Inflation or Productivity Loss?

I’m seeing a lot of online chatter about inflation lately, and naturally the commentary has the stink of politics on it. Republicans are trying to pin the price increases on Biden and the Democrats. The Democrats, in turn, are trying to minimize the problem by denying that the inflation is as bad as it seems — currently 6.8% ^[1]CPI-U, all items, per the December 10 BLS report., higher than it’s been at any other time this century.

There’s a lot of confusion about just what we mean by inflation. Some Biden critics are pointing at the high prices of cars and trucks, and the high price you have to pay to fill their tanks at the pump. Others point to the high costs of meat. There’s no doubt that these price increases make life more difficult, but these kinds of single-commodity price spikes are not what we normally mean by inflation.

Modern use of the term inflation refers to a sustained, wide-spread, across-the-board increase in prices, including the price of labor (wages and salaries). Everything goes up, all together. But those prices are expressed in US dollars, so it’s just as meaningful to say that the value of the dollar is going down. Sometimes referred to as monetary inflation, this is usually what we mean by inflation.

More limited types of “inflation” — increases in the cost of cars and trucks, meat, or gasoline — are understood these days to be the result of the ordinary changes in the prices of goods and services. If a major American oil refinery shuts down, reducing the supply of gasoline, the end users and intermediaries will bid up the price of the remaining supply, causing prices at the pump to increase. And if adaptations to the the COVID-19 pandemic make meat processing facilities less efficient, this will increase the labor and capital costs of getting meat into grocery stores, and grocers will have to raise prices to cover those costs. These changes have nothing to do with inflation: The US dollar didn’t lose value, oil and meat just became more expensive.

Measuring inflation is difficult. The measurement everyone’s talking about these days is the Consumer Price Index for Urban consumers (CPI-U). Every month, the Bureau of Labor Statistics gathers information from all over the U.S. about the store prices of over 200 representative products, weights those prices based on how much each product is consumed by households, and calculates an aggregate statistic representing the current cost of living. By looking at how much prices are changing each month, the BLS can estimate the inflation rate. That’s where the current figure of 6.8% inflation comes from.

There are several known problems with the Consumer Price Index. For one thing, while it’s easy to walk into a grocery store and look at prices, it’s a lot harder to get information about the volume of products sold. Consequently the weights applied to consumption are only adjusted every two years. This can cause the index to miss substitution effects. That’s what happens when people switch products because the prices have changed. For example, assume you buy a certain quantity of apples every week. If apples suddenly get much more expensive, you might realize that you don’t really want apples, you want fruit, so you switch to pears instead. If apples are tracked by the CPI, then BLS statisticians will calculate that your cost of living has gone up, when you actually avoided that increase by switching to less expensive alternative products. The CPI calculations will eventually catch up, but in the meantime, the CPI will overstate the cost of living and therefore the inflation rate.

Economists have been aware of this problem for decades, but it’s not easy to find a solution that doesn’t have other problems. There has been some tinkering with the way CPI is calculated over the years, but the biggest change is probably the use of Chained Consumer Price Index (C-CPI-U), which tries to detect substitution by re-weighting prices every month instead of every two years. In addition, there are other index such as the Personal Consumption Expenditures Price Index (PCEPI) or the GDP Deflator, which attempt to calculate similar indices in different ways.

By the way, the choice of index is more politically charged than you might think. Every time the BLS has adjusted the methodology for computing CPI, hard-money fanatics accuse them of trying to hide the “true rate” of inflation. Furthermore, things like Social Security benefits increase every year based on the CPI-U and switching to C-CPI-U would mean that benefits don’t go up as much each year. Seniors argue quite reasonably that they spend proportionately more on healthcare, so they benefit less from substitution effects than the general population. (The BLS also computes a CPI-E for the elderly, but there’s little enthusiasm to switch to CPI-E for some reason.)

I’ve tried to find out more about how the various index values are calculated, but it all becomes a blur — I can read the descriptions, but I don’t know enough economics or statistics to really understand what they mean. They differ on which products are surveyed, which expenditures count, how prices are weighted, how known seasonal cycles are adjusted for, and how all the data is boiled down to a final number.

I have no clue which index is best as a measure of inflation, but as it happens, they’re not too far apart. Here are values for all the indexes I’ve mentioned:

• CPI-U: 6.8%
• C-CPI-U: 6.7%
• CPI-E: 6.2%
• PCEPI: 5.0%
• GDP Deflator: 6.2%

Perhaps a more fundamental problem is figuring out if these price indexes are really measuring inflation. There’s kind of a chicken-and-egg problem here: We price goods and services in dollars, but dollars are only as valuable as the goods and services you can get for them — a dollar that couldn’t buy anything would be worthless. So when prices go up, is that because goods and services cost more? Or because the dollar is worth less? Only the latter is inflation.

Under normal circumstances the prices of some goods and services will rise and others will fall. Although these price changes are not really random, neither are they coordinated with each other. Some go up while others go down as the market responds to supply and demand changes. On average, these fluctuations should tend to cancel each other out to some extent. Furthermore, our economy’s methods of production and distribution have been getting more efficient every year.^[2]This is a recent development in human history, starting about two centuries ago. This means that the real costs of goods and services generally fall slightly every year.

So whenever there’s a widespread, across-the-board increase in prices, we have to ask ourselves which is more likely? That everything got more expensive at the same time? Or that inflation has reduced the value of one thing, the dollar, causing items sold for dollars to go up in price? That latter explanation seems much more plausible, which is why it seems justified to treat CPI-U and other price indexes as measures of inflation.

But these are not normal times. The entire world has been suffering under the COVID-19 pandemic for almost two years. Businesses are operating under special rules — social distancing, masking, etc. — that may impair their productivity. And there have been huge changes in consumption habits as many Americans have been working from home, ordering more deliveries, and so on. Our production capacity was developed to produce a different mix of goods and services from what we are consuming now, so it might not be working at previously high levels of efficiency.^[3]This, not hoarding, is the true reason for the toilet paper shortages at the start of the pandemic. The soft fluffy stuff you buy for your family is a substantially different product from the scratchy crap your employer was buying. It took time to make more of the good stuff.

It seems to me there’s a real possibility that our economic productivity has declined over that past two years. This would make everything more expensive, and it seems to me there’s a real possibility that our price indexes are reflecting that. In other words, the unlikely possibility happened — everything did get more expensive at the same time — and we are fooling ourselves into thinking it’s inflation, when it’s really a loss of productivity.

This might sound like I’m defending the Biden administration (“The inflation rate is not so bad…”) but it’s important to realize that a decline in productivity is much, much worse than inflation.

To see why, suppose Congress passed a law saying that at the stroke of midnight on New Year’s Eve, every person would get ten times as many dollars as they have, and all payments, wages, loans, and contracts would be adjusted to match. A family making $100,000/year on December 31, with $30,000 in the bank and a $250,000 mortgage on a $300,000 home would wake up on January 1 earning $1,000,000/year, with $300,000 in the bank and owing $2.5 million on a $3 million home.

It’s pretty obvious that the family is neither better nor worse off than before.^[4]Yes, I’m assuming tax brackets were also adjusted. They have 10 times more money, but everything they buy will cost 10 times as much. Fortunately, they also earn 10 times as much and have 10 times the home equity. The economy is still producing the same goods and services as before, they are all just priced 10 times higher. The nominal prices have changed, but everyone is able to consume exactly the same goods and services as before. Their economic well-being has not changed. Inflation — even 10-fold inflation overnight — is not in itself harmful.

On the other hand, if economic productivity declines, there’s less stuff to go around. Our civilization is impoverished. Goods and services are no longer being produced at the same quantities, and everyone is forced to consume less.^[5]Or work more, which is also bad. And we are all worse off for it.

That’s not to say we shouldn’t worry about inflation. For one thing, the change in the dollar’s value is never as coordinated as in this example. Prices at the grocery store may go up before wages do, and an increase in wages doesn’t automatically increase your savings, and bank interest rates may take even longer to go up. Meanwhile, fixed-rate mortgage loans will lose value because they can be paid back in cheaper dollars. That’s great if you have a mortgage, but it will send the industry into another tailspin.

When inflation is steady, we can plan for some of these changes, but an unexpected increase in the inflation rate can disrupt markets enough for the effects of inflation to spill over into the real economy of goods and services.^[6]And then there’s the possibility of devastating hyperinflation. But a productivity decline, by definition, always hurts the real economy. It’s never not a problem.

I don’t know how much of my story to believe. It certainly seems plausible that at least some of the increase in inflation is due to a decline in the real economy that is being misread by the CPI-U and other indices.

If so, it’s hard for me to say if the President is to blame. But there are certainly ways in which our presidents have not been helping. At least part of the price increase for cars and trucks (and home appliances) can be blamed on deliberate policy choices, such as Donald Trump’s idiotic steel tariffs, which drove up the cost of domestically produced vehicles and appliances. Of course, since our new President has delivered on his campaign promises about international trade, those are now Joe Biden’s idiotic steel tariffs (and quotas).

Other than that, I don’t know what’s causing the increase in prices. I’m hoping it turns out to be inflation, because we have some pretty good ideas about how to fight inflation. (Whether we have the willpower is a different matter.) Productivity, on the other hand, is mostly a mystery. We have some ideas about why it started growing so much faster a few hundred years ago, and we know that disasters like wars can destroy productivity, but we don’t know much about how to keep it growing, or what to do if it stops.