You Cheatin’...CPI, Oh Why Did You...

Long ago, we adapted to an inflationary world. The cost of everything from hamburgers to tuition faithfully rose each year. Then, in the ‘70s, as consumers complained about rising prices and an eroding standard of living, companies, labor unions and the U.S. government began to tie raises and benefit payments to an official inflation measurement called the consumer price index.

Inflation is now taken for granted. Although there has been less of it lately, we still assume that prices will creep up. With a collective sigh, we blame inflation for the trouble people have in living as well as their parents did. Indeed, the official numbers say that personal income is growing far slower than inflation, even today.

But wait.

What if everything Americans think they know about inflation were wrong?

What if the consumer price index, which guides cost-of-living adjustments on everything from Social Security payments to rents, systematically overstates inflation?

Wouldn’t that mean that billions of dollars in government benefits are misdirected? Wouldn’t it mean that some Americans are unfairly enriched while others are unjustly impoverished? Wouldn’t it mean that Americans as a whole might be somewhat better off than they think?

A group of the nation’s top economists says yes, that billions have indeed been misdirected and that the inflation rate may be overstated by as much as 2 percentage points each year. And these otherwise staid academicians are waging battle to bring the CPI--arguably the world’s most quoted economic index--in line.

Maybe economic debates seem dusty dry. But the stakes of this battle are enormous, with major political and financial repercussions. Everyone is affected.

That’s because the consumer price index, launched in 1919 to help set wages for dock workers, now exercises profound economic muscle. It determines how much to pay recipients of Social Security, disability and Supplemental Security Income. It affects the size of retirement benefits, monthly stipends for food stamp recipients and also subsidized school lunches.

It’s used to set wage rates. Landlords use it to determine rents. The Internal Revenue Service uses the consumer price index to adjust tax brackets, personal exemption rates and standard deductions. The Federal Reserve Board uses it to monitor--and influence--economic policy, so it has a ripple effect on loan and savings interest rates. Stock and bond brokers here and abroad watch it to measure the health of the U.S. economy and regard it as a harbinger of future interest rates.

Indeed, the power of the index is the very reason economists say its inaccuracies cannot be ignored.

“What we are doing are taking money away from present and future taxpayers and redistributing it to recipients of Social Security, government retirees and other people who have their wages indexed to inflation,” says Robert Gordon, economist at Northwestern University in Evanston, Ill. “It’s tremendously important to fix this.”

To be specific, retirees, government workers and welfare recipients whose benefits and wages are tied to the consumer price index are being rewarded with somewhere in the neighborhood of $90 billion a year more than they ought to be getting if their incomes were tied to the true cost of living, Gordon says.

It’s important to note too that Congress is now in a virtual stalemate because of the difficulty in cutting entitlement payments in order to reduce the yawning federal budget deficit. Yet fixing the index could do the job with minimal political pain. A blue-ribbon panel of economists appointed by Congress and led by Stanford University’s Michael Boskin estimates that fixing the CPI could beat back the federal deficit by an astounding $140 billion a year, presumably resulting in lower interest rates and a stronger economy.

Notable too is that almost every economist who has studied the index agrees that the CPI overstates inflation--and that includes economists at the Bureau of Labor Statistics, the government agency that compiles the index, says bureau economist Patrick C. Jackman. Even the American Assn. of Retired Persons, whose members could be the biggest losers in a CPI revision, agrees that inaccuracies in the index should be fixed. The problem is, fixing this index won’t be easy.

To understand why, you have to understand quite a bit about the index itself.

It is not a “cost of living” index. Rather, it’s a measurement of prices for a wide range of goods purchased by urban consumers. Why are those two things different?

Here’s the explanation:

Every week, about 400 government shoppers all across the country pop in their cars and visit a series of grocery and department stores, auto showrooms and gasoline stations in an effort to record prices on some 80,000 items. Each shopper has a set list of items and a list of retail outlets to visit or call. Each visits the very same retailers and jots down the prices of the very same items month after month, year after year for a period of five years. At that point, about a fifth of the retailers in the index are changed, so some shoppers will get new assignments.

After checking for mistakes, statisticians put the figures into a complicated formula that gives each item a weight that corresponds with its relative importance to the “average” consumer. The bureau’s average consumer, for example, spends nearly twice as much each year on apples than on canned fish. Consequently, a change in apple prices would have a bigger effect on the consumer price index than a similar change in the cost of a can of tuna.

To determine just what the average consumer buys so each item can be weighted, the Bureau of Labor Statistics periodically does an extensive survey of consumer buying habits in which roughly 40,000 consumers submit data about what and how often they buy virtually every item you can imagine. From that the bureau creates its “average” consumer, whom we’ll call Buffy.

Yesterday’s Shopper

So what’s the problem? There are several serious ones, according to the economists.

The simplest is mathematical. Because of the way the index is computed, goods that tend to rise and fall in price may be given the wrong weight when they’re introduced to the index. The calculation is complex. But the result is simple: An item that’s temporarily cheaper or on sale when it goes on the list will leave a residual mark of inflation when it returns to its normal price.

The bureau is already working on fixing that glitch. However, the Boskin panel estimates it could have accounted for half a percentage point of each of the annual upward moves in previous years.

Three of the other four problems revolve around a single issue. Simply put, the consumer price index measures what things cost, not what people buy. And, at various points in time, the things people are buying are so different from the goods that are being priced that the index substantially misstates the cost of living as most Americans experience it.

There are several reasons people don’t buy what’s being priced by the index.

One is that the index doesn’t price everything. It prices what the bureau’s average consumer, our Buffy, would buy. And Buffy is stuck in the early 1980s--a time of rampant inflation and high interest rates. It’s also a time when newfangled products like personal computers were purchased only by the sort of techno-nerds Buffy refused to date in high school.

Indeed, today’s consumer price index--the products it prices and the weights that products are given--are based on Buffy’s shopping basket of 1982, ’83 and ’84. As a result, the cost of a pack of cigarettes has more power to move the index than the cost of a computer. And, because of the strict way CPI shoppers must check the same items year after year, the computers that are being priced today are relics.

Why not track the expenditures of a ‘90s family? Because determining how Buffy’s habits have changed costs money--lots of it. The tens of millions of dollars required to do the extensive survey--that is, track the buying habits of the 40,000 consumers Buffy represents--is appropriated by Congress just once every 10 years. Then it takes several years to analyze the data.

The good news is that the Bureau of Labor Statistics is now in the midst of just such a survey. The bad news it won’t put the new data into the index until 1998. By then, Buffy will already be three to five years behind the times.

To put it bluntly, without some serious government expenditure, the bureau’s Buffy will always be yesterday’s shopper.

What’s most troubling about that, from an economist’s point of view, is that the price of a newly introduced good (or service) tends to be higher at the start than it is 10 years later, when Buffy starts to pop them in her basket. The index misses the initial--often steep--price declines, leaving a misleading picture that ignores all those who reaped the rewards of falling prices.

Buffy Buys Retail

But there are more problems with Buffy than simply her nostalgic tastes.

She’s also a static consumer in a fluid world. Buffy spends exactly the same percentage of her budget on precisely the same objects each week, no matter what they cost. You and I, on the other hand, sidestep a lot of inflation through substitution of items and stores.

For instance, you might wander into the market planning to pick up some breakfast cereal, a head of lettuce and a roasting chicken, but you see that lettuce prices have soared and the maker of Froot Loops has boosted its profit margins just a fraction beyond what you’re willing to tolerate. You buy broccoli and the generic brand of sugarcoated O’s instead. But there’s a sale on chickens, and they’re relatively cheap. You buy two: cook one; freeze the other.

Your grocery bill may drop as the result of these choices. But Buffy doesn’t substitute. She doesn’t bargain shop. She buys the 12-ounce can of Green Giant green beans no matter the price.

By and large, Buffy has also been insulated from the warehouse store trend that’s swept most of the nation into discount supermarkets and department stores such as Price-Costco and Wal-Mart. Buffy buys retail.

This problem could be fixed with a formula that assumes everything is easily substituted, but that wouldn’t be the answer either.

In practice, even items that can be substituted sometimes aren’t for reasons other than price. You may refuse to substitute broccoli for lettuce for no better reason than the fact that you’re itching for a salad. And brand loyalty--and taste preferences--might keep you from switching to a lower-cost version.

But the most serious problem index-makers face may be simply that it’s tough to track prices over long periods in a vibrant society in which new--and often better--goods hit the market each day, replacing old products that have price histories.

If researchers treat all these new products as if they’re the same as the old, the price increases would be overstated. It’s not fair, for example, to compare today’s radial tires to yesterday’s bias-ply models. Although today’s tires are more expensive, they last longer and are safer. Economists have to somehow discount the price rises on the new items to reflect the change in quality.

That’s not easy to do, particularly when the new product is vastly different from what came before. Consider air bags. E-mail. Microwave ovens.

Government economists do have a system designed to compare the prices of new goods to established products. But the process is necessarily subjective, and many experts maintain that it simply doesn’t go far enough to reflect gains in quality, durability and convenience. As a result, what America believes is a gauge of a static cost of living may actually reflect the cost to maintain a rising quality of life, Stanford’s Boskin says.

There is a small group of contrarians. Some argue that the cost of mandatory, universal quality improvements--such as auto safety devices and cleaner gasoline--should be seen as genuine inflation because consumers have no choice about them. Or some people may not see them as improvements. In addition, a small number of economists argue that new burdens that must be borne to live in the modern world--from home alarms to educational software--have an upward influence on prices that should be seen as inflation.

Regardless, the great majority of economists agree that the CPI has risen too fast.

Costly Mistakes

The bottom line: Analysts believe that the consumer price index overstates the true rise in the cost of living by anywhere from 0.2 to 2 percentage points per year.

That means that the entire basis for gripes about declining living standards is inaccurate. In fact, economists estimate that real, inflation-adjusted, income has outpaced inflation in almost every year, leaving today’s consumers considerably better off than their 1960s counterparts.

But more important, mistakes in the consumer price index cost the U.S. government a pretty penny.

Consider what’s happened to average Social Security benefit payments since Congress indexed benefits to the consumer price index back in 1972. In that year, the maximum retirement benefit you could collect was just $216.10 a month. Today, it’s $1,199.10.

Precisely how much of the increase is justified by true inflation, as opposed to phantom inflation caused by glitches in the index, is debatable. Even a 1-percentage-point overstatement in the index has enormous implications. The Boskin panel estimates that each percentage point translates into $98.3 billion in federal entitlement payments, including Social Security.

Tax revenues, meanwhile, could rise because federal income tax brackets, standard deductions and personal exemptions are also all boosted based on the CPI. If the CPI change has been too high, some people did not go into higher tax brackets even when they got real after-inflation raises.

Needless to say, there are plenty of people who would like to see the CPI stay just the way it is. However, Northwestern’s Gordon maintains that the ills of such an approach far outweigh the benefits.

At various points in time, particularly in times of high inflation, the consumer price index can overstate true inflation by several percentage points, he says. Since this index is used by the Fed to set monetary policy, the Fed may react by tightening the money supply, which leads to higher interest rates, slower growth and, often, millions of lost jobs.

Right now, no one professes to have the perfect answer as to how to fix the consumer price index. But by June, Boskin’s panel will issue a series of recommendations. And, thanks to Washington’s budget-cutting fervor and the obvious help a lower CPI could lend to that effort, the index might for the first time this century get a major revamp.

Still, even then it’s likely to have a few flaws.

“There are no perfect ways to measure inflation,” says Boskin. “There are only better and worse ways to measure it. We think we can do better.”

(BEGIN TEXT OF INFOBOX / INFOGRAPHIC)

True Measure of Inflation?

The consumer price index is often thought to be synonymous with inflation. Actually, the CPI measures only certain costs of the urban consumer. Annual CPI since 1959:

(please see newspaper for chart)

1995: 2.4% (does not include fourth quarter)

Source: Commerce Deparment

(BEGIN TEXT OF INFOBOX / INFOGRAPHIC)

Price Ups and Downs

Here’s how prices have changed for selected items over the last 15 years, according to the Bureau of Labor Statistics. As the quality of an item improves, the bureau tries to determine what percentage of the price increase is due to quality rather than pure inflation. For example, new cars may cost 54% more than they did in 1980, but the bureau has adjusted the percentage increase to account for improvements in quality.

*--*

Item Total inflation* Avg. annual rate of inflation College tuition 259.7% 17.3% Hospital room 259.3 17.2 Oranges 231.2 15.4 Tobacco & smoking 214.0 14.2 Auto insurance 185.6 12.4 Cereal 143.8 9.6 Lettuce 137.2 9.1 Newspapers 124.2 8.3 Local phone rates 117.8 7.8 Postage 110.4 7.6 Canned soup 98.9 6.6 White bread 91.9 6.1 Men’s haircuts 79.3 5.3 Baby food 77.7 5.8 Pork chops 68.0 5.0 Beer 65.7 4.4 Bananas 64.8 4.3 New car 54.5 3.6 Rice and pasta 48.7 3.3 Coffee 46 3.1 Men’s jeans 41.4 2.7 Whole milk 38.2 2.5 Carbonated drinks 33.3 2.2 Kids shoes 30.1 2.0 Refrig./freezer 26.3 1.8 Women’s dresses 24.3 1.6 Power tools 22.1 1.5 Ground beef 9.1 0.6 Tires 3.8 0.3 Fuel oil - 5.2 -0.4 Interstate phone rates -12.1 -0.8 Television -36.8 -2.45

*--*

* Figures are from 1980 through October 1995 and are rounded.

Source: Bureau of Labor Statistics