Joe Biden is facing re-election shepherding an unsteady economy through high interest rates. Is this Democrats’ bad economic luck repeating itself? Joe Stone finds that Democratic presidents have regularly expanded economies at the beginning of their terms but Republicans benefit from expanded economies leading up to the election. And partisan misalignment with the Federal Reserve also dampens economic performance. William Clark finds that the Fed raises interest rates as elections approach under Democratic presidents and lowers them for Republican presidents. And the Fed is more responsive to inflation under Democratic presidents. They both say the historical patterns may help explain why Biden finds himself in a more difficult position. 

Guests: Joe Stone, University of Oregon; William Clark, Texas A&M University

Studies:Presidential party affiliation and electoral cycles in the U.S. economy“; “Independent but Not Indifferent: Partisan Bias in Monetary Policy at the Fed.” 


Matt Grossmann: Does the Biden economy have bad election timing or an unfair Fed? This week on The Science of Politics. For the Niskanen Center, I’m Matt Grossman.

Joe Biden is facing reelection shepherding an unsteady economy through high interest rates, but some think he should be getting a better shake. If Democratic presidents oversee strong in economies, why don’t they benefit electorally? And why does the Federal Reserve not seem to lower interest rates in time to help?

This week, I talked to Joe Stone of the University of Oregon about his Journal of Macroeconomics paper with David Jacobs, “Presidential Party Affiliation and Electoral Cycles in the US Economy.” He finds that Democrats have expanded economies at the beginning of their terms, but Republicans benefit from expanded economies leading up to the election and partisan misalignment with the Federal Reserve also dampens economic performance.

I also talked to William Clark of Texas A&M University about his Economics & Politics article with Vincent Arel-Bundock, “Independent but Not Indifferent: Partisan Bias and Monetary Policy at the Fed.” He finds that the Fed raises interest rates as elections approach under Democratic presidents and lowers them for Republican presidents and the Fed is more responsive to inflation under Democratic presidents. They both say the historical patterns may help explain why Biden finds himself in a more difficult position this year.

Let’s start with Stone, who explains why Democratic presidents don’t get credit for overseeing better economies. Their advantage has been overstated and it’s not well-timed. Tell us about the main findings in your paper on partisan business cycles.

Joe Stone: I’ll try to be brief. First, we find a party difference in economic growth of just under 1%, but that’s only half of prior estimates. I should add, widely publicized prior estimates. Second, and consistent with prior work by Nathaniel Beck, we find higher growth of roughly 3% in the first half of terms for Democrats countered by higher growth of roughly 2% in the second half of terms for Republicans. Three for Democrats, their first half, two for Republicans, their first half, so the difference is the 1% we find as opposed to the 2% others find. Third, by allowing the late term pre-election growth to differ by party, we find a surge in pre-election growth for Republicans. Fourth, we find that party affiliation alignments between the President and the Chief of the Federal Reserve yield 1% growth for both parties. That’s my quick run through.

Matt Grossmann: You find that Democratic presidents have expanded economies at the beginning of their terms, so what are the potential mechanisms that you or others have forwarded for that difference?

Joe Stone: In work now, a couple of decades old, a co-author, Steve Haynes and I published a paper showing that both the early surge for Democrats and the late surge for Republicans, smaller, appear to be linked mostly to fiscal expansion. I’ll take it from the perspective of Democrats. You have a newly elected Democratic administration. They have lots of ideas for what they’d like to accomplish, traditional sort of liberal agenda for new programs, so they start in on them. And so, that takes funding, of course. They start funding these programs that expands the federal budget and spending, and so that works its way through to higher growth in that first half of their term. Eventually, you begin to see a little rise in inflation. And so, what you see is Democrats tend to have slightly higher growth, but also slightly higher inflation compared to Republicans.

Matt Grossmann: Does that lead into why Republicans benefit leading up to an election? Are they doing stuff at the end of their terms or is it just that the Democratic early advantage gets paid back a little bit later?

Joe Stone: It’s probably some of both. I think, in a sense, the Democratic propensity to pursue their partisan objectives for funding new programs early in the term, that’s great, but later in the term you see an advantage for Republicans from fiscal expansion and that’s coming closer to the election. And while, overall, there’s a small advantage for Democrats counting both parts of their term, the advantage for Republicans, even though it’s only part of the term, comes closer to the elections.

There are multiple threads of partisan differences and economic growth and elections. And so, I won’t go through all threads, but there’s the Nordhaus thread of political business cycles and the Ray Fair thread of economic effects, what determines election victories and economic terms, and then, the most recently, there’s the piece, the result we were intrigued by, which was the Blinder-Watson finding of a 2% overall advantage for Democrats.

We thought we had a better way to go at that question than they did. And so, we find about a half of what they found. Most of the effects, aside from the difference in the two terms, the early term and late term differences, the effects, overall for the terms, actually for both parties, are small, so that doesn’t necessarily set the headlines on fire, but they’re really pretty small.

Matt Grossmann: You also find that the aligned Federal Reserve matters. Tell us about that evidence and we supposedly have an independent Central Bank, so how does that work?

Joe Stone: We were really intrigued. We began trying to find what are the mechanisms that work. We thought the Federal Reserve has a role to play. And so, we started looking at the party affiliation of the president who appointed the Fed Chair, who’s in power at a particular time, and whether that aligns with the current president or not. And when it does, whether it’s Democrat or Republican, that gives about a 1% advantage in growth to that incumbent president.

Now, why is that? I don’t think it’s because there’s any kind of explicit collusion, though occasionally you’ll see, as in the famous case of Lyndon Johnson flying down, taking the Fed chair down and almost literally twisting his arm on his ranch, it’s more… I think, these are subtle differences and even Fed chairs really trying to walk down the middle are swayed a bit by their own political views and I think that’s what we’re seeing. Just a coincidence of common values. I have no suspicion that there’s any kind of direct… Collusion is too hard word. Coordination.

Matt Grossmann: You differ from the prior studies in that you are using comparisons of adjacent presidential terms. Without getting too far into the weeds, tell us about that intuition behind that and how it differs from how people usually…

Joe Stone: Yeah, I’ll do my best to take what we think is a sort of quasi-experimental design in as simple a way as possible. What is normally done, or what did Blinder and Watson do and others too, you take a binary variable that says it’s one if its Democrats, it’s zero if it’s Republicans or vice versa. And when that party’s in power, it gets a one. But those terms are often separated by decades. You might have someone in the ’50s, ’60s compared with someone in the ’90s or the 2000s.

The general environment is radically different across some of those terms and one can try to control for all those differences, but you can’t control for everything. We had the intuition, partly spurred by, I think, sort of new empirical methodologies emerging in the social sciences, of trying to look at changes of where… First of all, adjacent terms so that they’d be close together, assuming that adjacent terms are more alike than terms that are decades apart, and we did some tests to see whether that was true and it certainly did seem to be true. That’s the key difference, those adjacent terms.

A second difference is that we look at changes in the party affiliation of the president. When it goes from Democrat to Republican, we measure, okay, what effect does that have in adjacent terms? Then, if it goes from Republican to Democrat, what is the effect of that? And we assume that those effects are equal and opposite. We’re trying to get, what happens when there’s a party switch? And theoretically, they should be asymmetric opposites. But if you go from Democrat to Republican, that should be the negative of if you go from Republican to Democrat.

And so, using that technique of those party switches in adjacent terms, we get a reduction of about 50% in the standard estimate of a 2% differential between Democrats and Republicans. Did I confuse you enough?

Matt Grossmann: No, I got it. I got it. Another way of putting that is that periods of more Democratic presidents in a row might’ve just had different levels of economic growth than periods where the Republicans were in charge and that’s what the difference between your model and others-

Joe Stone: Yes. Thank you. You can explain my next paper.

Matt Grossmann: Your data goes up to 2013, so we’ve had two administrations since then that have been followed by election, and we have another one obviously happening now. I just wanted to get your qualitative take on how well those match the model. In 2016, I think we had a somewhat similar circumstance, where the Democratic administration had stimulated, but by the time of the election, growth was slower and Republicans won.

Then, Republicans came in and we had a strange economic term because, by your model, the Republicans didn’t wait until the end, but it was really a COVID-induced recession and then stimulus following it. How well did the most recent shifts follow what you think usually happens?

Joe Stone: I hesitate to revert to your field. Politics is a multidimensional issue and each of those elections had many issues and the economic part played a role, certainly, particularly the COVID recession in 2020. Certainly that played a big role, not to mention Trump’s own demeanor during that period.

Matt Grossmann: What about the economics? Did it seem like Obama was a fairly typical Democrat in what he was doing and its potential effect on the economy, but Trump may not have been or do you think it also matched the model?

Joe Stone: Obama had… The first half of his term fit the model relatively well in that there was an expansion and then tapered off in the second half. Trump’s term was a bit turned upside down, however. The first half of his term was better than the second half of his term, which was easily explained by COVID. I think events… Back to, it’s a multidimensional question and COVID just turned 2020 upside down.

Matt Grossmann: It seems like the Biden administration might be consistent with the pattern. They came in with fairly large stimulus and even might be kind of a lesson in not learning the previous lessons, if you think that the stimulus was so high that it actually contributed to the inflation pattern later. Is that the story that you’d buy?

Joe Stone: I do. And if you’re interested, I can add a little background to that, in that Biden’s domestic program and fiscal expansion for various things certainly was a very large fiscal expansion and it came on top of a substantial amount of spending by the Trump administration in the latter half of his term. But by the time the second big spending bills were passed in the Trump administration, the economy was already recovering from a very short but very deep recession.

In the very deep recession, earlier in Trump’s term, the Fed, in my opinion, sort of panicked. They saw the jump in unemployment and the dramatic drop in the economy and they said, we’re not going to do 2008 all over again. They really encouraged the banks to loan to be much easier in their loan policy. They eliminated reserve requirements. The banks didn’t have to have any reserves other than a little bit of technical oversight by the Fed and, in fact, the money supply just exploded.

It’s unfair to say that the inflation that really began to take hold in the Biden administration is primarily his fault. It clearly is not. Some of it is, because when he came in, I think it was not that difficult to see what was coming and the spending came anyway, and it just made things worse. That’s where we are now. Now, things are beginning to come around, but it’s a slow process and the election’s coming.

Matt Grossmann: Biden now has an unaligned Federal Reserve President, and I think it’s fair to say that for elections, at least, he would want some pre-election rate cuts that the Fed is not providing yet. Does that also match your model and what are the implications of that?

Joe Stone: Yes, it does, but I would add, quick-

Matt Grossmann: … implications of that?

Joe Stone: Yes, it does. But I would add quickly, first of all, it’s a little late. The effects of money on the economy are much slower than the effects of fiscal policy. So there’s not really, even if the Fed fairly quickly or even in the recent months had decided to ease up on interest rates, it would be too late to really show up much in the economy. It would show up in the stock market, but not the economy. And in addition, the scare of it, just as the Fed panicked a bit over that really deep recession that turned out to be very short, I think they panicked a bit over those really high inflation rates that you had to go back to the eighties to see inflation rates that high.

And so, they panicked… I shouldn’t say panicked, because their actions weren’t that dramatic, but they began to squeeze interest rates up. And I think that danger to the economy as a whole, and here I’m just speculating. I don’t think a Democratic, even if the Fed chair and the Democratic and the Democratic President Biden, even if they were aligned, I don’t think the behavior would’ve been that much different, just because the threat to the overall economy from the inflation binge we were on was so great.

Matt Grossmann: I’m also interviewing William Clark for this episode, and he sees more of a role for monetary policy than fiscal policy by comparison to you. But one of the mechanisms might be consistent. They don’t really look at alignment. They just say, overall the Fed during Democratic presidents is basically more sensitive to inflation, and during Republican presidencies is more sensitive to employment, and that the result of that has been that the monetary policy is less expansive as we get closer to elections under Democrats. So do you think that mechanism could be part of your story, or what do you think?

Joe Stone: Oh, yes. No, I’ve seen work along those lines. And it’s hard to separate out whether the Fed is responding to which party in power or they’re responding to the economic environment that they’re trying to deal with. And that just happens to differ by the party differences. But that result has been out there in one way or another for some time. And it would be useful to really go after it and understand why. Is it just the partisan alignment or are there a standard economic response to the environment would explain that?

Matt Grossmann: So your evidence might help to explain a question that political scientists have had, which is why Democratic presidents haven’t consistently done better, given that they do have a better economic performance overall. You have two explanations for that. One might be consistent with one that political scientists have given in the past, which is that while the Republican growth appears closer to the election and the other piece just that maybe we’ve overestimated that Democratic advantage. Do you think that explains why?

Joe Stone: Yes, both of those things I agree with. In addition, you add on the Democratic spending early in the term nudges up inflation in the second half of the term. So you get a little bit eaten away.

Matt Grossmann: So what should we be watching for this cycle if we are looking for the effects of the lower economic growth for Biden on the election itself, or if there is a party transition, what we should expect going forward?

Joe Stone: Going into the election, we’re getting so close that unless there’s a tank in the stock market, which is one of the factors in the Ray Farris style election prediction models, unless there’s a big tank in the stock market that triggers very rapid events, I think the economy is close to being baked in at this point into the election. I think we’ll be looking at… There’s a lot of inertia to inflation and I doubt that there’ll be big changes. The prices are stuck up there. And as we learned in the seventies and eighties, it’s really hard to pull inflation back down once it gets entrenched. As I said earlier, that’s not to say that of course the Republicans are going want to pin as much of blame as possible for that on Biden, but it was only part of the issue. Part of the issue, the blame is easily spread around from Trump to the Fed and other factors beyond policy like the supply chain problems.

Matt Grossmann: And is there anything we didn’t get to that you wanted to include?

Joe Stone: Well, I would probably add, if I were interested in this general area and were a young, ambitious graduate student or assistant professor, I would be looking at these patterns and saying, “No one’s really nailed down the mechanisms that link these.” The facts are beginning to emerge reasonably well, but the mechanisms are much harder to get at. And someone who could really crack open what the mechanisms are for why this is happening both on the political side and on the economic side, I think that’s a very right feel for some bright, ambitious researcher.

Matt Grossmann: Okay. I’ll ask one more question. One thing that seems, or that liberals would say is inconsistent with this model is that Republican claims to be fiscally prudent don’t actually hold up. If anything, they spend money and add in tax cuts. And so, you see what should be a big fiscal policy impact during Republican presidencies too. So is it possible that these models are about an older style Republican Party than we have now?

Joe Stone: Yes, the short answer is yes. Somewhat longer is that in recent decades, all of send and fallen short of the glory. Maybe for different reasons, but the end result on the fiscal side are pretty similar.

Matt Grossmann: So Democratic economic performance is not well suited to help and elections, perhaps reflecting their own fiscal decisions. But now let’s turn to my second interview with William Clark who says the role of the Federal Reserve is very important. Central banks are to the right of presidents and they especially fear inflation under Democrats.

So tell us about the main findings in your paper on partisan bias in monetary policy.

William Clark: Well, empirically we find that six things. We find that the Fed increases interest rates before elections when Democrats are in office, lowers interest rates when Republicans are vying for re-election. Third, we find that the Fed reacts to inflationary pressures when Democrats are in office, especially when elections are near. But the Fed does not react to inflationary pressures when Republicans are facing re-election. And somewhat relatedly or conversely, our fifth finding is that the Fed is insensitive to the output gap when Democratic presidents are standing for re-election. But the Fed is more sensitive to the difference between realized and potential GDP at the end of Republican’s terms.

So effectively, if you think in terms of the Humphrey Hawkins dual mandate of the Fed, the evidence seems to be that in the pre-electoral periods, the Fed is responding to employment shortfalls when the Republicans are in office but responding to missing its inflation target when Democrats are in office. So we conclude that the Fed is a conditional inflation hawk. It acts like we expect a traditional conservative independent central bank to act only when Democrats are vying for re-election.

But more important than these empirical results, or perhaps equally important, is that we argue that these observations can’t be explained by any existing theories, including my previous work on the effect of central bank independence on political business. But we do argue that adding a single, we believe entirely plausible assumption produces a new model that can explain all six of these facts.

This new model looks something like this. The Fed is more hawkish on inflation than all elected officials, and Republicans tend to be more hawkish than Democrats. None of those assumptions are new to the literature and they’re baked into most partisan models of the political economy and the very justification for central bank independence in the first place.

The perhaps more controversial claim is we say that the Fed is only partially independent, so that policy is a weighted average of the president’s preferred policy and the Fed’s ideal policy. This is based on work by my former colleague, Rob Francis, at the University of Michigan, who describes the relationship between the government and the central bank as a car with two drivers at the wheel. And so, the idea is that all bureaucracies that claim to be independent are actually only partially independent.

Combined, these suggest that policy will be closer to the Fed’s mandate to pursue price stability when Republicans are office than when Democrats are. And if that’s true, and this is really the only new part of our model, if the Fed cares about future inflation as much as it cares about current inflation, it will think about the electoral consequences of its current behavior.

And that means it faces an intertemporal trade-off when Republicans are in office because our claim is that all governments of all stripes will apply some pressure on the Fed to try to loosen policy in the pre-electoral period. When the Democrats do that, the Fed can, without conflict, deny the electioneering behavior of or counteract the electioneering behavior of a Democratic incumbent because doing so will only elect a Republican president whose policies they will find it easier to deal with.

But when Republicans stand for reelection, the Fed faces an intertemporal trade-off. They can compromise a little bit now with an electioneering Republican incumbent, or they could put the brakes on like a good independent central bank should do and wind up having to compromise a lot with the Democratic government that replaces that Republican. So really, the only difference in our model is the idea that the Fed thinks about the future consequences of current policy. Ironically, all of the rational models and the standard models in macroeconomics assume that the Fed is only thinking about current outcomes and therefore acts in a much more independent way than our model suggests they should.

Matt Grossmann: So let’s walk through these findings a little bit more. First, you’re finding that the Fed raises interest rates as elections approach under Democratic presidents relative to Republican presidents. How large is that advantage and how did you distinguish your interpretation from others?

William Clark: We estimate, even on our most conservative estimates, we find that the Fed tends to raise interest rates by 0.025%, or in the parlance of monetary policy is called 2.5 basis points a quarter during Democratic administrations. And it lowers interest rates by about the same amount during Republican administrations. So that makes a difference of about five basis points between Democratic and Republican governments.

Matt Grossmann: And that’s just in the last year or two or what is the pre-election?

William Clark: No, that’s quarterly on average across a four-year term. And in subsequent work joint with my A&M colleague, Paul Kelstad, and once again, I should note my co-author, Vincent Aurobundac at the University of Montreal is involved in both of these papers. And this new work admittedly hasn’t gone through the peer review process, so you might take it with a grain of salt, but we show that a one standard deviation shock in the Fed funds rate, which amounts to about four basis points. So that’s more conservative than what we predict they should be doing quarter by quarter. This would result a year later in a half percentage point change in presidential approval. And so, if the Fed pursues this policy, as we find they do over a four-year term, that lead to about a six percentage point swing in presidential approval, which is quite substantial given that in our period of observation, seven of 16 elections had margins of smaller than that.

Matt Grossmann: And you also find that the Fed is more responsive to inflation under Democratic presidents. So is that what explains the difference or how much of a role is that playing?

William Clark: Yeah. No, I mean that all pours into this 0.025 difference. Yeah, we just estimate it two different ways. We look at the Fed funds rate and the effect of the election on it, and then we estimate what’s called a Taylor rule, which is a standard empirical model of measuring Fed’s reaction to both to its dual mandate, to changes in output and changes in the price level. And we show that the Fed in pre-electoral periods, the Fed is responsive to the price level for Democrats, so it’s fighting inflation under Democrats.

So it’s fighting inflation under Democrats, but it’s actually pursuing its employment mandate under Republicans in the pre-electoral period.

Matt Grossmann: So you also address an alternative interpretation of your result or previous results that you call the rational partisan cycle. So explain what that distinction is and how you distinguish between them, and, I guess, is one way of putting it that the question is whether it just makes sense for the Fed to be more concerned with inflation under democratic presidents.

William Clark: There’s a lot to unpack there. So what we’re calling the rational partisan model is a model developed by the late great political economist Alberto Alesina and his co-author, Howard Rosenthal. And their argument, they were trying to solve the puzzle of how the Fed could systematically fool markets to bring about expansions and output. This is all in response to the kind of Lucas/Sargent critique of Keynesian economic theory in the 1970s, right? So in their model, what they do is they assume that employers write sticky wage contracts with workers that reflect inflationary expectations. But in pre-electoral periods, those wage contracts are written under uncertainty because if it’s the case that Democrats run more inflationary policies than Republicans do, which is standard in the partisan literature on macroeconomics. Then, let’s say, there’s a 50% chance that the Democrats win the next election and a 50% chance that the Republicans win the next election.

It’s not always the case, but just for simple math, let’s assume that’s the case. And let’s imagine that the markets assume that the inflation rate is going to be 4% under Democrats and 2% under Republicans. Then, in those wage contracts, the inflationary expectation is a 3% inflationary. If the election happens and the Democrats win and they implement their 4% inflation policy, then that’s going to lead, according to the so-called inflation expectations enhanced Phillips curve, that’s going to lead to an increase in real output. In the first half of the Democrats electoral term, and conversely, if the Republicans get elected, they’ll implement the 2% policy, which, given the inflationary expectations, pre-election of 3% would actually lead to a contractionary policy and induce a recession. And they argue that this explains why they’re finding that Democrats tend to have macroeconomic expansions in the first half of their term and Republicans contractions in the first half of their term.

But these die down so that there’s no partisan difference in the second half of the term because the wage contracts that are written after the election don’t suffer from this same uncertainty problem, right? So it becomes impossible for the Fed to surprise markets in the second half of the term because everybody knows who the government is. That’s interesting and a very important attempt to sort of bring this rational expectations’ framework into the study of political business cycles, and it was hugely influential work. However, it’s contrary to the results in our paper because our paper finds exactly what they predict for the first half of the term, which is Republican recessions and Democratic macroeconomic expansions. But in the second half of the term, we find the opposite. We find Republican expansions and Democratic recessions, and that’s only consistent with our argument that the Fed is responding differentially in the second half of the term to Republican and Democratic incumbents.

Matt Grossmann: So your data in the paper runs through 2008. Since that time, Democrats unexpectedly lost the 2016 election at a time of limited growth and low but rising interest rates. And then Republicans lost in 2020, obviously, following a large temporary COVID recession, but had very low rates at the time of the election. So how well do those recent circumstances qualitatively match your model?

William Clark: I am very cautious here. I really view our paper in some ways as a historical paper that covers the period roughly from 1950 to 2008. Because if you look at what happened in the run-up to the 2008 election, the way I would characterize it is the Federal Reserve basically broke the handle on the Fed funds rate. They lowered the Fed funds rate so low in the run-up to that election, trying to keep the presidency in the hands of the Republican Party. Our model would interpret that result as being; that the Fed funds rate was essentially; we entered this period of zero real interest rates, which is a bizarre macroeconomic phenomenon that we thought couldn’t happen and some people say will never happen again. So that period was a historically very strange period. We had the Fed experimenting with new instruments like quantitative easing, and it’s very hard to match up the data generation process from the period before 2008 and the subsequent period.

So I’m somewhat reluctant to claim that our model is doing a great job in more a recent period. Although, currently, the Feds continued tightening it, much to the surprise of markets that have been trying to predict the Fed funds rate. The Fed fund rate futures market pretty strongly suggested that there’d be more serious cuts to the Fed fund rate that happened this year than actually occurred. So you could argue that the Fed is pursuing a tighter than usual monetary policy under the Biden Administration, and that certainly fits our model. Fits our model with the important caveat that another confounding effect of this is the election of Donald Trump, and a changing of the political landscape, and the changing of what it means to be Republican and means to be a Democrat. And so, I mean, I’ve really spent a lot of time since Trump was being elected trying to figure out what the hell has happened, as has the rest of the political science discipline. And I can’t say that I’ve cornered that market just yet.

Matt Grossmann: So before we get to this year, you don’t want to take any more credit for 2016? It seems like it’s a good example where they really started raising rates. Yes, from a very low bar, but during the run-up to the election, out of fear of inflation.

William Clark: Yeah, that’s true. And gosh, I’m forgetting the person’s name, but one of the things that was amazing is that the former president of the New York Fed actually came out and wrote a newspaper editorial, I think in the Wall Street Journal, suggesting that the Fed should actually consider the electoral consequences of its behavior in trying to prevent Trump’s re-election. Which was, I never thought I would be able to get anybody associated with the Fed to say anything like the thought process that we’re claiming is going on in the minds of Fed policymakers. So that was quite astonishing to hear that.

Matt Grossmann: So let’s talk about this year. So obviously, Biden would be hoping for some pre-election rate cuts, wanted them, but before now as well. So where are we in the world of your model? What should we be expecting compared to what’s happening?

William Clark: I think, to be honest with you, I think we should be expecting exactly what is happening, which is that the Fed, well, again, if the Fed continues to think of the Democratic Party as the less hawkish party on inflation, then it would be perfectly reasonable for the Fed to be pursuing a tighter monetary policy than Biden would want, and that markets would’ve predicted in the absence of our model. And I hasten the ad that none of our argument is a suggestion that there’s some kind of conspiratorial intrinsic preference for Republicans over Democrats. It’s merely that if we believe the Fed is pursuing its low inflation mandate, which has been the de facto policy mandate since Volcker became the Fed Chairman in 1979.

This is just the way for the Fed to hit its target better: if the Republicans, if they have to compromise less with Republican presidents then Democratic presidents, they can produce, in their minds, better monetary policy with Republicans in office than with Democrats in office. So this isn’t any kind of deviation or what we might call a agency creep. This isn’t them pursuing a different goal. This is actually allowing them to pursue the goal that they see as their main goal, which is to fight inflation.

Matt Grossmann: And there does seem to be evidence for that in the Biden Administration that, after a very long period of low inflation, we finally got a high inflation period after COVID, but partially in response to a lot of continued stimulus at the end of the Trump Administration, beginning of the Biden Administration.

William Clark: That’s right.

Matt Grossmann: Would you agree with that?

William Clark: Yes, I would agree with that.

Matt Grossmann: Okay. So, I mean, let’s talk about that for a second. How do you think the Fed should be thinking about it now given that that part of the model seems still correct, the Democratic policies are a higher inflation risk? On the other hand, Republicans under Trump and even before that have not necessarily displayed fiscal prudence. Everybody seems to want to stimulate when they come in, just with different tools. So how should they be thinking about the difference between the left and the right?

William Clark: That’s a good point. I mean, to be honest with you, the real anomaly here is trying to explain why the Fed would hold on to this traditional belief that the Democrats are less fiscally responsible than Republicans are. Because if you look at the post-World War II period, it’s pretty hard to justify that. Republicans and Democrats spend their money on different things, but there’s not a lot of support for the idea that the Republican Party has ever been particularly fiscally responsible. And in fact, my earlier work on political business cycles made the argument that it’s not just in the US that in all OECD countries, there’s not a dime’s worth of difference between parties to the left and parties to the right in terms of fiscal and monetary policy. So the real puzzle is, my criticism of my paper is: why would the Fed ever have this crazy idea in the first place?

The interesting thing, though, is 90% of the literature assumes that this is exactly the way the Fed should be thinking about it. Add to that the fact that the Republican Party has been under increasing sway of a radical populist element, and populists tend to, in macroeconomic policy terms, favor consumption over investment, favor fiscal laxity and economic growth over price stability. So there’s nothing about the recent changes in the partisan landscape in the United States that would lend new credence to the old assumptions that the Republicans are fiscally conservative. So perhaps if the Fed admitted the reality that there’s not much difference between the two parties, then the Fed would cease placing its thumb on the scale for Republicans. But as you said, the stylized facts don’t suggest they’ve done so.

Matt Grossmann: So what would the Fed say about your findings? Are we not giving them enough credit for adjusting in response to data? I mean, if the inflation data had turned out slightly differently over the last six months, then maybe we would’ve gotten rate cuts inconsistent with your model.

William Clark: Yeah. Around the time we published this paper, one of my graduate students at Michigan and her colleague at, a graduate student at Hertie School of Governance in Berlin, wrote a paper, and they showed that the Fed systematically makes these mistakes when it comes to formulating their own inflate… The Fed publishes inflationary expectations, and they show that the Fed systematically overestimates the expected inflation under Democrats and underestimates it according to Republicans. Now, we could go even one more step in the conspiratorial direction and say they do that deliberately because it actually helps them pull off inflationary surprises under Republicans. I don’t know if I want to go that far, but there is this interesting pattern, it seems.

What the Fed would say about it? I mean, I can tell you so far, they’ve said absolutely nothing about what we say, and I’ve presented this paper and summarized this paper, and discussed this paper openly at conferences where Fed officials were present, and I’ve never been able to get them to take the bait. I think the thing that they would find most objectionable to our argument is the idea that the Fed is partially independent. If the Fed were wholly independent, then the Fed could still have these mistaken beliefs about the differences between Republicans and Democrats, but it wouldn’t affect their behavior. They just continue to implement their preferred policy. But we believe the idea that the Fed is wholly independent flies in the face of just too much data and too much theory, right? Even if you abstract from monetary policy and you think of any independent bureaucracy, the leaders of independent bureaucracies like the Fed, tend to guard their independence jealously, and the biggest threat to their independence is that the legislative body that gave them that independence at time T might at T plus N take that independence away, right?

So the Fed theoretically has to worry about the House of Representatives tearing up and in conjunction with the Senate of course, and the presidency tearing up the Federal Reserve Act. Or they could worry about the President and the Treasury undermining the 1951 Treasury Fed Accord that gave the Fed its operational independence. And because they don’t want that to happen, they have to trim their sales. When the preferences of their political principles change, they have to move the policy a little bit in the direction of the new principle or else they worry about this threat of losing the independence that they have.

Matt Grossmann: So place the American example in international context for us, it seems like your theory should apply pretty widely to left and right governments and supposedly independent central banks. Is the American Fed reacting similarly to elsewhere?

William Clark: Yeah, around the time we wrote this paper, we also wrote a comparative paper that used panel data from cross-sectional time series data from all OECD countries prior to the creation of the problem is in many OECD countries are part of the economic and monetary union in Europe, which means they all have the same central bank in Frankfurt. So that makes it difficult to compare the US with post-EMU Europe. But if you look at pre-EMU Europe, as we did in a paper that’s in an edited volume published in Germany, like I said, around the time this paper came out, we found roughly similar patterns in other countries where banks were independent. We saw them pursuing this preference for right wing parties. I should say that the idea for this paper came out of a conference I attended at the finance ministry in Berlin probably more than 10 years ago now, and my colleague and co-author, Lawrence Prose, at UC San Diego presented a paper where he argued that financial crises are more likely to happen under right-wing governments than under left-wing governments.

But he didn’t have an explanation for why that was the case. He just showed that empirical regularity. This was of course in the aftermath of the financial crisis in 2007, and I offered a off-the-cuff explanation for his finding at that conference. Quoting William McChesney Martin, who was really the architect, or at least the conservator of Central Bank independence from the Treasury Fed Accord up until Nixon fired him in the early 1970s. McChesney Martin said that the Fed’s job was to take the punch bowl away when the party gets out of control. And it occurred to me that the Fed might be less quick to do that when their allies are in office than when people who they have to compromise a lot with are in office. And that’s what led me to come back and immediately look at the Fed funds rate in this way.

So yeah, I mean, if Central Bank independence is this tendency to take the punch ball away, and if central banks are partially independent, then we would expect everywhere there is Central bank independence, we’d observe a similar pattern and Admittedly modest attempts to look at that empirically. We haven’t found any evidence to the contrary.

Matt Grossmann: So without getting too far into the scholarly history, bring us up to date of how this evolved from a set of theories developed around Nixon and then Reagan about the political business cycle and where we are today and what we know about that.

William Clark: Okay. Yeah, so I mean, prior to the work we’re talking about, well, really prior to the Alicina and Rosenthal attempt to build this all on rational expectations, the traditional literature fell into two camps. One said basically that parties at both types and act like drunken sailors on the eve of elections, right? They spend everywhere they can to try to stimulate the economy. The basic assumption is that even if parties have partisan differences, they realize that they can’t act on those differences when they’re not an office. So in the pre-electoral period, they put all of their ideological goals aside and just try to get reelected. If voters are as foolable as those traditional, we call them adaptive expectations models assume then voters, again, since according to standard theory and political science, since a voter can’t count on being the deciding vote in an election, the expected value of voting is actually negative.

And so voters shouldn’t be voting for instrumental reasons. It’s easy to believe, therefore, that they would vote on basically using a very rough, less than rational rule of thumb that says, “If the economy’s doing well on the eve of the election, vote for the incumbent. If the economy’s doing poorly, vote for the opponent.” This runs afoul of the sort of party ID literature popular down the street from you in Ann Arbor, right? That says people have these are socialized into supporting parties and they support the parties, the party that their parents supported, and they do that their whole life. Well, it’s possible that you could have half of the, well, 40% of the electorate be strong Republican parties, 40% of the Democrats electorate be strong Democrats and this middle group, if those groups were the same size. If the middle group was reacting to the economy, then incumbents would have a strong incentive to pull off inflationary macroeconomic expansions in the pre-electoral period.

So the traditional political business cycle literature, say, associated with Ed Tufte would say that we get cycles in macroeconomic expansions where you get booms before elections and busts afterwards. The competing explanation for that argument is this so-called [inaudible 00:52:52] political business cycle, which is a partisan cycle, which says Republicans are supported by largely above average income earners, people who derive their wealth from financial assets and Democrats get their support from people who derive their income from their own labor. The so-called partisan political business cycle would predict therefore that Democrats run systematically more expansionary fiscal and monetary policies, and Republicans pursue systematically tighter monetary and fiscal policies. All of those literatures, not to sound immodest, but prior to my work, all of those literatures assumed that political actors controlled monetary and fiscal policy. And my early work said, “Well, what if the Fed controls… What if independent central banks control monetary policy, how does that change those traditional political business cycles?

And in addition, how do changes in the global economy having to do with international capital mobility reduce the room for maneuver of domestic political policymakers?” And the results of my early work suggested that once you add those institutional nuances on who actually controls monetary and fiscal policy, there’s more evidence for the electoralist argument associated with Ed Tufte than there is for the traditional partisan cycle, all of which leads to different predictions than the paper we’re discussing now, which adds a new gloss to it when the Fed starts to think about future economic policy. There quite frankly hasn’t been a lot of work done since I published this work 10 years ago and since I published that book 20 years ago on political business cycles. It’s in some sense become a little bit out of fashion way to think about the link between electoral politics and macroeconomic outcomes. But as you suggest, recent events suggest that maybe it may be due for a bit of a comeback.

Matt Grossmann: So what should we be watching for this cycle and anything we didn’t get to that you want to do include?

William Clark: No, as I said, I continue to predict that the Fed will continue to… I’ll put it this way, if the Fed continues to pursue this low inflation, this anti-inflation policy, and leading to fewer reductions in the Fed funds rate than many market actors anticipated, that would be consistent with the idea that they continued to worry about expansionary policies in the Democratic Party. Like I said, I think there’s lots of reasons why they should probably unhinge themselves from those old beliefs, but they seem to be persistent.

Matt Grossmann: There’s a lot more to learn. The Science of Politics is available bi-weekly from the Niskanen Center. I’m your host, Matt Grossmann. If you like this discussion, here are the episodes you should check out next, all linked on our website. Will a Good Economy Save Trump? Inflation Hurts Presidents, and it’s not the Media’s Fault. How to Change American’s Views of Inequality Teaching in TV. Why Rising Inequality Doesn’t Stimulate Political Action and Interpreting the Early Results of the 2020 Election? Thanks to Joe Stone and William Clark for joining me. Please check out Presidential Party Affiliation and Electoral Cycles in the US economy and Independent but not Indifferent, Partisan Bias and Monetary Policy at the Fed. And then listen in next time.