Is the credibility of US government data at risk? Why it matters to everyone.

The credibility of U.S. federal government data may seem an arcane topic, but it is critical for a well-functioning economy. To explore the importance of reliable economic data and potential challenges to that reliability, and why it matters to businesses and the public, Ben Harris, vice president and director of Economic Studies at Brookings, joins The Current.

HARRIS:  We have this sweeping system that regularly produces statistics that relate to critical parts of the economy. And for decades, it’s really been considered credible, unbiased, and reliable. And I think that’s at at threat today.

[music]

DEWS: Hi, I’m Fred Dews, and you’re listening to the Current, part of the Brookings Podcast Network. Today I’m diving into a topic that may seem arcane — the credibility and usage of U.S. federal government economic data — but one that is critical for a well-functioning economy.

To help us explore the importance of reliable economic data and potential challenges to that reliability, and why it matters to all of us, I’m joined in the studio by Ben Harris, the vice president and director of Economic Studies at Brookings.

Ben, welcome back to The Current.

HARRIS: Thank you for having me on.

DEWS: So Ben, when we talk about federal economic data, what exactly are we referring to and what are some of the key statistics that these agencies provide?

HARRIS: So Fred, when we talk about federal economic data, we’re talking about the information collected by the federal statistical system that relates broadly to the economy. So some of the best known statistics are the unemployment rate, GDP growth, various measures of inflation. But the information collected and distributed is really much broader than that.

So let’s talk briefly about the federal statistical system. This is the sweeping, decentralized collection of over 100 different entities, but most of the activity is centered through 13 different principal statistical agencies. So even though the work is decentralized, it is coordinated. So it’s coordinated by the Office of Management and Budget, which sits in the White House. And it’s also guided by various laws and statutes that guide the work of these principle statistical agencies.

So I think one of the best known laws that guides the collection of this data is the Constitution. And so Article One and the Constitution dictates that we have a census that goes out and collects information about the number of Americans who are alive and where they’re located. But there are also other laws guiding activities of the system, like the Paperwork Reduction Act, which dictates some of the structure of the federal statistical system, including designating OMB as a coordinated agency.

So within these 13 principle statistical agencies, it’s tough to discern what exactly is economic data and what’s not. So we know, for example, the unemployment rate, that’s clearly economic data. But there’s other things too, like data on education and data on crime, which some people might consider economic data and some might not. But the central point is, is that we have this sweeping system that regularly produces statistics that relate to critical parts of the economy. And for decades, it’s really been considered credible, unbiased, and reliable. And I think that’s at at threat today.

DEWS: And I want to get into that question of threat in a moment here. But first, I’d like to ask you about how different sectors of our economy and our nation really use these data. Let’s start first with businesses. How do businesses use these data?

HARRIS: So the broad answer is that businesses use these data to understand the business cycle. So to understand where we are in terms of the general economy; the direction of the economy; whether or not unemployment is an issue; the extent to which we’re seeing expansions in consumer spending and possibly trends in prices as well. So inflation.

But it’s really much more nuanced and specific than that for many businesses. So, for example, for a major retailer like Walmart, they might want to make decisions about whether or not to open a particular store or whether to close a store. So they’ll look at spending trends and they’ll sometimes look at spending trends in geographic areas in order to determine where they should direct their activities.

You’ve got manufacturers like General Motors who might want to know more about labor market trends, so they can project future costs of employment.

You might have home builders, like Toll Brothers, might want to look at housing data to know about expanding or contracting their construction activity and where they should focus their efforts.

You could even have agricultural companies like Cargill, ask how they should plan for future production. And even looking at weather data, which maybe you don’t consider to be economic data, but is very important for an agricultural company. And also how to hedge their bets as far as future price movements.

You could have shipping companies like UPS, ask questions about planning for future shifts in e-commerce and where they should direct their operations.

You have airlines like United or Delta asking questions about how they should plan future routes, looking at different projections of economic activity, where to optimize a hub, should we be in one city or another?

You even have insurers like State Farm or Allstate asking questions about providing insurance policies related to weather, like we just mentioned, or maybe looking at mortality data for life insurance policies.

And, you know, I think that some of the most frequent users of federal data are those in the financial sector. So banks and hedge funds, which really pay very, very close attention to these releases. I used to work for a hedge fund briefly. These data releases are central to their their business, and their ability to project some of these data releases will dictate their future profits.

So business use of these data is just sweeping. It’s really hard to imagine a robust U.S. economy without these data.

DEWS: Right. So it’s a huge swath of economic activity, financial activity, across every sector you can imagine in the economy. But can you explain how these economic data impact day-to-day, kitchen table concerns of everyday Americans?

HARRIS: Yeah, so we just talked about how it impacts business. But in terms of the pocketbooks of households, I think the most obvious example is through the Federal Reserve. So the Federal Reserve has this dual mandate, it needs to control prices. So basically it needs to control inflation, but it also wants to maintain maximum employment. And so the extent to which jobs are plentiful, and the extent to which inflation is under control really is is a reflection of how well the Fed is doing its job.

And the Fed is very data dependent. The Fed use these federal data as an input to making their decisions. And if this data is not credible, if it’s not produced on time, the Fed can’t make decisions as well as it would like. So that’s kinda the first conduit for this data.

But then in even more direct ways, many of the finances for U.S. households depend directly on these data. Let me give you a few examples. So the tax code: the tax code has all these parameters which are indexed to inflation, so usually to CPI, the consumer price index. The extent to which, for example, the extent to which the tax brackets increase each year depends on CPI. Also things like contribution limits to 401Ks, how much you can put away each year, are indexed to inflation, or indexed to CPI.

For people who are Social Security beneficiaries, that’s indexed to one measure of inflation called the CPIW, which is the CPI for wage earners in urban areas. And so for Social Security beneficiaries, how much you get each month directly depend on this federally produced measure.

Other things too, like. People who invest in Treasury Inflation Protected Securities, or TIPS, that amount is directly determined by CPI. So you have all these different measures of of inflation, or how the Federal Reserve uses data directly flow into household finances.

DEWS: So is it safe to say, it’s hard to imagine an aspect of American life, American economic life, American personal life that isn’t impacted by federal statistics?

HARRIS: Yeah, I mean, we’re talking about things like from Social Security benefits to the labor market. It really is sweeping.

DEWS: Ben, earlier you mentioned the Census as one prime example, one well-known example of data collection. I think another very well-known one is the Bureau of Labor Statistics and their monthly job report, you mentioned unemployment earlier. We also call the Bureau of Labor Statistics the BLS for short. Those kinds of agencies have long guarded their independence, even though they’re lodged in federal cabinet agencies. They’ve addressed over the years issues like budget cuts, hiring freezes, staffing cuts. And we’ve seen that recently the BLS has been more in the news, over the summer especially and recently. How did those kinds of longstanding institutional challenges compare to some of the more overt actions that we’ve seen in the news lately such as the recent firing of the BLS commissioner?

HARRIS: Yeah, so I think the right way to think about it is two different sets of challenges. So the first one are the ones which are really not self-imposed by the agency. So sometimes we’re talking about staffing challenges, they just don’t have enough people to do the work. We have these broader changes in how Americans view surveys. And a lot of the BLS information is derived from direct surveys of Americans or direct surveys of businesses, and we’ve seen response rates go down pretty sharply over time. So that’s unrelated to these decisions made by the BLS, but still pose real challenges, not necessarily for credibility, but more for their ability to produce these statistics in a way which is accurate, and which is timely.

The second challenge has to do with political challenges to the agencies. And we just saw as you, as you mentioned, the firing of a BLS commissioner after the president was unhappy with jobs reports. And this in my view is the most egregious attack on the federal statistical system that we’ve seen.

So it is true that we have seen marginal infringements on credibility. And so one example that was given in a recent event we hosted at Brookings was during the Biden administration, and I served in the Biden administration, was around the time after which the data was released that federal officials can discuss the data. And so the example was given that the Biden folks wanted to go from one hour to 30 minutes as the kind of the blackout window for discussing the data. But that is so different in magnitude than firing the BLS commissioner because you’re unhappy with the jobs report.

So while it is true, we have seen these, I guess, you know, incremental threats, if you want to use that word to credibility, it’s nothing like we’ve seen in the past couple months.

DEWS: I’m glad you mentioned that event. I’ll tell listeners that it was held on September 18th. You can visit our website, Brookings dot edu, where you can see full video and download a transcript of that great discussion. It included two former BLS commissioners and some representatives from the business community. So check that out on our website.

Ben, I want to ask you a question that I think a lot of listeners might have. Why does the federal government need to produce so much data at all? Why can’t the private sector produce the kind of data that it needs to make good decisions? I mean, JP Morgan, it’s a huge company, it has its own data sets. ADP, the payroll company, has its own payroll data. Why can’t the private sector produce this data?

HARRIS: Yeah. So one analogy I could give is why do we have a national military? And we can all go out and hire our own personal security forces. But that’s not very efficient and it’s really not practical for a lot of households.

Just the same way as you mentioned JP Morgan, but most businesses in the United States are small businesses and they really don’t have the capacity to go out and have an in-house statistical team. Even JP Morgan, I think, doesn’t really have the capacity to go out and have its in-house statistical team. The BLS alone has 2,000 employees. I mean, can you imagine if JP Morgan had to reproduce all the statistics, not just BLS, but GDP and inflation? I mean, they would have to hire tens of thousands of employees. So it’s really not practical at the company level to go ahead and have these massive in-house statistical teams.

Also, the federal government has a different fundamental motive than businesses. U.S. businesses operate in order to create profit. And that’s a good thing and, you know, part of capitalism. But then we also have these public agencies that have a very different motive, which is to provide services for these businesses. They don’t care about profits, they care about producing reliable, and credible information. So, they just come at this from a very different angle.

But, you know, you mentioned ADP and I think ADP is also kind of the best example of a private sector entity, which produces a statistic which might seem on its face to be a substitute for a publicly provided statistic. But ADP is not nationally representative. And if your goal is to look at the economy on the whole and make conclusions about where we’re headed, ADP is really nowhere near as good as what BLS produces because it’s not designed to be.

Even the biggest banks in the United States, really don’t have the capacity and their motivation’s very different than the public sector entities.

DEWS: Ben, I want to wrap up with a with the big question here. Are you worried that federal statistical agencies like the BLS could lose their independence, could become politicized? And if so, what’s at stake for the economy, for businesses, for regular Americans, if that happens?

HARRIS: So, one of the best ways to answer that is to look at international examples, and there are two that I think economists like to raise, given all of the concern around credibility in the wake of the BLS commissioner’s firing. So we’ve seen, for example, in Turkey when there was real pressure from the president’s office to lower the reported inflation rate. And what we saw was that households that depended on measures of inflation because it would set their wages or set their pension levels. And, you know, inflation is what it is. If you say it’s half of what it is and people get less in wages, they get less than pensions, all that means is that their standard of living goes down because they’re not getting the raises that they deserve in order to keep up with inflation.

So what you saw was a real suffering at the household level for those not having enough information. And that could happen in the United States, for example, if you see political pressure to cut the inflation rate in half, that means for Social Security beneficiaries, they’re seeing their standard of living go down.

The worst example happened in Argentina when you saw the administration around 2007 through 2015 put real pressure on their statistical agency in ways that start to look like what’s happened with BLS. Now, the magnitudes are very different. But at that time, the president of Argentina fired the head of the statistical agency, tried to point to private sector statistics, which were already biased, fired lots of people at the agency, and really put pressure on the agency to report less inflation. And as a result, you had investors flying blind, which meant that they charged higher premiums for investing in Argentina. And you also saw a real loss in confidence by the people of Argentina. So when you would get an inflation report, they just didn’t believe it. And what happened was that inflation expectations went up over time.

And that’s a real worry for an economy because that means that people will stop behaving in kind of normal ways and start taking steps to avoid those future price hikes. So credibility is really central to to avoiding that.

And so, I don’t want to overstate what’s happened over the past couple months. We did have a firing of the BLS commissioner, but we haven’t had other firings. But it’s a path that we don’t want to go down. It’s an experiment we don’t want to try.

DEWS: Well, it’s such an important topic, and Ben, I want to thank you for sharing your time and expertise to help us unpack it today.

HARRIS: Of course. Thanks, Fred.

DEWS: To learn more about the importance of credible economic data, you can visit our website, Brookings dot edu. And a reminder, you can find the full video and transcript from a public event held at Brookings on September 18th.

More information:

• Listen to The Current on Apple, Spotify, YouTube, or wherever you like to get podcasts.
• Learn about other Brookings podcasts from the Brookings Podcast Network.
• Sign up for the podcasts newsletter for occasional updates on featured episodes and new shows.
• Send feedback email to [email protected].