With the holiday shopping season, we’ve seen a revival of the confident but completely wrong assertion: ”Consumer spending is 70% of U.S. economic activity.” No matter how many times I stab this vampire meme in the heart, it just keeps getting back up again. There are two reasons why it’s simply not true that consumer spending is 70% of economic activity:
* Consumer spending, especially this time of year, includes a lot of imported goods. So when you buy a toy, a shirt or a big screen TV, just flip it over and look at the label– your money flows out of the country to stimulate economic activity in China, or Korea, or Mexico.
* The government’s definition of consumer spending includes Medicare, Medicaid, and money spent by nonprofits such as political parties and religious groups.
So when you see pictures of Americans buying toys, clothing, and consumer electronics, think about happy workers around the world. U.S. consumers drive global economic activity.
For some of my other posts on the subject, see here and here.
Hall of Shame–news organizations who got it wrong.
For at least 15 years, 70 percent of economic activity in the U.S. has come from consumer spending. Atlanta Journal Constitution
Consumer spending accounts for 70 percent of economic activity. Associated Press
Retail sales are closely watched as a measure of consumer spending, which accounts for about 70 per cent of economic activity. Financial Times
Please follow up on this by giving a better estimate and definition: maybe “X % of the economy” is too vague. Perhaps it would be better to say, “Considering that so much of consumer spending ends up in foreign countries, consumer spending contributes only Y % to the U.S. economy.”
Still, it’s a significant percentage, isn’t it, and a necessary element of a recovery? And taking this closer to where it should go, how about comparing the dollar-for-dollar effect on U.S. economic activity, and jobs, of consumer spending vs. government spending on specific projects that necessarily hire mostly American workers?
Hi Urban
I give a range because the government’s input-output tables do not distinguish between imports and domestic-produced goods.
Wait a minute, you can’t just subtract induced consumption and claim that part doesn’t count but the other components of GDP are just fine. The most volatile part of GDP investment, which is also largely an induced quantity. Part of the changes in investment are induced by changes in consumption, which means that the non-induced portion of consumption has an impact beyond what is directly measured. You can’t just say that chickens hatch from eggs and therefore the egg comes before the chicken.
Also, import demand has repercussion effects, as well as exchange rate effects (which induce exports), so subtracting the entire imported portion of consumption is misleading with respect to the impact on domestic demand.
I also don’t see why one should not include consumption of in-kind transfers such as Medicare in the aggregate of consumer spending. Whether it’s relevant depends on what question one is asking.
If one wants to ask how important consumer spending is as a driver of economic fluctuations, then the whole percent-of-GDP statistic is virtually irrelevant, no matter how it is defined. What matters is (1) the variation in consumer spending, rather than the level and (2) the extent to which variation in consumer spending ultimately influences variation in other categories. On the basis of (1) consumption doesn’t look very important, since it tends to be quite smooth over the business cycle. On the other hand, it’s very difficult to tell what’s going on with (2). I don’t think one can rule out the view that variation in US GDP is largely determined indirectly by variation in US consumption.
Of course it’s true that, given a choice, it would be better to increase exports than to increase consumption, but increasing exports is a particularly tall order when one is a very large economy and supplier of the world’s primary reserve currency.
Hi Andy,
The short answer to (2) is that a) it depends on what kind of consumption and b) we don’t know. When you go and buy an imported shirt or a toy, roughly 40-60% of your money goes to the domestic retail-wholesale-transportation chain. The remainder flows overseas. On the other hand, if you go out and get a haircut, or pay for medical care, much more of that spending stays in the country.
The reason why we really don’t know is that the government’s input-output tables do *not* distinguish between imported and domestic-produced goods. So, for example, there’s no way of knowing whether the computers you buy in Best Buy have more or less imported content than the big servers that Wal-mart uses to track its inventory. That’s why I had to use a range.
How would you provide a benchmark against which one could compare consumer spending?