Imagine you are driving on the interstate. Suddenly the cars around you accelerate, and you join them. Scenery whizzes by. You look at the speedometer and it says your car is going 45 mph. What is going on? What you should believe: the speeding-by scenery or the speedometer?
We face a similar predicament when it comes to gauging the speed at which the U.S. economy is growing. Looking out the front windshield, we see a lot of rapid change: artificial intelligence, robotics, genomics and two-way mass media. But our statistics — the economy’s speedometer — claim that growth is unusually slow. The mismeasurement creates a significant risk that businesses, consumers, investors and policymakers will make decisions based on inaccurate information about market activity, inflation, and productivity.
The measurement of the U.S. economy – known as gross domestic product or GDP – has been in the spotlight in both regular and social media because of news that the economy grew at what is now considered a blistering annual rate of 4.1 percent in the second quarter. The Bureau of Economic Analysis or BEA — a section of the Department of Commerce — is in charge of producing the GDP numbers, using data from the national income and product accounts. The BEA’s July 27 report, which got all the attention, indicated the spring economy grew at the fastest pace since 2014.
But this is in the context of the very slow GDP growth rates of the 21st century. During the last quarter of the 20th century, during nonrecessionary periods, the average growth rate was 4.1 percent! How did an average growth rate gain the appearance of blistering growth?
As an economist at the Federal Reserve Bank of Philadelphia, I have been researching the question: How are new inventions in the free-media sector being recognized in our government economic statistics? The answer is, not very well. Simply put, we are trying to measure information-age output using industrial-age data gathering. I will explain later why this is not entirely the BEA’s fault. First, let me explain how free-media products used by millions of Americans tend to be overlooked in the GDP report.
Consider one industry that has changed radically in the past 20 years. Before the Internet, newspapers paid reporters, photographers, and editors to gather and edit the news. Then printers set up presses to churn out thousands of copies of paper that were distributed by truckers to newsstands and stores or delivered by paper boys and girls to individual subscribers who paid for their papers. All of these steps involved costs that could be tallied up and fed into GDP.
But in this new information age, newspapers — sorry, media outlets — gather news in a far different way. Yes, there are still reporters, photographers, and editors, but far fewer than before. “Citizen journalists” send in — for free — iPhone videos or tweets to become part of the news-gathering process. Stories are laid out on computers and uploaded to websites. No paper or ink — or printers — needed. Consumers still get news, but most expect to get it for free (although some outlets have paywalls). The cost of producing news is no longer a tallying of labor and printing costs, making it much more difficult to calculate the contribution of GDP from the media sector.
Additionally, the Internet has created new forms of “free media” that are also hard to quantify. Let’s consider a soap manufacturer, an entertainer, and 100 households. Initially, the soap manufacturer spends $550 to make the soap, spends $250 on selling costs with no value to households, and sells 800 bars of soap for $1 each. Meanwhile, the entertainer sells 100 tickets to her online show for $2 each. The 100 households each spend $8 for soap and $2 for entertainment. As far as GDP goes, the BEA would tally up $1,000 of consumer spending: $800 for soap and $200 for entertainment tickets.
Analyzing free media’s value in GDP
Now, suppose the soap manufacturer pays the entertainer $200 to include a video for soap in her show and cuts other selling costs by $200. The entertainer can now allow the same 100 households to watch her act without charging for tickets. Each of the 100 households still goes out and buys $8 worth of soap. The households are now receiving soap and entertainment (which includes the soap video), but they are paying only the $8 per household for soap. In the current GDP treatment, the entertainment is no longer measured as part of personal consumption, only the soap is. Effectively, $200 has disappeared from consumer spending. However, this appears to be a misrepresentation, because the households are still consuming the same entertainment — they just aren’t forking over cash to buy tickets.
I recently worked with other economists to estimate the added value of free media to GDP. In other words, we wanted to capture that missing $200. We found that marketing-supported information has grown more rapidly than overall GDP, and that this growth — led by gains in online and audiovisual marketing — has had a substantive effect on real output and private business productivity. In fact, we calculated that free content increased nominal GDP growth by 0.025 percentage point per year between 2005 and 2016. This is not a trivial amount, and, using our methodology, measured GDP in 2015 would be $294 billion higher ($18,415 billion versus the reported $18,121 billion). Similarly, because free media has no price attached to it, reported measures of inflation are also being skewed, probably to the downside.
Data mismeasurement is not an academic exercise of interest only to economists. Having wrong information about the economy has real-world consequences for how we spend, how we invest, and how we think about our financial well-being.
For instance, the Bureau of Labor Statistics — the agency in charge of collecting employment and price data — currently reports that inflation is running close to 2 percent. If you get a 2 percent raise, you seem to just be treading water financially. But what if the bureau is underestimating inflation because it cannot capture the correct pricing or quality improvement of the apps you use? If inflation is lower than reported, your 2 percent raise has put you ahead financially.
Inflation errors could also skew your investment decisions. Keep in mind that what’s important in investing is the real interest rate, the nominal rate minus inflation. Earning 3 percent on your portfolio may not sound like much, but if inflation is rising less than 2 percent, you are accumulating real wealth that can compound over time.
The mismeasurements also have policy implications. If current numbers show productivity — output per worker — is rising slowly, then we as a society may think the economy is not generating enough income to finance public programs that help workers displaced by rapid technological progress. But if output per worker is growing faster than we now estimate, then we do have the resources to help by, for example, offering more retraining programs.
Right now we don’t know how bad data mismeasurement is. What can be done? A good place to start would be higher funding for our data-gathering government agencies. Certainly the Bureau of Economic Analysis and the the Bureau of Labor Statistics know the economy has changed, but they don’t have the resources to conduct research or try experiments that would improve data-gathering in this new tech world.
Congress can change this deficiency by expanding the budgets of the government’s economic agencies.
Without increased resources to experiment with new collection methods, these agencies must rely on old methods and sources that simply aren’t capable of capturing our new economy.
Leonard Nakamura is vice president and economist at the Federal Reserve Bank of Philadelphia.