Not Another Politics Podcast – Episode 40

Infrastructure. It’s one of the hottest topics in politics today. But what does the research say about the effects and politics of infrastructure investment?

Political scientist Jon Rogowski from the University of Chicago has a surprising paper that shows the long-term economic outcomes of post office developments in the United States. But it also gives us a lot to think about when it comes to who benefits, misses out, or even loses when infrastructure gets political.

Listen on Apple Podcasts or wherever you enjoy podcasts.

Transcript

Will Howell:                 

I'm Will Howell.

Wiola Dziuda:              

I'm Wiola Dziuda.

Anthony Fowler:          

I'm Anthony Fowler, and this is Not Another Politics Podcast.

Will Howell:                 

All right, guys. All the talk right now in DC... that's not true, not all the talk. But a big chunk of the talk in DC right now is about the possibility of an infrastructure deal being brokered between the Democrats and the Republicans.

Reporter:                     

Every day, Americans drive over roads and bridges that are in need of repair. Every day, Americans commute on transit systems that are outdated. Americans have water coming out of their tap that isn't clean, and they use internet that is not high speed.

Anthony Fowler:          

It's a big conversation. I don't know about you guys, but I feel myself becoming desensitized to talk about trillions of dollars. I go to the grocery store, and I think the bill might be three trillion dollars! I'm not sure... I don't even know what that means.

Will Howell:                 

It's a lot, but that's all you know. But how are we supposed to think about the stakes of this? That somehow, this is meant to not just fix bridges and roads, it's meant to put communities on a more solid footing so that they can make their way forward, not just in this moment now, of economic precarity, but for decades to come? Wiola, you talked to somebody who has a paper that investigated different kind of investment, and it certainly didn't involve trillions of dollars, but it was a significant form of investment in the 19th century.

Wiola Dziuda:              

The big question in this paper is, does investment infrastructure pay off? Does it lead to better economic development, better living conditions for the people, higher welfare of the society? And one very interesting aspect of this paper, other than it's about the 19th century development, is that they looked at postal offices. And postal offices are interesting because they are not only like roads and bridges in that they somehow facilitate a business, facilitate interactions between humans and so-on, but they also have less tangible value. They facilitate human interactions, they serve as this place where people can meet and form connections. So it's this softer side of infrastructure that they also provide. And if you listen to the discussions that we have in the media nowadays, you realize that that's what Biden is trying to do. He's trying to invest not only in roads and bridges, but he's also trying to invest in less tangible infrastructure, which he calls “human infrastructure.” The paper speaks a little bit to, what are the effects of this kind of investment? Can you start by telling me what is the big picture question that you are interested in when you wrote this paper?

Jon Rogowski:              

Yes. There were a couple of big picture ideas that we had in mind. One is that most people will receive mail most days. And so when we think about the postal service, when we think about the post office, we all have, or most of us have, some kind of connection with those experiences, and what might be the short and somewhat longer-term implications of these experiences that are nearly universal?

A second motivation was that when we tend to think about infrastructure, the things that come to mind for most of us tend to be things like roads and bridges and highways and canals and things like that, which are all very important forms of infrastructure. They are very well studied through several disciplines, but we think that the post office is also a form of infrastructure, but that operates through somewhat different mechanisms. And so we wanted to get a sense of whether we can trace out any evidence for how the post office in general might be associated with economic implications. And then in particular, whether we can suss out any evidence for whether those somewhat different mechanisms might be responsible for any observed relationships.

Wiola Dziuda:              

Tell me more about your analysis: what kind of data you collected, what kind of questions you asked, and how you try to answer them.

Jon Rogowski:              

There are two main sets of data that we use and two main kinds of analysis that we conduct with each of them. First is a new data set that we've collected that characterizes the distribution of post offices across countries from 1875 to 2007. The second is a more granular dataset that characterizes the distribution of post offices across counties in the United States from approximately 1846 through 1896. And so, we use these data to look at the association between the presence of post offices and a range of economic indicators.

So the first type of analyses that we do is we look at the increases in economic activity that immediately followed the change in the distribution of post offices. For instance, if a county got a bunch more post offices this year, we would look at economic indicators measured a couple of years down the road to see whether there is any systematic relationship between that increase in post offices and the economic output produced in those counties in the years immediately following.

The second analysis that we do some more longer-term analysis, where we look at the presence of post offices that was in countries or in counties at the end of the 19th century. And we look to see whether there is any association with the economic activity in those countries or in those counties at the end of the 20th century. So here, we're looking at a century-long association between the presence of post offices and whether those countries are doing better or worse today because of the number of post offices that were there a century ago.

Wiola Dziuda:              

What do you find? Is it good to have a post office across the street, as I do have, or-

Jon Rogowski:              

You chose well, because post offices are in fact associated with increased economic outcomes. We find that there is a positive and significant relationship between the number of post offices that your locality receives and the economic status of that same locality, whether it's in the short-term or in the longer-term.

Wiola Dziuda: 

Can you give me some examples of numbers that you found, just so that I get the sense of how big the effects are.

Jon Rogowski:              

Our results suggest that an increase in counties post offices at the close of the 19th century by 25%. So this is increasing the counties' post offices by a couple of post offices on average. That was associated with an increase of about $600 more in median income in the year 2000 about a century later. Now, at the level of individual households, $600 a year is not an astronomically large sum, but it is meaningful. Once you scale that up to all of the individual households that live in a given county, then you really are talking about real money. So an increase in post offices of about 25% in 1896 in the United States would have provided the average county with about 20 million more dollars in total income for the residents of that county.

And that's pretty economically substantial, probably well worth the investment that was initially provided back in 1896 and in the years prior to that, particularly because post offices are also revenue generating in the sense that people don't simply receive their mail from the government, but they also purchase things from the government so that their mail can go from one place to another. So a government's early investment, both political and material, for creating post offices seems to generate economically meaningful returns that seem to persist long after they were initially created.

Wiola Dziuda:              

Great. So theoretically, in what ways can the post office improve development, affect our lives?

Jon Rogowski:              

Yeah. There's a couple of different ways in which we argued and then studied the evidence for mechanisms through which the postal service might be associated with developmental outcomes. One is that they reduce transaction costs, they allow firms to conduct business, and they allow individuals to exchange money. So when we think about great American companies, it's hard not to think of companies like Montgomery Ward, for instance, which started in the 1870s as a mail-order business. They needed the post office in order for people to receive their catalogs and then send in money so that they could receive the products and the goods that they were ordering from the catalog.

Relatedly, the post office was associated with the rise of secured money orders so that people could, instead of sending cash through the mail... The mail in the old days; 1840s, 1850s, 1860s, in the United States was frequently intercepted and your letter containing a bunch of cash could be stolen, so the money would never get to where it needed to go. And so the post office began offering securable money orders. So you could purchase the money order, you could send it to friends or family members or whoever you wanted to.

So we looked at the volume of money orders that were exchanged through the mail in each of these places in the United States. We did find some evidence that places with more post offices exchanged more money orders. And as you exchanged more money orders, that would be consistent with the idea that there was more commercial and financial exchange in places that had more access to the post office.

A second mechanism would be that it facilitated the diffusion of information and ideas. Newspapers, for instance, would be sent through the mail. You could subscribe to your newspaper, it would show up in your mailbox, or you would pick it up from the mailbox that you have at the local post office, so it allowed lot of people to exchange both ideas and information. This has been documented through some recent work in a related context where they showed that the availability of post offices was also associated with increases for patenting rates. So there was some reason to suspect that the post office might be an engine through which entrepreneurialism could flourish.

A third potential mechanism might be through enhancing social capital. When people live across large distances and there isn't air travel yet, it's hard to see one another frequently. And so changing letters would be a way that you could keep in touch with friends and family members that had moved away to allow people to maintain those social connections, even when their social network became more geographically dispersed.

Finally, post offices themselves were important local meeting sites. So if you lived in a small community, maybe a couple of hundred people, there's a post office that opens on the main road through town. The post office would be (this would be in the era before free delivery to your house), You would instead go to the local post office in your town, you would pick up your mail, and you would see your friends and neighbors, which gave you a sense of local community. This would also allow you to share information, share local gossip, figure out what's happening around town, and this would be a way to facilitate greater cohesion within your community.

So we used some data that characterized the number of civic organizations that were in each county in the United States. This is some data that some researchers developed to create measures in the 1990s. And we did find a pretty strong relationship that suggested that places with more post offices at the turn of the 20th century had more civic organizations a century later. That we interpret as some suggestive evidence consistent with the role of social ties and social capital for facilitating the increased economic activity through the post office.

We wanted to also look at the evidence that would suggest the role of information and the diffusion of ideas in facilitating economic exchange. And so we got some data on the number of newspapers that were housed in each county, as well as their circulation numbers. We actually did not find very strong evidence there. Well, we did not really find that counties with more post offices tended to have more newspapers located there, or tended to have more newspapers being read by people that lived in those counties. So diffusion of ideas and information seemed to be the weakest mechanism of the three that we posited, but we did at least find some suggestive evidence consistent with the other two mechanisms that we discussed in the theoretical section of the article.

Wiola Dziuda:              

Let's think a little bit about, should we actually trust the $600? Is there an alternative explanation for what you're finding? And the obvious criticism would be, well, perhaps the causality goes the other way, perhaps in places where business was booming and that had, let's say, favorable geographical conditions that were located in places that actually could have experienced anywhere into development, those were the places where the government predicted business would need some extra help, or actually any return to extra help would be high. And that's why they placed postal offices. So perhaps postal offices help a little bit those businesses to develop, but the large part of the fact that you're seeing we would have seen anyway because those places were special to begin with.

Jon Rogowski:              

We tried to address this by including as many indicators of business activity and other kinds of economic activity in the late 19th century as we could find. So all of the different variables that I described earlier for the 19th century panel analysis, we tried to throw in separate independent variables for trying to isolate the effect, per se, of the post office. And we continue to find the very strong relationship between the locations of post offices in 1896 and economic activity in the year 2000, even when we account for a variety of these other indicators.

Now, that does not mean that we can totally rule out the possibility that there is not some sort of reverse causality that's happening here, or maybe in fact, the story is the other way around. We tried looking at some other factors that might be associated with this. We tried to look at whether our effects plausibly could be driven by things that would reflect economic activity. For instance, maybe it's increases in patenting rates, for instance, during the 19th century, that is generating more economic returns later. And even when we account for factors like that, we continue to find the same coefficient estimate for the association between the post office and economic activity.

But this remains a huge limitation of our analysis, is that we don't have control over the mechanism through which post offices were assigned. They certainly were not assigned in a random fashion. And because we don't have precise control over the assignment mechanism, we are somewhat limited in the ability to make strong causal inferences on the basis of these results. So we tried to leverage what we do know about the assignment mechanism to gain a little bit of additional control over this. So we used everyone's favorite tool, instrumental variables, to try to do this. From some earlier work that I've done, I found that there's a pretty strong and robust association between local partisan politics and the allocation of post offices to that county. And the political connection seems to be driven through the president's partisanship.

So if you live in a county that was hugely supportive of the president's party in the most recent presidential election, you are predicted to receive more post offices. So what we did was try to leverage the stat to use election returns and use those election returns as an instrument for predicting the allocation of post offices and, in association with that, looking to see whether that instrument is associated with longer-term economic consequences.

Wiola Dziuda:              

So just to keep it simple, basically you're telling me, ideally, you would like the post offices to be assigned at random, and then we see what's the effect of having a post office. But here you can say, "Well, maybe I cannot think whether people voted for the party of the president is random," but it shouldn't be necessarily dependent on some other things such as location and business development and so on. So in a sense, if I can say that this place got a post office because this place voted for the party of the president, then I can treat this as a random assignment. This happened at random. And I compare it to another place that's very similar, but they did not get the post office because they did not elect a representative associated with the party of the president.

Jon Rogowski:              

Yes, that's exactly right. Presidential elections in the late 19th century were very hotly contested, like presidential elections today. In fact, the political landscape and partisan landscape was very similar during the late 19th century to what it is today. The parties were very polarized, there were very competitive presidential elections at the national level. And so we want to leverage these voting patterns through county election returns to treat the allocation of post offices as a kind of experiment.

Wiola Dziuda:              

As being driven by the political outcomes and not necessarily the things that we think affect economic development.

Jon Rogowski:              

Yes.

Wiola Dziuda:              

Let me ask you the question that's been hanging over our entire conversation, what lessons can we draw from your work for the importance of Biden's infrastructure bill and for the possibility that that's going to affect future development of the country?

Jon Rogowski:              

There are a lot of things I like about this question. One of the things I particularly appreciate about it is that, when you write a paper that's about the distribution of post offices in the 19th century, it invites the “so what” question: why on earth should we possibly care about this today? But because the post office has been in the news in particular over the last year, and this year we're talking about infrastructure generally, where the Biden administration has made a big push to say that infrastructure is not just about improving highways and building new bridges and things like this, but there are a bunch of other ways in which we can improve the status of human infrastructure.

So I think that our article offers a few lessons, a few insights into these contemporary politics. One is that there tends to be more agreement, more bipartisan agreement, about the importance of infrastructure than there is for most other issues that are politically salient today. Now, of course, there's been a lot of Republican opposition to parts of the Biden administration's plan. However, no one says that infrastructure is not important. There's disagreement over what forms of infrastructure are important. And we think that generally speaking, the findings from this paper show one other way in which infrastructure is good. And that's why there is general agreement across party lines that yes, of course, infrastructure is good at something that lawmakers' constituents will appreciate and will benefit from not only in the short-term, but potentially in the longer-term as well.

A second insight I think this offers is that where things that post offices generate these communication networks, which are not visible in the ways that roads or networks of highways are visible, but they do connect people and they allow for these economic returns to occur downstream. And parts of what the Biden administration is doing is trying to place emphasis on other aspects of infrastructure that is not always visible at the surface level. So by allowing more people to have access to things like broadband infrastructure, that is basically increasing the density of the 21st century post office network.

By allowing people to build stronger social ties with people who don't live down the street or in their communities, they can communicate instead with people who live all over the world. They can access newspapers and other sources of information through the internet online just as the post office allowed people to do 100 or 150 years ago, and they can conduct various kinds of economic exchange. So it seems like the Biden administration's efforts to categorize things like broadband as essential for strengthening human infrastructure has some parallels to these more invisible forms of networks that we study in this paper.

Wiola Dziuda:              

We are convinced that Joe Biden is our fan. So please expect a phone call from the White House in the next few days. I want to go back to this idea that placement of post offices was political. Do you think this is happening nowadays too? Do you think that when the bill passes, who actually receives it, which counties the money flows to, will actually be worked out later on? And is it possible that, or should we expect that if the bill passes under Biden's administration, then the places that will benefit will tend to be more democratic and more closely affiliated with Democrats? Is it possible that that's actually part of Republican opposition, that yes, we like the infrastructure, but given that we know all those investments will be channeled in the direction of the supporters of Democrats, we would like to be the ones that actually pass the bill once we are in power and we can channel the money appropriately to our constituency?

Jon Rogowski:              

Yes, I think that is very likely the case. When you give presidential administration, or you give bureaucrats discretion, there is always the possibility, maybe even the likelihood, that that discretion will be exercised in ways that privileges those bureaucrats' own political and partisan interests. Now, I want to also emphasize that there are some limits to the extent to which presidents and presidential administrations might want to go in this direction, for a couple of reasons. One is that if presidents push too hard in trying to benefit their own co-partisan constituencies, and by denying those benefits to constituencies affiliated with the opposite party, they're just creating opportunities for the opposition party to criticize the incumbent president for politicizing or otherwise being unfair about how federal resources are allocated.

Two, I think the second reason to be somewhat circumspect about the extent to which this factor will dominate the ways in which the Biden administration makes decisions, is that presidents want the country to do well. Generally speaking, they want the country to do well. They want citizens to be better off when the next election comes around relative to when the president was initially elected, because that's only going to benefit their own reelection chances. And so presidents, Biden included, likely will not do things that are going to undermine their own reelection chances. Partisanship won't be the only factor that dominates how exactly these funds were allocated and where these projects are built.

Wiola Dziuda:              

Well, thank you, Jon Rogowski. That was great.

Jon Rogowski:              

Thank you, Wiola. This was really fun.

Will Howell:                 

Wiola, do you want to just summarize for us what the core empirical findings are of the paper?

Wiola Dziuda:              

Yeah. Basically, what the paper looks at is how does the number of post offices or density of post offices correlate with economic development five years later? In addition, he looks at how the density or the number of post offices in the 19th century correlates with the economic development in the year 2000. And overall, he finds positive correlation and the effects he finds are quite substantial. For example, he finds that a 20% increase in post offices per capita in 1900 is associated with a 3% increase in GDP in the year 2000. That strikes me as a relatively large effect, given that this is a hundred years after a particular investment was made. Can we say this is definitely causal, and especially, can we say this entire effect is causal? Absolutely not, but the association is there and it's very strong.

Will Howell:                 

Okay. In the paper, when they think about mechanisms and pay attention to what they call the reduction of transaction costs and economic activity, the creation of social capital; that is, people may be learning more because they get newspapers that they wouldn't previously get because there's a post office near then, these kinds of things. But accompanying the investment of a post office may be other kinds of infrastructures that are built up simultaneously. When you build a post office, you may also be building a road and at the same time building up a school that's near it.

Will Howell:                 

And it can be all but impossible to disentangle the separate influence of each of those different kinds of investments, particularly when they're happening all but simultaneously. There's an investment happening into a community for one reason or another. And it may be... it looks like in the data, it's not obviously just a straightforward product of past economic trends; they look at that. But it may well have to do with announcements of future businesses or where they plan to relocate or expectations about future kinds of economic growth that are not just a straight forward extension of past economic trends. And so, this is a real challenge.

Will Howell:                 

I think that you're exactly right, Wiola, that they are perfectly upfront about the challenges that they face in identification, but this is... when we think about the mechanism, are we sure that we're estimating the right thing? Is this about post offices as opposed to the other kinds of infrastructure investments that might be happening simultaneously?

Anthony Fowler:          

We've talked about the possibility of some kind of reverse causation, we've talked about there could be a bundled treatment. The third related concern is that there's just a confounder here, which is that there was a lot of patronage politics, there's politics involved in these decisions as to where do you put the post offices. Maybe you put more post offices in places where there are lots of political connections, and maybe those political connections are actually driving some of these economic developments for reasons other than post offices. So it could be that political connections are good for your region, but not because of post offices, per se, post offices are just a by-product of that. So all of these seem like potential concerns you would have of this paper in saying, “it's definitely the post offices, per se, that caused this economic development.”

Wiola Dziuda:              

And I think this one is particularly interesting because it has consequences for what we think about the current infrastructure view. I wish there was a way to get clarity on that. In the paper, the authors actually try to use these political connections to get more at the causality. They claim, "Well, if a place is politically connected, then it's more likely to have a post office just because it's politically connected."

So we can treat this arrival of a post office in that particular place as uncorrelated with other development-related characteristics of the place. So we can treat it as if it was just built at random in this place. But what you're saying is, well, even if that's true, other things are probably flowing because of this political connection. And how do we know that this is the post office that's driving the development, or all the other things that are flowing together with the political connections.

Will Howell:                 

Yeah. Jon’s just got a whole bunch of papers on precisely that, about the politics that informed the allocation of these. And I think that he introduces and offers up these kind of political variables as an instrument for their allocation with appropriate hesitance that, "Here's something that one might do. I'm going to do it for you." But then in the next sentence in the paper, he says, "But look, I also recognize that this suffers from precisely the kinds of challenges that you've just laid out." I think that one thing that is useful when you think about the literature on state building in the 19th century where this is helpful—so much of that literature is thinking about politics to the exclusion of the provision of actual public goods. It's all about spoils, it's all about political alliances, and it is about those things in important ways.

But in attending to those things, what this paper then comes in and shows, some communities made out pretty well. It actually yielded some good. And so to say that... arguments about pork barrel politics are to the exclusion of the possibility that the public goods might be attended to as well. And so there you are. There's a hopeful spin that comes out of this article.

Anthony Fowler:          

So if you were a dyed in the wool libertarian, you'd push back a little bit and you'd say, "Well, okay. Clearly there were benefits of this government spending." You'd want to ask, "What were the opportunity costs? What were the downsides?" You could have had lower taxes and maybe just the country as a whole would have invested more in other things. And maybe those other things would have turned out to benefit people equally, if not more. What is there to say about that? Well, how can we... Can we take the magnitudes of these estimates and compare them to other comparable investments to get a sense of whether this was a good bang for the buck?

Wiola Dziuda:              

I think you're absolutely right that this is something that's missing from the analysis. And I think it's particularly relevant here because I can envision how, if I put a post office in place A, I'm benefiting place A, but I'm actually harming place B. So it's not even about the cost of taxation that I needed to impose in order to build this post office, but I'm just simply shifting business from place B to place A because it's easier to do business in place A. So what the paper is not looking at is some general equilibrium effects.

And now, if you think about other papers that try to look at the question of what's the return-to-investment, there are some other papers that perhaps have better causality identifications that they are exploiting, and overall, papers do find that infrastructure pays off. But even in those papers, I think looking at the general equilibrium effect is just almost impossible because infrastructure investment is local and usually, taxation is more global, more federal level. And then, how do you capture the impact of that? I think here we are in a situation in which we can stick to our political positions no matter how many papers we read on this topic.

Anthony Fowler:          

Is that a virtue of the cross-country analysis that Jon has in the paper? Maybe there's some other things about the cross-country analysis that we don't like as much, but in the cross-country analysis that's... One country is deciding to invest money from its tax base, build more post offices. And we can look at, at the aggregate level for the whole country, is the whole country better off? Whereas when we're doing the county level analysis in the U.S. it could just be the case that we're transferring money from all taxpayers and helping some counties and hurting others, and we never really know on net what's the effect. So is that something to like about the cross-country analysis?

Wiola Dziuda:              

I think that's why they have cross-country analysis, even though I think we tend to shy away from this kind of analysis, because we know that there are plenty of other things going on. The entire postal infrastructure varies from country to country. The needs of the country vary from country to country. So yes, it's the virtue of this analysis, but we know that there's this entire variety of problems that comes together with cross-country analysis.

Will Howell:                 

Hey, if you're getting a lot out of the research that we discuss on this show, there's another University of Chicago Podcast Network show that you should check out. It's called Big Brains. Big Brains brings you the engaging stories behind the pioneering research and pivotal breakthroughs reshaping our world. Change how you see the world through research, and keep up with the latest academic thinking with Big Brains, part of the University of Chicago Podcast Network.

Wiola Dziuda:              

Okay. I think we all agree, and I think the authors agree, that we can't really argue that we’ve measured precisely and convincingly the impact of building a post office in the 19th century on the subsequent economic development. But I think, at least I would say, this paper is suggestive that investments in post offices were profitable for the surrounding communities. And we have body of evidence from other disciplines and from other papers that suggest, yes, investments have some returns. How large they are, that varies from investment to investment, but investments in infrastructure have returns. Thank God.

Let's run with this idea. How should we think about what's happening right now in the U.S.? The discussion about Biden's infrastructure beyond the fact that Republicans are sort of signaling that they might get on board with at least some portion of that but are still opposing a large infrastructure bill. Does the paper make us think about this entire political drama a little bit differently, and about the potential consequences of passing the infrastructure bill?

Will Howell:                 

Yeah. I think there are two things. Two things immediately come to mind for me. There is, of course, the longer-term implications of the investments that go beyond just the jobs associated with the building of the bridge. And in that sense, every legislator from every party ought to be clamoring for those kinds of investments. But there's this other piece, and this has been a theme in our... We've talked about this in a number of our shows, is that there's a big urban-rural divide. And you could imagine that certain kinds of investments to the extent that they lead to greater economic growth, that then could lead to changes particularly in rural environments, in the composition of the electoral composition of those counties and states.

A big part of the reason why Texas is turning purple is just something to do with old-time Texans changing their minds and saying, "You know what? Those liberals have something to say." It's because there's all kinds of population movement into this region, which is in part a function of economic growth. The same thing can be said for Georgia. And so you can imagine, like, "Be careful what you wish for, Wyoming," in that if you get a whole bunch of infrastructure investment, and this story is right, and that leads to 150,000 people moving in over the next 20-30 years, that what was once solidly red may not be solidly red going forward.

Anthony Fowler:          

We might be overstating how much people care about the party of their Senator. I think if you ask people in Wyoming, "Would you like to have lots of money invested in your state and lots of economic growth and lots of... and cheaper, better internet, and better roads, and better bridges, and sewers and so on, but you might have a Democratic Senator?" I think most of them would take that deal. I could be wrong about that. But I think...

Wiola Dziuda:              

But I don't think it's only about political representation in this narrow sense that, "Am I presented by a Democrat or by a Republican?" But I think it's also about, am I going to have all these coastal elites [crosstalk 00:34:28] Exactly. And $5 avocado toast or probably $5 chips for today? And I think we have a lot of examples of something like this happening in different communities that people are aware of that.

Anthony Fowler:          

Certainly, you've raised an interesting point, which is that a lot of infrastructure proposals have differential effects. They benefit certain kinds of people and certain kinds of industries more than others. The rancher in Wyoming is not really excited about better broadband internet. They're not going to benefit from that as much as the new tech startup that might want to move to Cheyenne.

Will Howell:                 

And so there's a political logic for Democrats to get behind big infrastructure investments in rural communities, because the returns will be such that suddenly a bunch of hipsters and tech types will want to move to Montana. They already are. And it doesn't take that many more to a Montana, to an Idaho, to a Wyoming to pick up a seat. I think what we're accustomed to thinking about when we think about infrastructure investments is that it's a mad scramble. Everybody stands to benefit, so we all want the infrastructure. And you want to get more for yourself. And every dollar you get for yourself that you keep from somebody else is necessarily a win. This paper suggests that, well, there might be longer-term dynamics that suggest otherwise.

Anthony Fowler:          

Do you know what the equivalent of the hipster was in the 19th century? The kind of person who moved into a new place because there's a shiny new post office!

Wiola Dziuda:              

But there's also this direct effect that I think Jon’s paper hints at. Jon mentions that post offices were more likely to be located in places that were politically aligned with the party in power in the federal government. And I think there's reason to believe that the same effects are going to be happening right now. That yes, we pass this bill that broadly describes how much we invest and in what we invest, but then the nitty-gritty details will be decided by agencies. And at the end of the day, it might be that the money will flow to more Democratic places. If I lived in a Republican state, and even if I knew that I personally would benefit from some better broadband and a better highway, I might want to wait until Republicans implement a similar infrastructure bill, because then I know I'll be the one on the receiving end of this bill and not the Democratic counties.

Anthony Fowler:          

So, do we want to make any leaps from this kind of paper and say something about the economic effects of the infrastructure plan? Should it go through?

Will Howell:                 

I'm going to say the naive thing. It's going to be naive twice over. It's going to be naive first because I'm going to interpret the findings that are offering this paper causally. And to say that in fact, it may well be that, maybe not the entirety of the effect, but a good portion of the effect, is attributed to these investments of infrastructure on long-term and short-term economic growth. And then second, it's going to be naive because I'm going to take Biden at his word. And what is he saying? He's saying that this investment is not a jobs program. He's saying this is about our long-term economic well-being, our long-term economic growth. This is about our competitiveness as a country vis-a-vis other countries. This is an investment in our kids.

And the findings in this paper lend some credence to those kinds of claims. This is not about an economy in decline, and we just need to get people back out getting some kind of work so that they don't have long gaps in their employment history. That's not the kind of argument that's being made right now on behalf of the infrastructure bill. The kind of argument that's being made right now is in keeping with the findings of this paper. This is about short, medium, and in fact long-term economic well-being of communities.

Wiola Dziuda:              

I will pick up on this long-term versus short-term. We tend to think that politicians don't focus on policies that are good long-term unless they are also good short-term. Politicians have a deadline. There's an election coming and they have to perform before the election. So Biden can talk all he wants about the long-term effects, but my guess is that the things that he puts and will put in his infrastructure bill are things that he thinks are going to have some immediate payoff so that people see they're working out. And then the question is, should we worry about that? We are spending, how much money again? And tell me it's trillions. 5 trillion—

Will Howell:                 

Trillions!

Wiola Dziuda: 

Exactly. We are spending the enormous amount of money that even cannot comprehend what it is. And I think we should worry about the priorities. Is this money going to go to investments that have only short-term effects, or also to those who have long-term effects? And I think in this respect, there's a little bit of hope that comes from Jon’s paper. I bet that when post offices were being built, no one argued, "Look, this is for your future in the year 2000, when your grand great, great, great grandchildren are going to be alive." I'm sure that it was sold or introduced because everyone thought there will be immediate benefits.

And what the papers shows is that, well, those immediate benefits somehow managed to translate into long-term benefits. It wasn't obvious that they would, because you could envision that new technology would come about and would completely outdo whatever the post offices did. But somehow there's a mechanism through which the development today grows and trickles over and results in development later on. And I think that makes me a bit more hopeful that a big infrastructure bill is not going to be just money wasted for short-term gain.

Will Howell:                 

In the paper, the talk is about one thing; post offices. And there's this notion that a post office as a post office. A feature of the debate right now is that, well, infrastructure means 74 different things. It's about—sure, it’s about roads and bridges, but it's also about broadband, and it's also about investments in human capital, and it's also about relieving debt for college, and on and on and on. And I think the possibility of these broad kinds of investments that you've articulated, Wiola, is tantalizing. But this question about, what are the individual returns from these different kinds of investments, lingers in the contemporary setting in ways that it was a little more... at least it seems like it's a little more straightforward when it came to the 19th century, where it's just about, "Shall we have a post office? Shall we not?"

Anthony Fowler:          

Well, yes. I think a thing that you think about when you read this paper is the fact that infrastructure is supposed to be for long-term growth. And so I think to the extent that we are pushed to think about that in the current infrastructure debate, I think that's a good thing. And so for some of these infrastructure proposals, I think there is a good argument to be made, and we should make that argument. For other infrastructure proposals, a lot of Democrats are lumping in lots of everything under the sun in infrastructure. Some of those things, there's not a clear argument about long-term economic growth. Some of them are just cash transfers.

I think that is a dangerous precedent to say that we're going to call all of our preferred spending projects infrastructure. So student loan cancellation, I don't see what the compelling argument is that this is going to have some long-term economic benefit. It just seems like a big cash transfer from taxpayers to recent college graduates. It's not a progressive proposal nor is it an investment in anything. So that part... That's one example. I think that's unproductive to lump that in with infrastructure discussion.

Wiola Dziuda:              

I think you're pointing out another interesting thing, that there are some investments that are in this bill, some items that are in this bill, that on the surface seem like overall beneficial. You might ignore the fact that you have to tax people in order to support those investments. You might say, "Well, we have income inequality, we have progressive taxation." Hopefully. That's not obvious, but let's assume. So, taxation is going to be fine, it's going to equalize people's chances. And it seems like certain investments are good, like debt relief or investments in childcare and so on.

Wiola Dziuda:              

But if you dig a little bit deeper, the distributional consequences of those investments are actually not as straightforward as one would hope. I think the same way, the introduction of post offices probably had distributional consequences that were not obvious to the people who advocated for that. At the same way, I think Biden's proposal have distributional consequences that are not as obvious as Biden would like them to be.

Wiola Dziuda:              

For example, one piece of data that I have is our own colleague, Ingrid Gardener. She has a paper on the rollout of broadband in Norway. It was semi-random so they have a nice identification strategy. And of course you would say the rollout of broadband, yes, you have to tax people to cover it, but let's put it aside. It should benefit everyone. And what they show is that actually, it harms unskilled workers. Companies are moving away from using those workers to using broadband to achieve certain goals and to have certain tasks performed. So there's a group of people that probably was a constituency of the politicians who are advocating the route of the broadband who actually lost from this infrastructure investment. And I think one should talk about this more.

Will Howell:                 

Yeah. It's super interesting. Now, yes, I've got a better connection to watch my movies, but I've lost my job by virtue of the rollout of broadband.

Anthony Fowler:          

And isn't that such a central story in American politics today, right?

Will Howell:                 

Yeah.

Anthony Fowler:          

You think of the debates we have about trade, the debates we have around immigration, we can implement some policy that on net, on average, benefits the country, but does harm certain groups and certain people, and they're going to be disaffected and they're going to be upset and their votes are going to be influenced by how they feel about that. And they're not going to be comforted by the fact that a bunch of liberal academics tell them, "Don't worry, this is, on average, economically efficient for the whole country. And tough luck for you."

And I guess a corollary is, even if we decided to go through with it, we don't have to be smug elitists who talk down to people. We can say, "Okay..." We don't have to say, "Just deal with it. This is good for us on average."

Will Howell:                 

There need to be compensatory programs.

Anthony Fowler:          

Right. We can say, "Okay, maybe there's something else we're going to do for unskilled workers who are harmed by this policy." I don't know. We could at least talk about that.

Wiola Dziuda:              

You know what I find myself thinking every time we get to the “what's your bottom line?” My bottom line is actually the opposite of what I thought at the beginning. I don't know whether that speaks negatively to how prepared I am, or how informative these discussions are for me.

Anthony Fowler:          

What's your bottom line, Wiola? What was your bottom line before, and what is it now?

Wiola Dziuda:              

My general inclination is, yes, we have to invest in infrastructure, especially because the infrastructure in the U.S. is really lacking. I came to this country having this image of a very well developed country, and I envisioned Manhattan to be shiny glass buildings. And I arrived and I saw Penn Station, and the A-Train from the JFK, and the rats on the streets. Personally, my personal politics is yes, we have to spend more money on infrastructure. And I like a lot of proposals that are in Biden's plans. But this entire discussion made me think that there's a lot of arguments to oppose certain provisions of this bill or the bill altogether. And I shouldn't have such a strong opinion without reading more carefully and thinking about the items that are in this bill more carefully. But overall, the paper, I think, contributes a little bit to our understanding of the importance of good infrastructure investment.

Will Howell:                 

Yeah. The paper provides some reassurance to those concerns. We don't have data from the paper that speak to the distributional consequences of these kinds of infrastructure investments, but we do see is, both of the short long and long term, it meaningfully contributes to growth on average. That seems like a good thing.

I guess as my bottom line, I'd reiterate a point that I offered earlier, which is that we're accustomed to thinking about pork barrel politics as just being about politicians scratching the backs of their political advocates. And even if there are politics, and there surely are politics that inform how infrastructure investments are being made, there are larger things at stake. And this paper productively points our attention to those larger things, which have to do with not just the jobs that are available today, but the kind of growth that is rolled out over decades.

Anthony Fowler:          

Those were so well put. I don't think we should end with my bottom line, because I'm not going to have as much to say or as eloquent things to say, but I agree with everything you said. I think this is an interesting paper. I think the concerns we discussed about the paper are genuine, and that it's hard to know how much we can attribute these estimates to directly the effects of post offices versus some of the other confounding factors that we discussed. But to the extent that we are in the middle of an infrastructure discussion, we should be talking about these long-term economic benefits as much as possible. And this paper does a good job of pointing us in that direction.

Will Howell:                 

Ladies and gentlemen, it's now time for another segment of Not Another News Roundup.

Wiola Dziuda:              

Yes. We talked a lot on this podcast about the fact that when you look up the data carefully, it seems that moderate candidates win in the U.S. politics. And it seems to go against our gut feeling. It seems like we turn on the news and it seems that it's the other way around, but here we have another data point, which is Eric Adams.

Reporter:                     

What should Democrats across the country take away from your victory?

Eric Adams:                  

Oh, we can't be so idealistic that we're not realistic. Cities are hurting all across America, and New York personifies that pain; the inequalities, the gun violence, the lack of really looking after everyday blue collar workers, I like to say.

Will Howell:                 

Yep. In a field with plenty of progressive liberals, the Democrats saw fit to put forward one of the more moderate candidates as their nominee in keeping with the papers that we've looked at. As you say, Wiola, there are electoral rewards associated with doing so.

Wiola Dziuda:              

Especially because they didn't have to bother. They didn't have to think about the general election. We should have seen more progressive candidates having an edge.

Will Howell:                 

When we think about, are there downsides associated with electing more extreme or more moderate mayors around the country? I don't know. I think it much depends on their ability to negotiate with their city councils. And if they're all liberal all the way across the board and progressive liberals or people on the far right, all of equal advance, better policy, what if that might be, then we don't have a problem.

Wiola Dziuda:              

But we know that they won't be because all the papers that we looked at tell us that they will be moderate.

Will Howell:                 

Thanks for listening to Not Another Politics Podcast.

Wiola Dziuda:              

Our show is a podcast from the Harris School of Public Policy, and it's produced by Matt Hodapp. Thanks for listening.