The latest episode of On the Evidence features Mark Zandi, the chief economist at Moody’s Analytics. On the episode, Zandi speaks with Mathematica’s President and Chief Executive Officer, Paul Decker, about immigration reform, artificial intelligence (AI), labor shortages, remote work, the merits of pursuing a nonacademic career in economic research, and how Zandi seeks to influence politically charged policy debates with data and credibility. Zandi is the author of two books related to the Great Recession and hosts the Inside Economics podcast.
During a discussion about the policy implications of an aging workforce in the United States, particularly on current labor shortages and the ability of younger workers to keep the Social Security Insurance and Medicare programs solvent, Zandi highlights two “wild cards”: immigration and AI.
“Because of the aging of the population, the retiring of the Baby Boom generation, my generation—[the] labor force is coming to a standstill, and it’s going to decline. So we’ve got this very significant weight on the working-age population,” Zandi says.
“Immigration could be a way to get more growth and address some of these longer-term fiscal issues,” he says, adding that he hopes Congress will eventually pass a new immigration reform law to help manage the flow of immigrants into the United States.
Zandi says AI and other technological innovations could also help with the current tight labor market by improving labor productivity, but he also acknowledges significant uncertainty about AI’s impact, including the risk the technology poses to the economy if it eliminates jobs without creating new ones. Accounting for the potential impact of AI on the economy, his own team’s real GDP growth forecast for the next 10 years shows an increase of about 0.15 percentage points (“So it’s consequential, but I don’t think it’s game changing,” he explains.)
“If history is any guide, it’s more likely than not that AI is going to be more a plus than a negative,” he says. “The technology is going to diffuse through the economy over an extended period of time, allowing labor productivity growth to improve enough to help us out but not come on so fast that it’s going to wipe out all these jobs.”
During the conversation about labor shortages, Decker notes another factor that is likely bolstering the supply of workers: the rise of remote work as an option for employees. Many of Mathematica’s staff were already working remotely before the COVID-19 pandemic, but even more have embraced remote work or a hybrid-work setup since 2020. To date, the company has not mandated employees return to the office in person, but has kept offices open across the country and resumed in-person activities such as team retreats and public convenings.
“We take a very flexible approach at Mathematica and partly for some of the reasons that relate to the supply of talent,” Decker says. “Individual preference for flexibility is strong enough that I thought it was worth at least seeing if we can achieve both benefits—benefits of flexibility but also finding ways of maintaining corporate culture, creating connection even if we’re not necessarily doing it daily on a face-to-face basis.”
For his part, Zandi says his team at Moody’s is fully remote and, “it feels like productivity hasn’t accelerated, it hasn’t decelerated, and people are happier, so why not?”
In the interview, Zandi also explains his rationale for using op-eds, media appearances, and visits to Capitol Hill to be a visible and vocal champion of data-informed policy.
“If you're not out there in the public domain and expressing views, you’re not credible. You don’t have a brand,” he tells On the Evidence. “You need that brand and credibility to be able to establish the trust in what you’re saying and maintain the perception of being unbiased. You want to be unbiased, but it’s also really key there’s a perception of being unbiased.”
One of the newer ways Zandi maintains a public profile is his Inside Economics podcast, which he calls “a really effective way of reaching people…at the same time, [I’m] really having a lot of fun doing it.” Zandi says launching the podcast in 2021 was good for his staff, who join him as co-hosts of the show, because “we’re talking about issues that they’re passionate about and you can see how talented they are, and they can express their views in a kind of more relaxed way.”
On the role that evidence plays in influencing policy debates, Zandi notes with concern the long-term decline in response rates on surveys and the damage that lower participation in surveys could cause to the quality of information policymakers use to make decisions.
“We can't make good, informed decisions unless we have good, underlying economic information and data,” he says. “At the end of the day, the government needs to be able to provide consistent, accurate, sets of data so we can really understand what's going on.” He hopes federal policymakers will provide more funding to ensure data sources are more resilient, comprehensive, and timely.
Watch the full episode.
View transcript
[MARK ZANDI]
If people aren’t quoting you, if you’re not out there in the public domain and expressing views, you’re not credible. You don’t have a brand, and you need that brand and credibility to be able to establish the trust in what you’re saying and maintain that unbiased – the perception of being unbiased. You want to be unbiased, but it’s also really key that there’s the perception of being unbiased; and that goes back to your credibility. And all those other things that we were talking about that I do is an effort to establish and maintain that credibility.
[J.B. WOGAN]
I’m J.B. Wogan from Mathematica and welcome back to On the Evidence.
We have a very special guest for you today, Mark Zandi, the Chief Economist at Moody’s Analytics. Mathematica’s President and CEO, Paul Decker, is stepping in as the guest host for this episode. Mark and Paul discuss immigration reform, artificial intelligence, the labor shortages, the merits of pursuing a nonacademic career in economic research, and influencing politically-charged policy debates with data and credibility.
If you like this episode, you may want to check out the podcast that Mark cohosts, which is called, “Moody’s Talks -- Inside Economics.” I hope you enjoy the conversation.
[PAUL DECKER]
Mark, thank you for joining me for the On the Evidence podcast. I’ve watched and listened to you for many years, and I’m really interested in your professional journey. You and I share a common background in that we’re similar in age, and we earned economics Ph.D.’s, yet we both chose nonacademic careers. What do you think about your career path and how it reflects your interest in economics and data, and why did you chose a nonacademic career?
[MARK ZANDI]
Oh boy, well, it’s good to be with you, Paul. Thanks for the opportunity. That’s a big question. Where did you get your Ph.D.?
[PAUL DECKER]
Johns Hopkins.
[MARK ZANDI]
Oh, great university, my deputy chief economist – I’m the chief economist at Moody’s, and my deputy chief economist got his Ph.D. from John’s Hopkins.
[PAUL DECKER]
Yeah.
[MARK ZANDI]
So great, great university.
Well, the question that’s easy to answer is why didn’t I become an academic; and that’s because I lived in a family of academics. My dad was professor of systems engineering way back in the day – civil engineering – at the University of Pennsylvania. That’s where I got my Ph.D. I did my undergrad work there at the Wharton School and then my Ph.D.
There’s a lot to like about the academic lifestyle, but there was a lot I didn’t like. I’m not a very patient person, and I think you need to be a very patient person to be a good teacher or professor or a mentor.
Then, different circumstance – when I was getting my Ph.D., I was working at a firm called Chase Econometrics, which was a firm that was established by a former Penn professor who had been a colleague, Larry Chimerine. Larry Chimerine was a Nobel laureate who kind of was the father of kind of econometric work and model that we do today. So I started working to earn money. I also used their mainframe and data to do my Ph.D. thesis. I just got engrossed in the business of the economic forecasting and the analysis and data work that that firm was doing and then was off and running.
Soon thereafter I got my Ph.D., I started a firm in 1990 that ultimately had competed with Chase Econometrics and Wharton Econometrics and DRI – they’re the firms in business. Those firms are no longer with us. They’ve all been acquired and assumed by others and have been broken apart. But our firm has grown. We were acquired by Moody’s around 2005. So 15 years, we were a privately-held company, small business. By the time we sold, we had 100 employees; and now we’re part of the large multinational organization, the Moody’s organization.
We’re all about data. We’re very data-intensive. We collect data from all over the world – government sources, trade groups, the private sources, consortiums – data consortiums. We clean the data. We provide that data to others for their use, and of course we use it for our own modeling and work that we do – forecasting, scenario analysis, policy analysis. So we’re very intensive users of economic financial demographic data information. So that’s kind of sort of the career path and why such a focus on economic information and data.
[PAUL DECKER]
When you chose to go down the nonacademic path, did you run into any interference from your family or from your advisors at U of Penn?
[MARK ZANDI]
No, no, I mean because there was Chimerine; they’re was all the other folks, academics that were very empirically oriented. They had their own firm, so I didn’t really – my dad, he – he was a very good academic, but he understood the kind of the issues and limitations. Actually, he was an immigrant from Iran; and he left the country because of the political environment. He was the head of the student body at the University of Tehran. So he was very politically active and had to leave because of the overthrow of Mossadegh. So if he hadn’t made that trip here to the United States, he probably would have ended up being somewhere in the poli – of course Iran went down a very different path, but he might still be part of what’s going on there and not an academic.
So, no, I didn’t really receive any resistance. The resistance was economics. My dad – I remember because he was an engineer – civil engineer, systems engineer – he goes, “What do you do with an economics degree?” That was the question (laughing). Of course, I had no answer whatsoever for that. I go, “I don't know, but I really enjoy it; so this is what we’re going to do.” And I think that’s an important lesson. I mean, it’s hard to know exactly how these paths play out and what’s going to work and what’s not going to work. You’ve just got to follow the thing that you enjoy and you’re reasonably good at.
[PAUL DECKER]
Yeah, my father was a theater professor.
[MARK ZANDI]
Oh!
[PAUL DECKER]
And similar question about, “What do you do with an economics Ph.D.?” (laughter).
[MARK ZANDI]
A theater prof – that’s interesting. Where did he teach?
[PAUL DECKER]
At MacMurray College in Illinois – small private college in the Midwest.
[MARK ZANDI]
Well, that’s so cool. So were you – you must have been on stage then.
[PAUL DECKER]
No.
[MARK ZANDI]
Oh, you never did that.
[PAUL DECKER]
No, I avoided that.
[Laughter]
[MARK ZANDI]
That’s too bad, I bet that would have been – you would have been good, good actor.
[PAUL DECKER]
I appreciate knowing more about your background, and I’m interested in the ways in which you communicate your work.
At Mathematica over the years, we shifted to having more of our focus on how we disseminate our findings/insights and how we do that as effectively as possible with the intent that don’t want to just create the evidence. We want to make sure the evidence is as useful and as informative as possible to those who are going to use it – policymakers and other decision makers.
One thing you should know about me is that I’m an avid watcher of CNBC and have been since the 1980s.
[MARK ZANDI]
Oh.
[PAUL DECKER]
So I know you as a prolific and popular voice on economic data and evidence, and I’d love to hear more about your approach to communicating data insights and influencing decisions. I know you’ve used a variety of venues to promote evidence. I’ve seen your name and your face and heard your voice everywhere, whether it’s through op-eds that you write, appearing on TV, using social media, or now through your weekly podcast. How do you think about the benefits of using those different venues, and have you found that different media that you access reach different audiences that you’re trying to target?
[MARK ZANDI]
Well, that’s a good question, thank you. I follow all of the above strategy – whatever media that is available, I’ll try to utilize – in part because I enjoy engaging in all these different – with all these different approaches, from op-eds to podcasts to webinars. I tweet. My handle – I’ll advertise that. So I’ve been tweeting. I’m not quick to adopt new forms of communication. I don’t move quickly; but once I see that some form of communication is popular and people are listening to it, I hear people discussing what they learn from the different platforms that they’re on, I’ll adapt and adjust and start to contribute.
Clearly each platform media has its different advantages and disadvantages. I mean, op-eds, I write those because policymakers and others that are thinking about these issues that I’m writing about. I generally have a point I’d like to make and get across, but that takes more energy, more work in the sense that it’s not – I try to vet it. It’s an idea, and I send it around to other people to see what they think and bring in their suggestions and comments. So it’s a fair amount of work.
Of course the media outlets that publish the op-eds, some are very – they take what you write pretty much as is. Some not so much – they really have a heavy hand in the editorial process. They don’t want to change what you’re saying, but they really pay very close attention to how you say it. They’re very careful about making sure that everything is annotated and well-documented, that kind of thing.
On the other end of the spectrum – and by the way, I also write and do more basic kind of research. They’re not peer-reviewed academic research, but they’re more involved. I call them “white papers” on specific issues that kind of deal with things that I’m deeply interested in but are kind of in the weeds. So for example, I wrote a paper – and I usually do this with colleagues both internally and externally – wrote a paper on – there’s a lot of controversy around the Federal Home Loan Bank system and its (inaudible) and should there be reforms and that kind of thing. So I wrote a few white papers around that. They go into more detail with regard to thinking about that particular issue.
A lot around reform of Fannie Mae/Freddie Mac – the GSEs, mortgage finance, housing, those kinds of things. But that’s very involved – takes a lot of time and energy. So right now, I’m working on a research piece, white paper, on the private credit markets because that’s been a rapidly-growing part of the global financial system.
Then on the other end of the spectrum are things that are just easy, they’re fun actually and I really enjoy. Like tweeting – I actually enjoy it to some degree. I actually enjoyed it more when it was 280 characters because that’s actually quite therapeutic. If you want to get down a real clear thought in 280 characters, that means you’ve got to be quite – you’ve got to really think about what you want to say. You say it very clearly, succinctly. It’s almost like – this is overstating the case, but it’s almost like writing a haiku. It’s like--
[PAUL DECKER]
Disciplined.
[MARK ZANDI]
Yeah, it’s a discipline. Now that they don’t have that constraint, I still kind of try to stick to the 280 characters because of that discipline.
Then, the podcast I just enjoy. I don’t really have to prepare for it. I do a little bit of – like I’ll write an email the night before or two nights before to the guest, if a have a guest, and say, “Hey, this is kind of broadly what I think we’ll talk about. What do you think? Any suggestions? Anything you don’t want to talk about,” so forth and so on. But that’s it; and then it’s a conversation, and they have a lot of fun. I think it’s really good for – people really enjoy it. I get a lot of comments from my clients and other parties, external parties, that listen to the podcast; and it’s really good for my staff because they’re engaged and we’re talking about issues that they’re passionate about. You can see how talented they are, and they can express their views in a kind of more relaxed way. So I’ve found that to be a really effective way of reaching people but at the same time really having a lot of fun doing it.
[PAUL DECKER]
Is there any particular experience that you can put your finger on where you thought your insights and the way in which you communicated those insights was influential?
[MARK ZANDI]
Well, I think I do testify in Congress a fair amount – generally more when there’s economic issues. I guess if I had to pinpoint one example where I thought I had some impact was – and generally, congressional hearings are more staged. They’re not like where you actually make policy. People are making points, but it’s not really making policy.
But there was one hearing in the financial crisis around the auto bailout. If you remember back, the auto industry got bailed out during that period; and this was the second auto bailout hearing. The first one was a disaster because all the CEOs came down in their Lear jets for a bailout.
[PAUL DECKER]
I remember.
[MARK ZANDI]
People got pretty upset, so they called a second hearing. Then they had the CEOs – the three CEOs of the major companies – and they had me. I remember this because this was a Senate – I think it was Senate banking. I’m pretty sure it was Senate banking, and they were trying to figure out how to do this. First of all, should they do this – should they bail out the auto industry and how they should do it. I remember it was like a five-hour hearing, and there’s no bathroom breaks (laughing). You have to – and I didn’t know that, and I’m looking around at the other CEOs two-and-a-half hours in and saying, “Hey, are we going to get a break here?” They kept coming for five hours (laughing).
Anyway, they were looking for a path forward on how to do this – the structure of bailout. I felt like at that hearing – that actually was policy being made at the hearing and had an impact, I think, on how people thought about that. So I take pride in that. That was a really interesting experience. But if I had to pick one experience where I thought I had some significant influence, real-time influence, that would probably be the example.
[PAUL DECKER]
When we talk about Mathematica’s work, we talk about the importance of Mathematica being advocates for evidence; that is, we want to stay objective and nonpartisan in our work, and we want policymakers to incorporate evidence in their decision-making process. It can be a tricky balance to strike – politically-charged environment – being objective, nonpartisan, and yet still being a forceful and influential voice to drive the kind of impact that you just described.
So I’m very impressed with how you do that. I wonder, how do you strike that balance? How do you maintain an objective nonpartisan posture while engaging in those policy debates?
[MARK ZANDI]
Well, that’s very kind of you to say. I appreciate that.
Well, a few things – one, I think all the work/analysis/all your opinions has to be rooted in data and research. If you can point back to here’s the datapoint or the time series or here’s kind of the preponderance of the economics literature – or you call upon institutions that are nonpartisan, like the Congressional Budget Office and they do fab -- I’m on the Board of Economic Advisors for CBO, and they do fabulous work. If you can call upon that, that’s also very helpful.
Second, I call it like I see it. I’m not going to say something that I don’t believe in. I’m going to say things that I believe in, and I try to give people a sense of how confident I am in what I’m saying. Like some things I’m confident in, some not so much. And I try to make sure that people understand that – that there’s a – this I feel really strongly about. I’m confident in what I’m saying over here. Over here, I’m not sure; but here’s my kind of view and my opinion. I’m not afraid to – you should be afraid to say something that the person you’re to or the group you’re talking to may not like. In fact, that adds credibility to what you’re saying. I mean, if you agreed with everything that the other people were saying or thinking, then that’s not – that’s just generally not credible. So the fact that you’re saying something that’s not kind of in their thinking, that they would disagree with, then they value and put more weight on everything else that you say because they know that you’re going to call it like you see it.
Then the third thing is you need to be transparent and clear. You can’t obfuscate. You’re not a politic – I’m not a politician. I’m not going to do that, and I’m going to try to be as transparent in what I say and how I got to what I’m trying to say so that people know where I’m coming from, that kind of thing.
And then I think it’s important to get the message out there and use those different platforms and media venues to get your views out there and expressed in a way that is consistent with what you’re trying to say. Because everything gets spinned, and you want to make sure that you have the platform and the voice to get the message out that you’re trying to get out in the way that you’re trying to get it out there and the framework so they can’t be pushed in a certain direction that makes it less – makes it more biased and less useful.
So all those things, I think, are important, Going back to all the different media and the platforms that I use, I mean at the end of the day, a big part of that is to establish credibility because if you don’t – if you’re not – if people aren’t quoting you, if you’re not out there in the public domain and expressing views, you don’t have – you’re not credible. You don’t have a brand, and you need that brand and credibility to be able to establish the trust in what you’re saying and maintain that unbiased – the perception of being unbiased. You want to be unbiased, but it’s also really key that there’s the perception of being unbiased; and that goes back to your credibility. And all those other things that we were talking about that I do is an effort to establish and maintain that credibility.
[PAUL DECKER]
Yeah, I think your characterizing it as brand is really important; but that takes an effort in recognizing that brand and taking action that makes that brand consistently recognized in the market.
[MARK ZANDI]
That’s a good point, yeah. Mathematica’s brand is stellar. It’s gold-plated, and you guys work at it really hard; so you’re really good at that.
[PAUL DECKER]
So in talking about influence, although you and I are committed to the evidence as you know in policymaking there are lots of factors that enter into the picture and considerations. Years ago, my friend Ron Haskins used to have a presentation where he had one of these pie charts that was representing all the different factors that play into policymaking. The pie chart was intended to give you a sense of what were the most influential factors. I think for research and evidence, I think that slice of the pie he had representing about 1% contribution to the process.
[MARK ZANDI]
(Laughing) All right, I’d like to see that pie chart – interesting, yeah.
[PAUL DECKER]
So in reflection on that, do you see signs that data and evidence are becoming more central to the way the Federal Government works today; and are there areas where you’d like to see a stronger focus on data than we have right now?
[MARK ZANDI]
Well, I’m surprised it was 1%. That’s not my sense of it. I think data and kind of the analytical basis for the policy it’s making is absolutely critical to success. I think to succeed legislatively, there’s a lot of elements to that for sure; and there’s a lot of politics involved for sure. But the idea has to be a good one. It has to be rooted in something that is real in data, in research, in analysis. That’s why in Washington there’s so many different think tanks that are focused on just that – establishing the kind of intellectual basis for the legislation that’s being put forward.
That kind of ebbs and it flows to some degree. I mean, it depends on kind of the political environment – who’s president of the United States and kind of their perspective on things. But generally I find that to be true. I think that lawmakers are very interested and really focused on what the data and what the analysis and research say. So I find that I’m less jaded than that; 1% sounds too low to me. I’d say maybe one-third of the pie should be that way. I don't know that that – again, it ebbs and it flows. I don't know that it’s changed – at least in my kind of 30-year experience with Washington, the nexus between economics and politics, i haven't noticed that changing in kind of a trend-like way, structural way. but it definitely ebbs and it flows. Sometimes it’s more important, sometimes not so much.
I do think it would be very helpful if there was more resources for better data collection. I mean one thing that’s happening that’s a real problem – and this is going to become a bigger issue going forward – is that the quality of the data that we’re using is starting to erode because most of – a lot of it is based on surveys, and survey response rates are way down across the board – not just government surveys, all surveys. A lot of discussion/debate as to what’s going on – I’m sure some of it’s survey fatigue. I don't know about you, but everything I buy or every plane I get on or every car I rent or hotel I go to, I get a survey saying, “How did I do,” because everyone’s focused on these so-called NPS scores and promotor scores, that kind of thing. It’s a way of evaluating your business’s kind of objectives and scorecards and that kind of thing.
There’s concerns around privacy, cyber issues; but for whatever reason, the response rates are way down, and that’s beginning to affect the data to a significant degree. That’s why to some degree, I don't know if – you said you listen to CNBC, so you’re a consumer of economic information and data – you’ll see – you may have noticed there seems to be a big revision to the employment numbers that we get every month. That partly goes to the fact that the response rates are low, and the responses are lagged. Businesses are responding later and later to what’s going on.
Funding for government agencies that collect the data, clean the data, disseminate the data really has not kept up. In fact, to some degree, it’s gone backwards. So I think that’s a really serious problem. We can’t make good, informed decisions; we can’t do the kind of work we need to do to make good policy unless we have good underlying economic information and data.
Now fortunately, private institutions – companies and other private institutions – are kind of coming in and creating their own sets of data. For example, we use information from credit files from the credit bureaus – anonymized, obviously – and Moody’s has its own consortiums. We collect a lot of data on how companies are performing, their balance sheet and income statements, defaults on different types of loans; and that’s helpful. You can think of a lot of different other companies that are providing more information and data; but nonetheless at the end of the day, the government needs to be able to provide consistent, accurate sets of data so that we can really understand what’s going on.
In fact, you can see it or the problem it’s creating. Just go over to the UK. The British can’t even collect good employment rate data; and if you can’t have a good fix on unemployment, how do you accept monetary policy? So that’s a real problem for them, and that’s the direction we’re headed here in the United States if we don’t provide more funding and try to figure this out.
So if I had one thing – if I were king for the day and I could devote resources to one thing, that would be it. Let’s go make our data sources more resilient, better, more comprehensive, more timely.
[PAUL DECKER]
I think the other thing in addition to some of the five-minute or big data sources that you touch on, which can be the result of transactions that are trapped more than they’ve ever been in the past, the factor that at least plays into our work is the availability of the administrative databases that are effective tools for research in ways that they weren’t when my career started.
[MARK ZANDI]
Hmm.
[PAUL DECKER]
So a lot of times when we’re studying programs, we’re studying it based on the database that’s created by the program that’s being run. So that’s really changed the equation too in terms of the data sources that we use.
[MARK ZANDI]
Yeah, I’m on the board of this nonprofit. Maybe you’ve heard of them, the Coleridge Institute?
[PAUL DECKER]
Yeah.
[MARK ZANDI]
So it tries to facilitate the use of government data within the government – state, local, federal – and set up the different platforms for researchers to use data and share data – government data – within the government. I think that still hasn’t gotten to scale, and you probably know this space better than I because I just recently joined the board. But that feels like a positive development – to try to – because there a lot of silos within the government itself with regard to sharing information and data, which is going to be very productive if we could break down those silos and allow the researchers to access different datasets cross the government. That’s the mission of Coleridge.
[PAUL DECKER]
Yeah. Mark, I’d also like to get some of your insights on the key policy issues that we’re working on and that we face as a country. For a number of years, we’ve had an aging population in the United States. This long-term trend has created downward pressure on the supply of workers. Also it has brought into question whether we can provide adequate support for retirees through programs like Social Security and Medicare going into the future. I can remember when this was being predicted several decades ago in my Econ 101 class.
[MARK ZANDI]
Yeah.
[PAUL DECKER]
So this threat has been with us a for a long time, and we are feeling the effects of it. Now, the pandemic may have exacerbated the problem in some ways by sending older workers into retirement earlier than expected; but it affected workers in other ways as well. But I’m curious as to how you see the future implications of our demographics in the labor market. Are the worker shortages inevitable from those demographics, and what impact has the pandemic had on your thinking around that trend?
[MARK ZANDI]
Yeah, as they say -- I can’t remember. You may know, Paul. Who said demographics is destiny? I can’t remember the economist that said that, but that’s certainly true. In the long run, demography determines a lot in regard to the economy’s performance and then everything else that depends on the economy’s performance – like Social Security and Medicare, the Government’s fiscal situation.
Abstracting from foreign immigration for a second, immigration remains kind of consistent with what it has been over the last – until recently – over the last couple/three decades – about a million immigrants per annum. It appears that we’re going to suffer kind of a perennial labor shortage because the native population is – because of the aging of the population, the retiring of the Baby Boom generation, my generation – labor force is coming to a standstill, and it’s going to decline. So we’ve got this very significant weight on the working age population, labor force, that’s going to create these potential significant labor shortages.
To some degree, we’ve been experiencing this already. I mean even before the pandemic, back in 2018-2019, the number one business problem was lack of labor. Unemployment was very low. The pandemic scrambled things, exacerbated the labor problems. Here we are on the other side; we still have a very tight labor market. Again, unless abstracting from immigration, you look forward. This is going to remain an issue/a problem for businesses and for the economy to grow.
Now having said that, there’s two possible wild cards here. One is immigration. We’ve seen mass immigration obviously in the last several years. The Congressional Budget Office, the CBO – I mentioned them earlier – they came out a study, an outlook on demographics, because as you rightly said, the budget – the fiscal outlook which the CBO is focused on – depends on these demographic trends. They’ve estimated 3.3 million immigrants came into the country in 2023 on top of 2.6 million the year before. So remember, we were at a million pretty much every year give or take -- both legal and undocumented; and now last year, we’re at 3.3. We’re going to be at least that this year, and it’s going to be very high for the foreseeable future.
So one way or the – one way where this kind of picture of a very tight labor market might change is if we see continued significant migration into the country. Now hopefully, we have reform so that we have rational immigration, get the right kind of immigrants into the country. We’re able to handle and manage the immigration that’s coming. Right now, it’s unmanaged and creates enormous challenges to communities across the country and the nation. So there are benefits to it, but there’s cost to it as well; and we could do a lot better there. Hopefully, we will eventually do that; but immigration could be a way to get more growth and address some of these longer-term fiscal issues that you mentioned given our domestic demography.
The other obvious wild card is technology, artificial intelligence, and a lot of kind of hammering about this. There’s – AI, you’ve got folks on both sides of this. Some folks thinks AI is just great because it’s going to improve labor productivity just enough to ensure that the labor market doesn't get so tight that it constrains our ability to grow. And goodness knows, we want productivity to grow. That’s the key to creating wealth in people’s incomes and profit and is the fountain of growth in our living standards. But the (inaudible) view is that the AI will come on so fast that it’s going to wipe out all these jobs, and we’re going to have not a problem with finding labor; the problem is going to be not enough work because AI is going to take all those jobs.
You know, it’s a wild card so hard to know. But if history is any guide, it’s more likely than not that AI is going to be more a plus than a negative – that the technology is going to diffuse through the economy over an extended period of time allowing labor productivity growth to improve enough to help us out here but not come on so fast that it’s going to wipe out all these jobs and the problem is going to be high unemployment. So that’s the lesson of history based on other technologies – the Internet, electricity, all the other big technological innovations that have occurred in the last century. But that is a concern, so that’s another wild card – those two things.
The final thing I’ll say, I think a prudent planner would plan on we’re going to back to about a million immigrants a year. AI is going to be helpful but not a game changer, and the problem going forward is more likely going to be labor shortages than a surfeit of labor. But that’s a lot of uncertainty around that particular forecast.
[PAUL DECKER]
Yeah, appreciate your point in AI and immigration; and maybe a third factor that doesn't have the same magnitude of impact is work from home or workplace flexibility. I’ve been interested to see the female labor participation rate is higher than its been historically, and I’m sure some of that can be attributable to the availability of people working from home in cases where they might otherwise drop out of the labor force.
[MARK ZANDI]
So you’re a proponent of remote work? Many CEOs are not.
[PAUL DECKER]
Well, we take a very flexible approach at Mathematica and partly for some of the reasons that relate to the supply of talent. So rather than being one of those companies that kind of prods its employees in three days a week or something like that, we pretty much leave it up to staff; but we provide the offices, or the office space, for those that want to work in the office on a regular basis and also recognizing that we want to continue to create connection between our employees. It’s an important part of the corporate culture. So we’ll navigate that over time. I didn’t want to be in a position to assume that because the way we’ve been successful in the past is by being face-to-face on a regular basis if that was the only way to do it going into the future.
Individual preferences for flexibility is strong enough that I thought it was worth at least seeing if we can achieve both benefits – both benefits of flexibility but also finding ways of maintaining corporate culture, creating connection even if we’re not necessarily doing it daily on a face-to-face basis.
[MARK ZANDI]
Hm, yeah, we’re fully remote – not all of Moody’s but each unit can decide what they want to do. But the Economics Team, a couple hundred people around the world, we are fully remote. We decided to be fully remote. But there are interesting challenges because the rest of the organization isn’t quite – we’re part of a large multinational, right? And if the rest of the organization isn’t quite there, it complicates things for us because we can’t take full advantage of being remote because of these constraints.
[PAUL DECKER]
I understand.
[MARK ZANDI]
Yeah, but it’s interesting and so far, so good. We were fully remote since the pandemic hit based on our measures of productivity; and that’s a whole other ballgame, how to manage productivity well. It feels like – productivity hasn’t accelerated, it hasn’t decelerated, and people are happier so why not?
[PAUL DECKER]
Yeah, that’s our finding as well.
Speaking of one of the issues that you touched on, immigration. You highlighted how it could be one of the solutions to our labor shortage, and I agree with that. But the political conversation around immigration doesn't reflect that economic reality. So when you make the case for immigration reform, have you found that there’s specific datapoints or kinds of data that are particularly compelling in that discussion; and is there something you think can cut through that partisan gridlock on immigration?
[MARK ZANDI]
Well, I don't think we can make much progress here until we kind of – use the hackneyed phrase “secure the border.” As long as you’ve got three million plus people coming across the border, it literally – I don't know if you saw that 60 Minutes episode where they had a camera on one part of the border wall, and people we’re just walking through. By the way, what I found so fascinating was it wasn’t just kind of folks from Venezuela or other parts of Latin America. It was a lot of Chinese – middle class Chinese – leaving China, finding their way to Latin South America and making their way up. So I found that quite fascinating.
But as long as that’s the case, I don't think we can have a good, rational conversation around this because it’s just creating such havoc in different communities across the country. You can see it in New York and Chicago and San Francisco. You can see it in Florida and Texas. That really makes the conversation very, very difficult and impossible to come to come to any kind of a resolution. So hopefully, we can nail that down. It felt like we were getting pretty close to a piece of legislation recently, but that got shot down in the political process. So we’re not going to get it this go-around. But hopefully on the other side of the election, we get something that helps the next president secure the border.
Then once that happens, I think – I’m an economist. I think of everything through the prism of economics. If we do have a tight labor market and perennial shortages, I think the politics will shift, right? I mean, if we have perennially low unemployment, then labor isn’t going to be so concerned that the immigrants are going to take their job; and the jobs that are being taken aren’t ones that they’re going to want anyway. And businesses will clamor for rational immigration reform, and I think we’ll get it. We came pretty close once in 2013, and I think we’ll get there again.
The datapoints that could really help – the best datapoints I think are other examples, like go look at Canada. Look at the Canadian immigration process. It’s pretty good, and it’s reaping enormous benefits. The Canadian economy has ups and its down for various reasons. It’s tied to the U.S. economy. But the kind underlying trend growth in that economy has improved dramatically because they’re allowing a lot more immigrants and the right – when I say right kind, kind that are going to bring talents and skills. Although we need immigrants with all skill levels – no skill to doctors and lawyers and technicians and everything else. But they’re really good at it, so I think if we have case studies of – and then you go to other countries, where they don’t allow immigrants altogether.
You go to like China, for example, or even in Japan – Germany has trouble with immigration. You can see these contrasts between economies that are allowing immigrants to come into the country in a rational way and those that don’t; and the economic performance and prospects are very, very different. So I think that’s probably – if I had one thing I could use to make the case, I think that’s a pretty compelling set of information and data -- just looking across country at what’s going on and looking at those immigration policies.
[PAUL DECKER]
Yes, I think the case of Canada conjures up a philosophy about this that we have to think a little bit more about being in competition for immigrants as opposed to just focusing on the burden of immigrants. Because as a business leader who talks to other business leaders, business leaders are going to find the talent; and if they have to go overseas to find the talent, they will. So it’s better for the U.S. if we have them doing the work over here and paying U.S. taxes than if we just rely on people doing it remotely from other countries.
[MARK ZANDI]
Absolutely. I started an office in Prague many, many years ago looking for talent. Eastern European folks are very talented and very technically competent people. They have programmers and statisticians, data people, economists – economic modelers. I heard this one fellow and I couldn't get – I was thinking about bringing him here, but it was very difficult to get a visa. So I said, okay, Prague’s cheap. This was 15 years ago – maybe even longer, almost 20 years ago. I was very cost-effective, and we set up shop there. Now we’ve got a huge office in Prague where – and that guy has now left and gone off and done other things. If that guy had come over here and done that here and gone off and done other things, he would have created a boatload of jobs here, created a lot of income and wealth and tax revenue and so forth and so on. But the reason it didn’t happen is because of our crazy immigration laws just as an example.
[PAUL DECKER]
You mentioned AI; and obviously AI is a hot topic right now and something I think you write about well, and you’ve been doing it for a while now. So I appreciate your familiarity with the technology. You talked a little bit about projecting out what it may mean for the changes in the labor market, and I agree with you that I think on balance that the historical experience is that technology is a positive in the labor market. How significant do you think the impact of AI will be in the labor market? You mentioned past technological advances like the Internet. Is AI going to be bigger than the impact of the Internet? How do you see it?
[MARK ZANDI]
Yeah, I think it’s a big deal. I don't know that it’s a gamechanger. It’s not Internet or electrification, but it’s going to show up in the data. So just to give you context, we’ve increased our real GEP growth forecast over the next decade by about 15 basis points, 0.15 percentage points, on average over that 10-year period, closer to zero over the next couple-three hers, higher than that as you make your way into the second half of the decade.
Just to put that into clearer terms, without AI say the economy would grow 2% per annum over the next 10 years. With AI, with a boatload of assumptions, 2.15 percentage points. So you say – one perspective is, oh, that’s no big deal; 0.15 percentage points per annum adds up to real money, so that’s consequential. And that’s additive; that’s juice on top. So it’s going to be – productivity boost from AI is going to be more than 0.15, but that’s 0.15 above kind of the typical without that added boost. So it’s consequential, but I don't think it’s game changing. It’s not like we’re going to restructure big parts of the economy wholesale because of the influence of AI.
Again, I say this going back to an earlier point. Some things I’m confident in, some not so much. It’s very difficult to gauge, and I’m relying on historical experience here and other. I think of it more like – it’s not the Internet, but it’s kind of like wireless. Wireless was important and it did add a lot to economic growth, but it wasn’t this game-changing thing.
The other thing – I mentioned this earlier, I’ll just mention it again – it takes a while for the technology to kind of diffuse through the economy. In fact, initially it could be counterproductive. Like take our business – we’re trying to figure out ways to use AI, but we haven't yet. But we’ve invested in trying to figure out how to use AI. That’s lowering productivity; that’s not adding to productivity. Now I expect that to change, and I think it’s a really good investment to make. We have to make it; but at least initially, it’s more of a restraint than a source of growth.
Then ultimately, the real benefit to any technology – and I think this is going to apply to AI – is when new business is formed and optimized around the new technology, around AI, because existing businesses when they try to gerrymander what they’re doing to the new technology, they don’t quite get it right. They don’t have the right people. They don’t have the right resources. You have to restructure your organization, and that’s difficult to do. But if you’re a new company starting de novo, you can say, oh, okay, let’s organize this way. We’ll make these kinds of investments because this will optimize this technology. So I think that just takes time. As new businesses form, they’re going to optimize.
Just like you’re going to optimize around remote work in general, I don't think anyone is going to optimize around the traditional office space. I just don’t think that’s going to work. People are going to – new businesses are going to optimize around the use of AI and other technologies. That’s when you get the productivity gains, and that’s why it takes – it’s diffused through time. It’s not this game-changing – we’re going to get a five-percentage-point increase in productivity on one year. It’s something that’s spread out over time as new businesses form. So that’s kind of how I’m thinking about it; but again, I say it with a lot of intrepidation because forecasting how technologies evolve and how they’re adopted by businesses and whether that lowers productivity or not is a pretty intrepid affair.
[PAUL DECKER]
One of the things that’s different about AI versus past disruptions is the degree to which it represents change in the labor market; that is, the group that expects to be affected is different than the profile of workers that expected to be affected in the past. AI is expected to have an impact on workers that are in occupations that are highly educated, highly skilled, highly paid; and that could have unique pressures versus past disruptions. But I wonder how you’re already seeing it in our field in terms of how it's changing our work around data-driven social science.
[MARK ZANDI]
Well, so far the biggest change is around the value I think companies are placing on data itself, the underlying fodder for the AI. I mean, AI only works and is trained on underlying information and data. So that has kind of advanced the value of that underlying information and data. So I think businesses are really focused on, oh, what are my data assets? What are my analytical assets? What do I have that could be useful to training AI to perform functions better?
Companies are working on – I mentioned the data consortiums and working together to come up with bigger datasets and information and platforms so they can use that to train their AI algorithms. Then, I do think the other thing that’s going on is we are investing. Businesses are starting to invest in people and in the technology necessary. That’s why Nvidia is now worth $2 trillion because everyone’s buying their chips to drive these AI engines, and so we’re making a lot of investment in that.
Then, I do think a lot of experimentation is going on trying to – I think kind of the next easy step is to come up with AI tools that are helpful in guiding users through the blizzard of information and data that companies have. Moody’s is like this. I can’t even keep up with all the things we’ve got. It’s all over the place. There’s all kinds of stuff that I go, oh, whoa, what’s that? It’s like detailed information on foreign direct investment – really, we collect that data information? I didn’t even know that. So we’re creating AI tools to be able to understand that information and data, bring that all together in a way that we can use it more effectively – and I think AI guides, things to kind of help us.
So those are the kinds of steps we’re taking so far, and there’s a lot of incentive to do it because shareholders are enamored by companies that are able to show that they’re taking advantage of AI and improving the kinds of services and information they provide and also improving productivity. So there’s a lot of interest in pursuing it. So I think I expect a lot of good things to happen here in the not-too-distant future. But those are the steps that, at least in my experience, have been taken so far.
[PAUL DECKER]
Mark, I rarely have discussions with folks that are experts at your level on macroeconomics. So I’ve got one final question and it’s federal debt.
[MARK ZANDI]
Yeah, fire away.
[PAUL DECKER]
As an organization that’s dependent on federal spending for a lot of the work we do, you can understand the interest. We now have accumulated U.S. debt that’s surpassed $34 trillion, which is equivalent to 123% of annual GDP. I presume that’s the record for that ratio. If it’s not, it’s at least the record probably in non-war times.
[MARK ZANDI]
Yeah.
[PAUL DECKER]
What do you think is the most likely endgame for U.S. debt and budget deficits? Can we simply grow our way out of this current predicament, or is there something that’s going to happen in the future that we should begin worrying about?
[MARK ZANDI]
Well, this is a problem. I mean again, going back to CBO, they do projections of the fiscal situation – that debt to GDP ratio you were talking about. If there’s no change in policy – we just take existing policy and extrapolate forward – the outlook is daunting. The debt to GDP ratio is going to rise about 80 percentage points over the next 30 years; and that’s when the forecast ends. You can do your own forecast after that. It’s just going to continue to grow very rapidly. That means that we’re going to be paying increasing amounts on just the interest on that debt. Interest is a share of GDP or interest in a share of tax revenue is already rising pretty quickly. In fact, I think we’re pretty close – if I have my facts straight, someone could check it – the amount of interest we pay on the debt is pretty close to our budget on defense.
So I think if that happens and people recognize that it makes no sense – like really, we’re spending more on interest than we are on our own defense and fact, half of that interest goes to foreign investors – Chinese, Japanese, Middle East, UK, not just American investors -- that people are going to say that doesn't make a whole lot of sense. We should do something about this.
I don't know exactly how this plays out; but if history is the guide, we come up to some kind of period where there’s a forcing mechanism where lawmakers have to come together and make some decisions and at that point in time, they’ll make some decisions that will start to bend that forecast to make it more manageable going forward. For example, in early 2025 after the next presidential election, we’re going to have to raise the debit limit again. That’s a foreseen mechanism. The Trump tax cuts for high-income/high net worth households, they expire at the end of ’25. Something has to be done about that. Obama Care tax subsidies come due. That has to be resolved.
So I suspect that will be a point in time when lawmakers will have to make some decisions and can make some decisions. They don’t have to make big ones right away, just things to kind of bend that curve and make it moving in the right direction and hopefully show the world that we do have a strong enough governance structure that we don’t do stupid things – like you shut the government down or, God forbid, breach the debt limit.
Now having said that, to really solve this problem or address it sufficiently so it’s a problem for two generations down the road – and that’s okay because things happen and then who knows – we may need to see some kind of – I want to say forcing of that crisis, where bond investors say, “Hey, guys, I’m afraid you’re not going to pay me back on time. You’re going to have to pay me a higher interest rate to compensate for that risk.” The interest rates really jump and take off and stay up. They don’t go up and come back down. They go up and stay up.
That may be necessary because for lawmakers to connect the dots in the minds of the electorate as to why we need to change our fiscal path, they need something to point to and say, “Look, if we don’t change our fiscal path, this is economic ruin; and you can see it right here in that we’re paying 6%-7% on a 10-year Treasury Bond.” So we may need to see something like that to generate the political will necessary to do it.
But, Paul, I’m going to say one last thing; and I’m going to butcher it. But a quote from Winston Churchill said something like Americans – and this, he’s talking about back in World War II trying to get Americans into World War II and help them out –he goes, Americans, they’ll try anything and everything and then they’ll ultimately do the right thing (laughing). Now, that’s not exactly what he said; but that’s what he meant. Based on my 30 years between the nexus of the economy and politics and observing and trying to participate in what’s going on in Washington D.C., I firmly believe that to be true.
We try everything. We look down every road, every path. We game out every single scenario, and then ultimately we come up with something that works. So I think that’s the history of America, and I think that’s the history we should expect. That should be the future that we should expect for the United States, for our country.
[PAUL DECKER]
That’s great, I love ending on a Winston Churchill quote.
[Laughter]
[MARK ZANDI]
There you go. Thanks so much for having me. I really appreciate it, Paul. You’re very kind.
[PAUL DECKER]
Mark, thanks for your insights. I appreciate your (inaudible) a range of topics as well as you do.
As we wrap up, could you share with our listeners where we can find you each week?
[MARK ZANDI]
Yeah, you mentioned my podcast. It’s the Inside Economics. It’s on all the platforms, so you can find that. We do that every week. I’m on Twitter, @MarkZandi, and you can find us – I publish a lot of stuff free and paid on a website called Economic and You. It will cover all the indicators around the world. But I post a lot of my articles, kind of deeper thought pieces, on there for free; so people might be able to enjoy that website as well. So those are some good ways to stay in touch.
[PAUL DECKER]
That’s great. Thank you, Mark.
[MARK ZANDI]
Thanks, Paul, take care.
[J.B. WOGAN]
Thanks to our guest Mark Zandi. Thanks to Mathematica’s Paul Decker for stepping in as the guest host for this episode.
Last month, we celebrated the five-year anniversary of Mathematica’s On the Evidence podcast. Whether this is the first time you’re hearing us or you’ve been with us since 2019, thanks for listening. If you’re a fan of the show, please consider leaving a rating and review wherever you listen to podcasts.
To catch future episodes of the show, subscribe at Mathematica.org/OntheEvidence.
Show notes
Listen to the Inside Economics podcast that Zandi hosts for Moody’s Analytics.
Read Paul Decker’s blog about how Mathematica used evidence to guide the company’s approach to reopening in-person offices after the COVID-19 pandemic.
Read an op-ed co-authored by Paul Decker for the website RealClearPolicy, which proposes comprehensive immigration reform and other solutions to address current labor shortages.
