Tim Graf (TG): This is Street Signals, a weekly conversation about markets and macro, brought to you by State Street Markets. I'm your host, Tim Graf, head of macro strategy for Europe. Each week, we talk about the latest insights from our award-winning research, as well as the current thinking from our strategists, traders, business leaders, clients, and other experts from financial markets. If you listen to us and like what you're hearing, please subscribe, leave us a good review, get in touch, it all helps us to improve what we offer.

With that, here's what's on our minds this week.

Simona Mocuta (SM): When I always say like, if there is something worth investing in, it's education. As an economist, I would say, if there's something worth investing in, it's good data. This, to me, is the raw material. It's the foundational element, and it deserves attention, it deserves investment, it deserves recognition.

TG: Until recently, there was a narrative of the United States economy, and especially the US labor market, that went something like this. Things were undeniably slowing down, but fiscal policy was set to provide a tailwind. Tariffs were going to add some amount of inflation into the mix. Jobs were being added to the US economy at an expansionary pace. Despite a ton of political pressure, the Fed had pretty good reason to be cautious in easing policy any further.

But large downward benchmark revisions to labor market data over the 12 months to April 2025 have completely changed that narrative. We now know that the US labor market was growing way slower than previously believed at a rate that's more like stall speed. Forget about acceleration. This caught the attention of the president, who understandably enough didn't exactly love the fact that the jobs market now looks like it slowed a lot more than he or any of us previously believed. It caught the attention of the Fed, who pivoted to a more dovish stance in resuming rate cuts last week.

How does this happen? How does such a big shift in narrative come from a process that we used to almost take for granted? Well, we're going to talk a lot about that this week, as these events have brought focus onto issues in economic data collection and revision that have been brewing for a long time. Issues which we talk about with my guest, Simona Mocuta, chief economist at State Street Investment Management.

How are you?

SM: Okay, good, I'm good, thank you. I'm glad you thought this was an interesting topic to cover. It's kind of funny because it's been so front and center for me. And in fact, I wanted to run this by you that there are a couple of anecdotes that I thought might bring this to life a little bit.

And one is around that Retail Sales data. It was then, we got the initial data, 17.5 percent month on month. And at that point, I went to Lori [Heinel, global chief investment officer at State Street] and I said, as the revisions were coming in and dramatically lower, I went to Lori and I said, you need to give me a little more money so I can access the revision series. So I can show you in a chart what actually has happened. Because otherwise, some of these are so dramatic that people think, oh, you must be dreaming, you know.

TG: Yeah. I wanted to make this really kind of straightforward and for a lay person, because I think it's a really important topic. That's kind of why I was curious to have you on or keen to have you on this week is your new blog post, “Why the Global Macro Data Crisis is Everyone's Problem.”

And I think my intention with this is to make it as much educational for people. I think there are people who listen to this podcast who aren't necessarily in the weeds as much as we are and invested in things like data quality. And understanding why it's important, I think, is really an important way to start.

So can you kick it off and just talk about why public economic data especially is so important to our lives and not just as economists and strategists, but to everyone?

SM: Yeah, you know, it's sometimes it's good to get out of your own bubble, because for me, this has been front and center now for some years. I think really after COVID, it was a clear decline in the quality of data.

But why should everybody else care? Well, frankly, I think it comes down to some very basic reality, that you cannot make good decisions without good data. And this goes in your personal life, in business, in every sphere of life, right? I mean, from simple things such as, you know, how do I, what's the weather? How do I get dressed today, too? Well, there is a medical diagnosis. Clearly, I want the right information and the right data, right? So it's the same for the economy. You know, how can you run a business if you don't have the correct information about your sales or how you're doing in the marketplace? It becomes a little fuzzier, I think, for the average person when you talk about the economy as a whole, but it's the same reality.

Policymakers have to have a sense of where are we in the business cycle. The Fed, you know, cut interest rates recently. How do you make those decisions? You have to have data to guide you, right? So the better the data, the better ultimately the decision is. And that's irrespective of political views, irrespective of professions. It is a fact.

TG: Thinking about it practically in terms of current economics, and you mentioned the Fed, and the Fed responded last week to deteriorating labor market data. And of course, that's a big topic when it comes to data quality, especially given some of the things we'll talk about with respect to the Bureau of Labor Statistics, who produce the data that in many ways might have driven, or at least partially driven, that decision.

Can you talk about some of those issues? What are the specifics in terms of how is quality of data deteriorating, both in the collection and maybe even the presentation? Where are you seeing this evidence deteriorating quality?

SM: There are two things that are noticeable. Perhaps the most important and visible one is the size of revisions. You're dealing with big revisions to data. In a way, you're told, well, we thought it's like this, but actually it's a little bit different. Revisions are always a fact of life when you collect lots of data. It's never going to be perfect as you collect more information, you're going to make revisions.

But over the last few years, it has been obvious that the magnitude of these changes is in a sense unprecedented. And it's not just BLS data. I want to be clear about that. It's not just BLS, it's not just the US, it's visible across the world.

The other element that comes into play here is that you notice inconsistencies in data sets that really should speak to the same underlying reality. So as an economist, I really like to utilize as many data points as possible. Why? Because I cross check the information. So in the labor market, you have many series. You have the initial claims data, you have ADP, you have payrolls, you have other private sector data. Do they tell a consistent story? For instance, I'm looking at things such as hours worked, the quits rate, all of these tell me a story about the labor market. Not the whole story, but very, very important bits and pieces. And to really make up the whole narrative, you want to make sure that these things are telling a consistent story.

So when I hear on one hand, funding cuts, on the other hand, big gains in government education, in employment, that seems inconsistent. And that was part of the story behind the recent payrolls revision. So you see these things, not just in labor market data, you see them elsewhere. So these are red flags. If A is true, then what would I expect B to do? And when B does something different, then at least you have to ask yourself the question why.

TG: Really ironic to me because we live in supposedly the age of big data and we've been telling ourselves this for the last, I don't know, 20 years probably, having access to lots of data sets, probably at a lower marginal cost of production, although maybe that's debatable.

Why do you think data collection has become so fraught with some of these problems?

SM: We do live in a world of big data and lots of data, but the reality is that our, at least in the space of macroeconomic data, a lot of the collection mechanisms are really not very different from 20 years ago. So at the core, I think part of, why are we seeing all these problems now? Well, there's some very objective reasons.

Number one is COVID. The COVID shock was such an unusual shock to the economy, the extraordinary magnitude. So how do you do seasonal adjustment in its aftermath? It's a very valid question and you have to have a lot of compassion for the people who have to tell you, this is what the typical April looks like. Well, you know, because you had April 2020, that completely changed the world. So subsequent months, you know, the April of 21, 22 will look different. How do you calibrate that? You know, that's a very, you know, it's an objective problem. And then you have the problem of very low response rates and dramatically low response rates. So that has, you know, and in many cases is really the response rate depends on a person. At the other end, actually filling out a survey that's time consuming.

So in many ways, technology makes things easy. But I think in the data collection process, we are probably not deploying that as effectively as we should. And, you know, maybe I'll give you an anecdote, but about a year, year and a half ago, I got a call from the local Bureau of Labor Statistics Office here in Boston. They were in the process of updating their quarterly census of employment and wages. And, you know, they reached out to me to sort of see if I can help ensure that State Street responds on behalf, you know, being a big employer in this area, you want to make sure that these big players respond to your surveys. Because if they don't respond, you know, how do you fill in the gaps? So, even in the age of big data, somebody has to do it. It's a challenge.

TG: That's the really one of the striking charts you have in the piece that you wrote, which is the declining response rate to these surveys and particularly BLS surveys. And it, as you mentioned before, it kind of cuts across political leaders. It's not a Republican thing. It's not a Democrat thing.

There may be some budget discussions we can have that are, you know, with respect to the BLS, that maybe are a little bit more political in nature. And the changes to the leadership we'll talk about are a little bit like that. But this problem of collection is really not new. It goes back even before COVID. It was a big problem that was starting to decrease before COVID.

Has the methodology itself just not caught up to the way people work? Is that really what’s behind this?

SM: I think there's probably quite a bit of that because where we are has changed, right? Where the consumers are has changed like 20 years ago, you're not purchasing stuff on your phone, now you do. Well, are we deploying our surveys where the respondents are? Are we making it easy for them to respond? Probably, we are not doing as good a job as we could.

And of course, the first step in improving the situation is to recognize that you have a problem. That there is a genuine challenge. And you know, I've worked a lot, many, many years in the emerging market space. And then I started covering developed markets, and especially the US. And you feel grateful that you have a data abundance. It's amazing to go from, you know, limited information on an economy to a situation like in the US, where you have so much data.

But at the same time, you cannot, you know, rest on your laurels and say everything's perfect because clearly, there's been erosion in the quality of our data. So let's admit that that's the case. Let's not, you know, jump to assigning blame, but I focus on the problem. And I would say, you know, on a personal level, I always say, like, if there is something worth investing in, it's education. As an economist, I would say, if there's something worth investing in, it's good data. This, to me, is the raw material of, you know, everything that we do. So how can I provide a good forecast if my underlying data is incorrect? If you don't have that good forecast, what kind of advice can you give to a business or an investor as to what's the best course of action for you here, right? So it's the foundational element, and it deserves attention. It deserves investment. It deserves recognition.

TG: Things that struck me when I was thinking about questions to talk to you about was, and you mentioned the pandemic, actually, and I'm really glad you did, because that was this environment where we weren't in offices. People were, there were anecdotes of people doing multiple jobs without often telling other employers they were doing work for other companies. But the notion of the gig economy really came to the fore, I think, and people became a little bit less tied, first of all, to offices, but I think more, less tied to jobs themselves or companies themselves.

Do you think that is creating some of the volatility, particularly in revisions? I know there’s a response issue, but the revisions being so volatile, is that an element of it in countries like the US or the UK where that is a very prevalent means of employment?

SM: Yeah, I don't know if it's necessarily feeding straight into the data revision question, but I do think about this a lot because I do believe there is a growing segment of economic activity that we don't fully capture. And I'll give you a little recent story. So my daughter recently discovered the amazing world of Facebook Marketplace. And sometimes she keeps telling me these stories about all these things that happen, all the transactions that happen. And I listen as an economist and wonder how much of that do we capture in our statistics? Because these are person to person transactions. I believe more of our economy happens this way now. And I think we are, our systems of data collection are not fully geared to capture that appropriately. You know, how many deliveries are there daily, right?

So all of these things are genuinely transforming the way we do things, and our statistics, I don't think, have caught up with that reality.

TG: Well, the pandemic also brought in, and we ourselves did it, we started a whole publication around alternative data sources, so private data sources. You know, we'll talk about PriceStats, is the one everybody kind of knows us for. Indeed is another one we're using a lot. I'm sure you're using a lot. But we were also looking at things like restaurant bookings, all of these.

SM: Yes, OpenTable.

TG: Yes, that was like the big data release when you didn't have good data.

SM: Exactly. I remember those early days, there was you're like, you know, scrambling for every piece of private source data that you could get your hands on, trying to piece it together to understand what's happening. It was OpenTable. It was the TSA traffic. Do you remember how we were tracking that? And then there was Homebase for return to work. You really were trying to get your hands on any piece of information.

And that's where you see both the value and the limitations of private data. On one hand, you can really get a lot of granularity, specific industries, very timely, high-frequency data, right? OpenTable was a daily, weekly data set. You know, much better than you get in official statistics. The same with Homebase.

And then one day, you come to the office and you try to update your chart, and well, the home base data is no longer available for free. So that's always a risk, right? Because these companies are not obligated, they're not in the business of providing the data as a public service. It's a collateral benefit that we may get, perhaps only temporary. So, this is where you need the consistency and reliability of official statistics.

TG: But how much do you think public officials rely on these now? Because that was the great eye-opener as well. It wasn't just us on the buy and the sell side looking for alternative data. It was, and this is in the minutes of central bank meetings, people citing PriceStats as what they look at. Not even in the pandemic, before the pandemic, but of course during it, and then after, we've had lots of traction and have gotten, you know, we were one of those data providers who was, who was providing food price data during the pandemic to central banks to help them monitor shortages and to monitor supply chain issues.

How dependent do you think they are now on private data sources despite some of these limitations and risks that you mentioned?

SM: Officials would not do their job properly if they did not try to leverage all the information that's available to them, right? So I take it as a given that you, you at the very least look at and then decide for yourself, how do you prioritize? How do you rank the importance of different data sources for your own self?

But you have to, number one, be aware of what they are telling you, right? So you have to scan the whole horizon as to, you know, what information is available there and utilize that. I think it's the only way to go.

TG: Well, let's start to think about solutions to this. And this is, you make some suggestions at the end of your blog post, and I want you to go through those.

But first of all, actually, are there countries or agencies that so far have avoided some of the pitfalls you've mentioned? Are there any, is there a gold standard out there that you would point to and direct policy makers to or public officials towards? Like, this is what you should be doing.

SM: I wish there was a simple example to say, okay, that's the solution, right? I think by and large, I would still say and argue that the US is the gold standard in terms of macro data collection. If only for reasons of abundance, we do collect a tremendous amount of data, so you have a lot of insights. This is probably going to be a process, right? We should think about this not as a one time, like this is the one thing that you're going to do and that's it. It's a process and I would say, I would start by trying to increase the response rate. If there is one thing that gets you better quality, fewer smaller revisions is actually getting the initial data more reliable, the way you do that, you make sure that more of your sample actually responds.

So how do you do that? I think there are lessons we can learn from other countries. In some cases, you can either mandate the responses. In some cases, perhaps you're willing to offer some compensation for either the individuals or firms to respond. It is a time-consuming process, so you have to recognize that it's not costless for a company to spend two hours to respond to a survey. Then you ask yourself, well, actually, how do I actually make it such that they don't have to spend two hours, but perhaps they can do this in like 20 minutes. And then on your own end, as you are a statistical organization, you have to look at your internal processes.

It's not just getting people to send you the data, but then internally, do you get these data in systems that easily talk to each other, so you gain efficiencies on the processing side, or do you then end up needing to spend a whole week figuring out, putting the pieces together internally? I think there is a lot that can be done on that side, through digitization, through investment and upgrade of that data infrastructure on the government side. But externally, what can you do to get your respondents to actually respond?

TG: I mean, on that note, and not to make this in any way political, but one of the big issues with revisions in the labor market data a couple of months ago led to a change at the top of the Bureau of Labor Statistics. On the surface, it looked a little bit political, but there were some of these legitimate concerns that you've raised about quality of data, size of revisions, volatility of revisions.

How confident are you that a change in leadership is a good, the right first step? Or is it mostly cosmetic?

SM: Well, it's hard to tell where we end up with this, but I would say this has now gained prominence in the public debate. And to my point that in order to fix a problem, you first have to recognize that you have a problem. I think now we are recognizing that there is something here that needs fixing.

Beyond this, I would say, can we please put the political angle aside, because this year we are talking about the public good. Correct, reliable, unbiased data is a public good, the same way that clean water is a public good. Maybe in a different space, but equally fundamental.

So then what I'd like to see now is a genuine, like the UK, is undergoing a similar situation, where you have hearing and a whole official public review of the processes, and independent commission suggesting changes. I would love to see something like that in the US, and I of course would love to see it on a bipartisan basis. And Congress has a role here because frankly, at some point these things require money, and Congress should fund these efforts.

TG: I am only glad to see that this is getting recognition as an issue that requires attention. One final question, and bringing the Fed back into it, I'm pretty confident you've just updated your quarterly forecasts, and we did have a Fed meeting last week, and the Fed cut rates as was mostly expected, but they also had a somewhat dovish take on the future path of rates relative to their previous set of projections. And as noted before, a lot of it was based on weaker labor market data.

So, taking into account some of these issues, do you think they actually did the right thing in sounding a bit more dovish going forward? And what is your outlook based upon the labor market data that you’re trying to make sense of despite all these issues?

SM: Tim, what can I say? I'm glad you asked the question because I most certainly think they did the right thing. In fact, we've been consistent on our end since the beginning of the year, calling for three rate cuts from the Fed in 2025. We would have voted for a cut in July, precisely because I was seeing these inconsistencies in the labor market data that told me, well, it's not just a strength in payroll, there is more under the surface that's a little bit softer.

So it's good for the Fed to take a little preemptive action here and make sure that the soft patch stays nearly a soft patch and not something more. So happy for them to see that they cut. We are on board, as I said, with two more cuts this year, and the median dot plot is now indicating that, even though it's a huge dispersion of use there. So the way I think about it, it's quite constructive on next year's growth in the US, but there is a soft patch that the Fed, they are not the third party independent observer here. They have a role to play, and now they are stepping up and playing that role to support the economy to ensure that things don't deteriorate.

TG: Very good. Well, I don't want this to deteriorate any further, this discussion. So it's probably a good place to pause it. Not to end it, to pause it, because we always love having you on the podcast. But the piece is called Why the Global Macro Data Crisis is Everyone's Problem, written by Simona Mocuta. It's on the website of State Street Investment Management. Please do check it out as well as check out everything Simona does. It's brilliant, as are her calls. So Simona, thank you so very much for joining this week.

SM: Thank you, Tim.

TG: Thanks for listening to Street Signals. Clients can find this podcast and all of our research at our web portal, Insights. There, you'll be able to find all of our latest thinking on markets, where we leverage our deep experience in research on investor behavior, inflation, media sentiment, and risk, all of which goes into building an award-winning strategy product. And again, if you like what you've heard, please subscribe wherever you get your podcasts and leave us a review. We'll see you next time.