From machines to manicures, what goes into the GDP these days?

BY Steve Landefeld  April 29, 2014 at 4:23 PM EDT
Its critics call the GDP an outdated short-term metric that's ill suited to assessing America's economic health. But it's a constantly evolving measurement, says Steve Landefeld, head of the agency that calculates it, that as of recently, includes the value of movies. Photo by Flickr user Skil Fibber.

Its critics call the GDP an outdated, short-term metric that’s ill suited to assessing America’s economic health. But it’s a constantly evolving measurement, says Steve Landefeld, head of the agency that calculates it, that as of recently, includes the value of movies. Photo by Flickr user Skil Fibber.

Editor’s Note: Gross Domestic Product rules the world. It’s the “king of all statistics,” but shouldn’t necessarily be, author Zachary Karabell told us. The limitations of the GDP have been widely recognized. Robert Kennedy called out its close cousin, the Gross National Product, for measuring everything “except that which makes life worthwhile.” And as Making Sen$e explored, there are alternative measures of economic growth that strive to better capture the economic and environmental welfare of average Americans.

We presented those alternatives to Steve Landefeld, who oversees the organization that calculates the GDP. He’s well aware that GDP will never present a complete economic snapshot of America; that is not its purpose. But as an evolving statistic, the GDP is consistently revised to reflect changes in the economy, he explains. And what about arguments that GDP is inflated by gains not felt by most Americans? Landefeld’s Bureau of Economic Analysis is working with the Census Bureau to put out annual measures of median income and income distribution to supplement the economic picture painted by the GDP.

Landefeld appears in our Making Sen$e segment on alternatives to the GDP, below, and our extended conversation with him follows.


What’s the history of the GDP?

GDP, or the Gross Domestic Product, is the broadest measure we have of transactions in the economy. It includes everything from machines to manicures.

It was started during the Great Depression. At that time, economists and President Franklin Roosevelt had to rely on very partial and sketchy data, from freight car loadings to dime store sales. They had no comprehensive and consistent picture of the economy to discern whether it was investment spending that was lagging or consumer spending.

So a team of economists at the National Bureau of Economic Research and department economists, like a gentleman by the name of Simon Kuznets, later a Nobel Prize winner, put together this wonderful set of comprehensive measures, which later came to be known as the GDP.

They were so successful that in the spring of that year Roosevelt was using it in the formulation of the federal budget.

How has GDP evolved?

GDP is designed for economic policy in business decisions, so it has to change a lot to adapt to circumstances, policy needs and the structure of the economy. At the time of World War II, for example, they went from just recording national income — incomes earned in production — to a GDP number which included all final expenditures because they needed that for World War II planning purposes. Later on, for example, in the 1970s and 1980s, when hyperinflation became much more of an issue, we had to adopt more sophisticated measures of inflation, inventories and count the stocks and the like. When we were going through the high-tech era, we had to improve our measurements of computer software. More recently, as a result of the Great Recession, where we needed to better integrate our financial data and its effect on real phenomena like GDP, we worked with the Federal Reserve Board to do that. So the GDP is this constantly evolving number which changes to meet changes in the economy.

What’s an example of something that didn’t used to be included but now is?

Some of the new things included that go beyond research and development were the value of movies and long-running TV series. The point is that while those are included in GDP in the year in which they are sold, what’s not included is what we actually add to our capital stock by having those movies. The “I Love Lucy” series, or for younger people, “Star Wars,” those go on for years and produce value, so they actually need to be accounted as investment and added to our capital stock, and then we depreciate them over time. After all, we’re the country of Hollywood. We should be capitalizing those kinds of things.

We enlarged our definition of investment. It’s been something that we’ve been trying to do for some time. As innovation became even more important to economic growth – it’s always been important, but it seems increasingly in this information age that it’s more important, international accounting standards embodied this notion that we needed to capitalize research and development. So we went ahead and did that, in concert with our partners at the National Science Foundation and at the Census Bureau.

We’re working in the future to try and measure not only scientific and engineering research and development, but what you might think of as nonscientific or business research and development. You know, the cool look of the iPad, that design kind of thing, is an important component of U.S. competitive research and development and our intellectual capital.

How important has GDP become?

It is used by the Federal Reserve Board for economic policy, and our consumer inflation rate comes from this. The entire federal budget is based on the baseline that comes out of the Bureau of Economic Analysis’ data. It’s used to allocate over $300 billion worth of federal funds. States and localities use it for budget projections. And businesses use it for location decisions because it’s not just GDP — we have an entire set of regional industry and international accounts. And GDP has a large effect on financial markets, exchange rates, stock prices and interest rates – in effect, on every American who takes out a loan, saves for retirement or buys imported goods.

Is it the most important economic measure we have?

Yeah, I think it is. Of course, I sit here at the Bureau of Economic Analysis rather than at the Bureau of Labor Statistics, where they produce the jobs figures.

How hard is it to keep up?

It’s a tremendous process trying to keep up with changes in the economy on a quarterly, annual and trend basis. The fabulous staff here at the Bureau of Economic Analysis does a terrific job. It’s not just these long-term changes in the economy, like globalization, that we’ve got to track; it’s things like the Affordable Care Act, which we’re tracking right now in our data. Earlier, it was the Recovery Act, where we tried to separate paper losses from actual losses as the economy went down in the Great Recession.

How accurate is it?

We like to think it’s the most accurate in the world. It does get revised because it’s an early estimate based on partial data.

Is GDP the best measures of progress?

Well, there’s a lot of controversy about that, and there has been since Kuznets started the accounts. What we aim to measure is the market sector of the economy. It should not be forgotten that a lot of your economic welfare does come from having higher GDP. If I look at countries with higher life expectancy, better health status, education -– all those kinds of things we prize and think about as contributing to economic welfare — they do come from economic growth.

But GDP is not a complete measure?

No, it is not a complete measure. From the outset, Kuznets said it, we’ve always said it: we are not a measure of economic welfare. We are a measure of economic transactions. Many of the things people often cite as missing from the GDP, whether it’s research and development or human capital or household production, we’ve worked on measuring. But because they’re not market transactions with defined prices, they’re very hard to measure.

If I add, for example, household production, which we produced estimates on, we could add 25 to 40 percent to GDP. That’s a very large number. Which estimate I choose can produce a big difference. It’s very hard to value these things, and putting those inexact numbers in GDP would diminish the accuracy and usefulness of GDP for monetary fiscal policy, business decisions, all the things that are currently used to manage the market sector of the economy.

Would you like to see GDP revised further?

The main ways in which we would like GDP revised further are along the lines of better capturing the things that cause economic growth. Way back in the 1950s, Nobel Prize winner Richard Solow discovered that when he looked at economic growth, conventional inputs like labor and capital explained about half of growth. So a lot of it’s unexplained. Over the years, we in GDP accounting have been trying to reduce that unexplained portion of growth.

Other things that we’d like to include would be in the form of satellite accounts. Household production is terribly important. We now have an American time use survey in the United States and we’ve produced some estimates. They should become a regular set of estimates, resources permitting.

Finally, distribution of income may be the answer to this problem of GDP per capita going up and most Americans feeling down. GDP per capita is raised by the high end, but many Americans may not be experiencing that gain. We need a better measure of median income and the distribution income, which would be put out perhaps once a year with GDP. We are currently working on that and trying to address gaps in the tax and household survey data, which are present in most estimates of the distribution income. This is joint work we’ve been pursuing with the Census Bureau. And we hope by the end of this year to have some estimates out on that.

Are we overly reliant on GDP?

It’s very important, but you don’t want to rely on any one indicator, including the GDP, for measuring these things. You’ve got to have different indicators. The problem becomes people want aggregate numbers. From the Organization for Economic Cooperation and Development, there’s the Better Life Index. We have the Maryland Genuine Progress Indicator.

What’s your opinion of some of these alternate measures?

The subjectivity is the Achilles heel of it. Maryland’s GPI has 20-some indicators, for example. What weight do I put on each? How much do I subtract for the commuting time? And if I enjoy my commute, maybe it’s one thing, maybe not. I have a problem too with the fact that as an accountant, if I subtract something, I need to put it in before I subtract it. So if I clean up the environment, that should count as a plus, and then the depletion of it should count as a minus. In other words, these alternative measures end up being systems of indicators that are troubling for an economist who’s trying to put together objective accounts, or at least not make normative judgments about what should be.

What about well-being measures?

That research should continue, but the difficulty with them is that they are self-reported. And as you look at them over time — whether it’s wars, recessions, whatever — they don’t move very much. And a lot of that has to do with the fact that people adapt. There’s this famous example where if you ask sighted people what their happiness is on a 1-10 scale, they answer maybe 6, 7 or 8. And then you ask them what would happen to their happiness if they went blind, and they say that it would be a 2 or something. But if you ask people who are blind, who have accommodated to being blind, their happiness is about like other people’s happiness.

Looking at the Lehman Brothers’ collapse and the recession, that happiness dropped like a rock, but within about a year, it was pretty close to where it was before, despite the fact that U.S. unemployment was still up around 10 percent. Most Americans had lost almost half of their life savings at that point. It had not been rebuilt by increases in home prices or equity values. So you see this adaptiveness even in economic variables.

Subjective wellbeing indexes are perhaps a better measure when they’re targeted to specific things, like perhaps a disease intervention or a particular public policy over a short-time period. But I am not convinced that they can be useful as an aggregate measure that would act as a replacement or a complement to GDP easily. The better route to go is the use of distribution of income as a complement to your regular GDP estimate, so we can see not only how total GDP is doing, but what how median households are doing.