical costs account for the fundamental irrationality of the U.S. health-care market, where standard back surgery, as an example, can “cost”, out-of-pocket, anywhere from zero to $150,000.

For an unbiased look at medical cost inflation, PwC calculates that U.S. medical costs have increased an average 6.1 percent annually since 2014. Over this same period, the CPI medical care increase is 2.8 percent.

Adding to the problem, the Fed and economists pay primary attention to “core inflation” which is defined as the CPI excluding energy costs and food. The theory is that food and energy prices fluctuate due to forces independent of inflation. As a result, the 50 percent jump in energy prices since the 2020 election does not count in the Feds’ inflation assessment.

Whatever the historical merits of this position, the climate-change agenda of the Biden administration will systemically and materially increase energy costs. Germany’s premature pursuit of “sustainable energy” has resulted in $0.37/kwh electric costs, more than triple current U.S. average electric costs of $0.12/kwh. On the present course, U.S. energy costs will be systemically inflationary and cannot be excluded from core inflation, as in the past.

Pieced together, it is plain to see that reported CPI inflation will inevitably be “tame.” The Fed excludes food and energy from consideration. The index controls housing and medical cost through methodology. Highly competitive goods such as apparel, cars, and household furnishings present no material inflationary risk. Together these categories account for fully 82 percent of the CPI. The 18 percent balance is for services, primarily education and communications (6.3 percent), transportation (5.1 percent), and recreation (3.7 percent), which, tuition aside, also present no material inflation risk. It is a rigged game.

To see where inflation truly resides in the present economy, consider which of the following are CPI inflationary: a rise in stock prices, bitcoin, income taxes, art and collectables, or a decline in the dollar? The answer is none of the above.

Economists in the standard view consider asset-price increases as impacting savings and investment, not consumption. Income-tax increases reduce consumption but by definition, are excluded from consideration in the CPI.  The value of the dollar affects purchasing power relative to imports but is neutral for domestic consumption.

The difficulty is that the past provides no analog to the current tax and spending orgy: $7.4 trillion in Federal Reserve printed money, the recent $1.9 trillion spending boondoggle, a forecast $10 trillion one-term Biden administration budget deficit, including $2.3 trillion in planned “infrastructure” outlays, and sharply higher planned income and other taxes.

This gets to the real definition of inflation. In the present economic environment, are you willing to hold cash and U.S. dollars, confident that their buying power will be maintained in the future? Here the answer from the markets is a resounding no.

Investors are doing everything in their power to avoid holding cash, which accounts for the historic bubble seen across all asset classes — stock markets, housing, Bitcoin, art, collectibles, really anything except dollars of eroding value. Equally, investors are increasingly unwilling to hold the U.S. dollar in relation to other currencies, as seen by the one-year 10.1 percent decline in the dollar against all currencies.

Fed Chair Powell blandly promises that inflation is no threat. By the refusal to hold cash, investors demonstrate that inflation is far higher than measured by the CPI and that the threat is growing.

The ultimate puzzle of inflation is that it depends on expectations. Monetary authorities worldwide have sold the proposition that inflation is not a problem. Properly understood, markets worldwide say it is a lie. At some point, impossible to predict, this tension will resolve itself. Unwinding of the inflation deception, hiding in plain sight in the price of assets and the value of the dollar, will be brutal. Its name is stagflation.


The Latest