Jonathan Rodden on the Geography of Voting
Oct 22 2012

Jonathan Rodden, political science professor at Stanford and a senior fellow at the Hoover Institution speaks with EconTalk host Russ Roberts about the geography of voting. The main focus is on the tendency of urban voters around the world to vote for candidates on the left relative to suburban and rural voters. Rodden argues that this pattern is related to the geography of work and housing going back to the industrial revolution. He also discusses the implications of various voting systems such as winner-take-all vs. proportional representation, the electoral college and how political systems and voter preferences can produce unexpected outcomes.

Don Boudreaux on Public Choice
Don Boudreaux of George Mason University talks with EconTalk host Russ Roberts about public choice: the application of economics to the political process. Boudreaux argues that political competition is a blunt instrument that works less effectively than economic competition. One...
David Brady on the Electorate and the Elections of 2010 and 2012
David Brady of Stanford University talks with EconTalk host Russ Roberts about the lessons of the election of 2010 and what we might expect from the elections of 2012. Brady draws on political history as well as survey results from...
Explore audio transcript, further reading that will help you delve deeper into this week’s episode, and vigorous conversations in the form of our comments section below.


Greg G
Oct 22 2012 at 9:13am

This trend for urban areas to be relatively to the political left of rural areas just might be the least surprising fact in all of politics. Regardless of your ideology, when people live in areas of high population density their actions create more externalities for others than they would if they lived further apart. This creates relatively more need for collective decision making and collective problem solving.

In rural areas, it makes sense for everyone to have their own wells and septic systems. That would never work in a city. Competing private toll roads might work in rural areas but would not be practical in a city. Common firefighting and policing are much more critical in a city than in the country.

Add to this the fact that individuals vary in how much personal space and collective decision making they prefer. Those individuals tend to self select for which system and location they prefer, further exaggerating these effects.

Once you have an urban area, these effects are self perpetuating even if the original reason for urbanization (like having previously been a center of manufacturing) is long gone.

Mark Crankshaw
Oct 22 2012 at 2:44pm


This creates relatively more need for collective decision making and collective problem solving

I beg to differ. I view leftist politics not as “collective decision making” and instead see it as “elite decision imposition”. The left has never really been about “collective decision making and collective problem solving” and the urban areas of the US have never been either. However, top-down “planning and control” by the urban elite with little to no collective input from the majority of citizens is the socialist and urban norm. Our urban areas have the lowest levels of social trust or cohesion in the country, and to an extreme.

I think Russ and Mr. Rodden missed half of the equation in their explanation of why urban areas in the US are so far to the economic and (and more importantly) social left. These urban areas strongly attract people who are leftists and it strongly repels those who are not.

A large majority of Americans, myself included, wouldn’t want to live in major effete liberal cities like San Francisco or New York even if they weren’t currently priced out them. And, even for those who do like such places, the minute that get married and have kids the overwhelming majority must leave for economic reasons. Consequently, there is a steady and rapid rotation of young, single people in these heavily Democratic cities as well as a rapid flow of the previously young but now married with children back to the suburbs. Only those few with sufficient wealth to send their kids to expensive prep schools and those dependent on the government who apparently aren’t concerned with the dismal urban public schools can stay in urban center long term.

For the more dire Democrat urban bastions (Detroit, Cleveland, Baltimore, St. Louis, to name just a few) there is just a slow movement out. Anyone who can escape the urban blight does so. What remains votes Democrat.

I find the urban/rural explanation put forth by Mr. Rodden unconvincing. There are parts of the country that are very rural and very Democrat such as the Missippi delta, the Indian reservations of Eastern Arizona, the border region of Texas-Mexico. What do they all have in common? There aren’t too many White Anglo-Saxon Protestants of the Evangelical persuasion in the immediate area.

Take a look at Census data and religious affiliation data ( would be a good place to start). There is something else that makes our major urban areas different than their suburban/rural counterpart (other than high population density and a large Democratic consensus): namely, these areas are completely lacking the demographic that predominates the rest of the country– the aforementioned WASP evangelicals.

Unlike Europe, affiliation with the Left in the United States has always had a racial, ethnic (and hence religious) aspect. Those on the left have always been racial minorities and, if white, almost exclusively Catholics and Jews. Those countries in Europe that successfully transitioned to a form of socialism were much more homogeneous with respect to ethnicity and religion than the US. Therefore, in the US, opposition to socialism could be neatly packaged as opposition of the socialism of racial minorities and those who are not of owns’ own faith. That packaging of ideology and politics was (and remains) a powerful obstacle to the hegemony of left-wing political movements. And as one who is both a WASP and a staunch opponent of socialism, to that I say: thank goodness for that.

I don’t think that dense population makes socialists (or the Democrats who love them). Our urban areas are just not where WASPs and married couples with children–the base of the Republican Party– can (or, in most cases, want to) live.

Oct 22 2012 at 5:39pm

I agree whole-heartedly with Greg G’s comments above and wish only to add to them:

First, the ratio of public spaces (roads, sidewalks, parks, water, sewer) relative to each individual’s personal space is much higher in the city compared with the country. Relatively speaking, therefore, each city dweller has more contact with public spaces and therefore more contact with the government which manages those spaces. (One of the valid responsibilities of government is to management public goods since individuals frequently mismanage them—ie. the tragedy of the commons.)

Psychiatry teaches—see Daniel Kahneman’s “Thinking, Fast and Slow”—that repeated exposure to a thing leads to the brain categorizing that thing as normal and not threatening. In exactly the same way that frequent exposure to cars leads people to feel safe around cars even though they kill as many people each year as guns and maim a far greater number of people than guns; frequent exposure to government activities—legitimate or otherwise—will lead people to think of government activities as safe even though government activities are–historically and philosophically–the greatest threat to their life, their liberty, and their standard of living in existence. This is why socialism–which I define as the belief that central planning can produce good outcomes—is not persuaded to change by the negative outcomes produced by government programs. Socialism begins with the feeling that government activities are benign. The benign feeling, in Danield Kahneman’s words, is cognitive-ease. Many people, according to Kahneman, are predisposed to substitute the question of “How do I feel when I think of the government?”—which for socialists generates cognitive ease–for the question “Is it a good idea to ask the government to try and solve this specific problem for me?”—which generates cognitive-discomfort because it is complex, nuanced, time consuming to consider, and time consuming to answer. Given that people are lazy thinkers, they are predisposed to answering the easier question because “easy” just feels right. If Americans are to overcome the evils of socialism, we must understand its root in feelings, not reason. Psychiatry can help us to understand how and why those feelings develop.

Second, redistribution programs are vote-buying schemes. The democratic party started them with FDR and continues to expand them with Obamacare. The republican party is unlikely to overcome such a direct financial incentive to such a large number of people when it is illegal to buy votes for anyone other than the federal government. Is it any wonder that the highest concentration of recipients of redistributed income (inner city) is staunchly democrat? If you gave me thousands of dollars a year in direct and indirect government subsidies I might just vote democrat too.

Greg G
Oct 22 2012 at 6:22pm

@Mark and SaveyourSelf

My point was not to argue for more or less government action or socialism in general.

My point was that, without changing his political philosophy, a rational person living in a city should want government to take care of some things (for example sewage disposal) that he should not want government to take care of if he lived in a rural area.

Perhaps this would be clearer if I stated the same point in the opposite way: A rational person living in a rural area should want government to take care of fewer things than he should want government to take care of if he lived in an urban area. For example, the rural resident should not want the government to take care of his sewage disposal.

The economics are different. In the city it is more efficient to have a common sewer and water system. In the country that would be more costly for everyone on average.

I see this as rational self interest and not as the result of lazy thinking or ethnicity.

Mark Crankshaw
Oct 22 2012 at 7:35pm

@Greg G

My point was not to argue for more or less government action or socialism in general.


However,my point was that I don’t believe that our cities contain people who have changed their view over time about the role of government. Rather than “wanting government to take care of some things”, the truly rational choice of those with a desire for limited government trapped in a big city would be to move to the suburbs or the rural areas.

If the only difference between the rural/suburban areas and the urban populations were population density, then the ‘rational self-interest’ argument might make sense.

However, the urban population of the US differs dramatically with respect to age, race, ethnicity, marriage rates, rate of heterosexuality, and number of households with children with two parents vis a vis suburban/rural America. There is a sorting mechanism at work here rather than mass adaptation in “rational self-interest” towards socialism. Those who don’t like big cities and the big government typically found in big cities have either left or actively avoid same cities.

I’ve lived in a couple of big cities (London and Washington, DC), I didn’t become less disgusted with big government in anyway (on the contrary, my hatred for the Left was intensified) and I never wanted government “to take care” of more things. I just left.

Oct 23 2012 at 7:25am

I found this talk was intriguing, but thinking about some electoral geography, I wonder about the fit of the stats outside the US. Some of the safest right of centre electoral seats in both the UK and Australia are in very dense areas: Potts Point in Sydney, for example, is Australia’s densest suburb, and (while not quite as safe) is in the solidly safe Liberal (in Australia, mainstream right of centre) seat of Wentworth; in Britain, the seats encompassing Kensington, City of London and Westminster are some of the Conservatives’ safest; in other Australian cities there have generally also been right of centre wins in some of the densest areas.

At a deeper level, has anyone undertaken an analysis of longevity of continuous office. Once a party has a stranglehold in a district, it will tend to control the powers of patronage – service access, local community groups etc – so that, even as districts change, the incumbent party remains at an advantage if it is able to adapt to new voting constituencies. Drawing again on Australian examples, the main centre-left party (Labor) has managed to hold many areas once “solidly working class”, as they have been subject to extensive gentrification.

Oct 23 2012 at 7:34am

One more point … more on Condorcet!

Oct 23 2012 at 3:21pm

I am very disappointed Russ mis-used the phrase “exception that proves the rule.” It is a perfectly valid logical and legal principle, but misunderstood by too many people. Russ should not have contributed to its misunderstanding.

I would also like to point out that the work of Arrow only applies if there are more than two choices. It’s perfectly reasonable for a family to vote on, say, whether or not to go to the beach for vacation. Such a vote is not subject to Arrow’s work.

Andrew Karre
Oct 24 2012 at 9:35am

I’m one of those city-dwelling left-leaners, and I’ve thoroughly enjoyed EconTalk thus far for its thoughtful challenges to my views. I’ve listened to several episodes twice, so deeply interesting was the content. In this interview, the thoughtfulness and depth vanished completely to be replaced by whatever was in the water at that notorious Romney fundraiser. In all your speculation about why people live in cities, you arrived at proximity to drug dealers, sushi, and coffee (what, no Volvo dealers?) before you arrived at a preference for proximity to places of employment? A preference for a diverse living environment? Instead, you posit that people live in cities because they have a greater preference for government services. If any hypothesis deserved the Robertsian “I’m skeptical” (meaning: “what nonsense”), it’s that one. Or are the millions of miles of roads that exist for no other reason than to allow people to live thirty miles from where they work not signs of a “taste for government services”? (How is it possible for economists to so quickly see public transportation as a government service for the poor and not in the same thought to see suburban roads and eight-lane highways as their middle-class equivalents?) Where is the data to support this astonishing claim? Maybe it’s next to the data that supports the claim that people who live outside of cities are closer to their neighbors because they raise barns together (because when I drive over those millions of miles of non-government-service roads, I’m astonished by the barns I see springing from the earth). Or perhaps it’s under the drug-user I’ve never stepped over in more than a decade of city living. Keep looking, Russ.

In the end, it is probably I who owe you an apology. After the bipartisan decent into vapidity that was the debates, I simply couldn’t take another half hour of non-thinking in one week, and thus I couldn’t even listen to this episode once and aborted the podcast. I’m sorry if I missed you coming to your senses in the second half.

Oct 24 2012 at 4:25pm

@Andrew Karre

The facts presented in this podcast: Most areas of high population density vote “left” (I am assuming here that voting left means voting for policies most similar to the democratic party in America. The generalization of democrats to “leftists” was necessary because the speaker’s observations involved countries where political parties had different names).

The question posed in this podcast, Does Population density cause people to vote left? If so, Why? If not, what other factors can account for this strong pattern?

That people want to live near their work or desire a diverse eating environment might explain why some people choose to live in the city, but does not explain, at least to me, why they would vote left. That people live close together does not give an intuitive reason for voting left either. Adam Smith’s “desire for a sympathy of sentiment”—Econtalk podcast 6/27/2011—might explain the strong coming together of tastes, but does not explain what originated those tastes in the first place.

The causes of voting “left” is important because the left uses redistribution schemes–ie stealing from some to increase the standard of living of others–and centralized decision making–government knows better than you how to make your decisions–to solve societal problems. The trouble with these schemes is they create perverse incentive structures which lead to horrible unintended outcomes such as low standards of living for everyone–except the government elite–and widespread moral decay–evidenced primarily as a generalized acceptance that injury to others is acceptable if it helps someone needy out.

As the population of the world increases and population density rises everywhere, is the human-species doomed to socialism? Was Marx right? Are we doomed to self inflicted poverty and mutual self destruction in communist-Russia fashion?

Regardless, Kudos for listening to Econtalk. I am curious if you can help answer the question even if you did not listen to the whole conversation. Most of the posts seem to list to the right. Some different views on the question would be appreciated.

Russ Roberts
Oct 25 2012 at 12:05am

Andrew Karre,

I think you’ve misread both Jonathan Rodden and myself. My remark about drug addicts was a poor attempt at humor. Obviously people live in cities because they are diverse interesting places to live.

Nor do I think either of us have any disdain for city dwellers or their voting preferences. It’s just a fascinating pattern.

Your point about suburban government services is an excellent one. Yes, roads into the suburbs are part of government spending just as public transportation is. The question is an empirical one–do cities spend more money publicly than the suburbs that surround them? Good question and you’re right to ask for evidence rather than perception.

Oct 25 2012 at 8:00am

bravo econtalk.

apropos to this topic, steven pinker offered a treatment story in the nytimes yesterday:

Why Are States So Red and Blue?

“Another cliché of the cowboy movies also had an element of historical truth. As more women moved west, they worked to end the lifestyle of brawling, boozing and whoring they found there, joining forces with the officials in charge of the rowdy settlements. They found a natural ally in the church, with its co-ed membership, norms of temperance and Sunday morning discipline. By the time the government consolidated its control over the West (and recall that the “closing of the frontier,” marking the end of American anarchy, took place just more than a century ago), the norms of self-defense through masculine honor, and the restraint of ruffianism by women and church, had taken root.”

Oct 25 2012 at 8:01am
Ryan D
Oct 25 2012 at 1:34pm

I was a bit puzzled by the perceived unfairness of all states having two senators. To me, having two chambers of congress with one weighted by population and the other with equal weight given to each state is just a sensible way to ensure checks and balances against federal government overreach. I am by no means an expert on the subject but I believe the rights and independence of the states was a primary concern when our system was established. In theory, this system should help contain federal legislation to areas where the majority of the public and the majority of the states are in favor. Personally, I think it would be equally unfair to states like Wyoming if their representation in the senate was weighted on population as their voice would not hold significant weight in either chamber. Perhaps I am off base on this…I would be curious what others think.

John Berg
Oct 26 2012 at 1:47am

While Ryan D’s spirit seems to match mine more than other commentators, he seems to have overlooked how the Progressive movement is attempting to corrupt the US Constitution. (See the Hillsdale College series of free internet lectures on the Constitution.) The second major corrupting influence in opposition to our Constitution is currently under observation by the BBC’s 2012 Riatt Lectures (Niall Ferguson) and Mark Levin’s “Ameritopia”. The third is the clear steps through judicial actions (See Chief Justice John Roberts recent decision on Obamacare.) and 20th Century Amendments that have lessened the power of States.

While “States Rights” was the clarion call of those wishing to keep Blacks subjugated, Federalism which insists on power to the individual states and limiting the Federal government to the “Enumerated Powers” provides the most optimistic way to ensure maximum liberty for a people who seek individual rights rather than human rights. The idea of fifty states trying different experiments with gay marriages, abortions, processes to provide public K12 education, care for the less fortunate, and controlled immigration seems more correct than being forced to make a binary decisions every 2, 4 and 6 years.

John Berg

[John, you keep mis-entering your website url, using two http’s instead of one. Please check that your link works next time.–Econlib Ed.]

John Berg
Oct 26 2012 at 1:57am

Saveyourself writes:
Second, redistribution programs are vote-buying schemes. The democratic party started them with FDR and continues to expand them with Obamacare. The republican party is unlikely to overcome such a direct financial incentive to such a large number of people when it is illegal to buy votes for anyone other than the federal government. Is it any wonder that the highest concentration of recipients of redistributed income (inner city) is staunchly democrat? If you gave me thousands of dollars a year in direct and indirect government subsidies I might just vote democrat too.

In the JonesTwitter podcast, we are taught that earmarks also discipline the office holders who soon loses independence in order to sit up and beg from their party for passing on earmark benefits.

John Berg

Oct 26 2012 at 11:00pm

I apologize if I am duplicating someone else’s post that I missed, but I think that so far everyone missed the biggest problem with majority rule. (Although I very much agree with Russ’ point about not letting (e.g.) pepperoni lovers outlawing veggie pizzas or vice versa, the fact is that point is largely about busybodies; I have something much more serious in mind, without dismissing the seriousness of fighting against coercive busybodies).

Democracy is two wolves and a lamb voting on what to have for dinner. There is nothing fair whatsoever about that equation.

Those checks and balances such as two senators per state are not some quaint anachronism without objective merit; they are there to help avoid the majority riding roughshod over the minority in an exploitative way.

One example is how Nevada has a Senator (whom I love to loathe) who has to a large degree kept the rest of the country from unsafely (from the point of view of those who live there) using his state as a dumping ground for nuclear waste. Of course one could also argue that he has kept the rest of the country from *safely* storing nuclear waste there too, but that misses the downside of democracy I am referring to.

If that example doesn’t do it for you, then think of slavery as a past example of the ills of majority rule.

Sure, those checks and balances may have been selfishly voted for by some self-interested framers of the Constitution; however those checks and balances really were deliberately put in place to avoid two wolves and a lamb voting to defraud the lamb of his natural rights.

Oct 26 2012 at 11:04pm

P.S. I am quite aware that the pizza discussion was initiated by Russ’ guest. It doesn’t change the fact that my example of pizza type outlawing is exactly what Russ meant…or at least as near as I can tell. 🙂

John Berg
Oct 27 2012 at 1:58pm

Somehow the commentators never get near the fact that supporting the US Constitution, especially the enumerated powers clause, controls much mal-ware. Of course getting control of the Federal level makes possible misusing the tax-payers money since that is where we pool it for collective use. A party can use the taxpayer money to control/bribe its office holders. A office holder can use it to bribe the electorate. The electorate can use it to influence special interest groups. (Funny, how does one explain bailing out the creditors of too-big-to-fail banks, unions, and corporations?)

Was the action of Chief Justice John Roberts best for the American taxpayer?

John Berg

Oct 27 2012 at 11:28pm

I think mulberry bags has econtalk programmed into their phone and misdialed a message!

Thanks for the podcast Russ. I learned a lot. I’m one of those who refuse to remain in the city. I was recently in a large west coast city and was amused to see city-provided plastic bag dispensers for collecting animal waste on corner light posts, but found myself having to step over the human waste left on the sidewalk by overnight bums.

I agree with the self-selecting theory, since people like myself deliberately avoid the city, and since one of the repeated charges in arguments about urban decline is that those who can, flee the inner city. It is useful to note that the ‘left’ population within the cities segregates itself into elite high-end and slum districts. They may vote for the same party, but they are not the same voters. One might argue they are the dependents and those who profess to provide for them if only they are given the power to do so.
Perhaps adults self-select and children raised there are products of the city leftist agenda.

Old-style political machines thrive in the city because the public sector has a stranglehold on services. If a district votes the wrong way they’ll find the garbage collection or snow-plowing less reliable. Public sector employees are often required to live within the district that employs them, thereby providing a powerful voting block that politicians must satisfy. It is for that reason also that unions are organized as ‘locals’ even though the dues may end up in the national organization coffers for political purposes.

In the sticks we pay a private sector contractor for construction, septic, well, and garbage collection, or you do it yourself depending on your skill-set and tools. It’s a simple transaction with no political strings attached. And taxes are lower.

“Keep Manhattan just give me that country-side.”
budump budump bump, BUMP BUMP
Thanks Russ.

[I’ve removed the mulberry outlet spam now. We had a particularly bad spam attack on EconLog and EconTalk in the last week, and several spam comments managed to get past our filters. The usual spam at this time of year turns to variations on popular clothing lines as catchwords, but the ultimate goals are usually driving people to drug or sex websites. I’m working on fine-tuning the filters to address the latest spammer subterfuges. I suggest not using the words “burberry,” “mulberry,” or “acai berry” in a comment for the time being. 🙂 Not to mention, “cherry.” I’m trying to avoid filtering out any use of the word “berry.” Feel free to email me if you see a spam comment get through. It’s probably more effective to email me than to remark on it online.–Econlib Ed.]

Oct 29 2012 at 1:49pm

I do alot of consulting work for a few large, global companies. We look at alot of trends to guide some of our ideas and recommendations. We are always looking for trends that are ‘global’ in nature, and not just isolated to the US or other specific markets. In recent years, the US has finally fallen in-line with the rest of the world’s trend towards urbanization. In the past decade we have seen the rise of ‘mega cities’, and more city centers. The US was an outlier since the 1950’s, when the trend was toward suburban communities subsidized by investment in infrastructure and incentives like the mortgage interest tax deduction. The ‘urban renewal’ efforts if the 80’s and 90’s have worked and most urban areas have lower crime rates (which have not risen even in this recession!) and attract the affluent and young people. Fewer young people say they wish to return to a rural or suburban community now than ever. As the economy has moved away from manufacturing toward media, financial, and creative services, the city has remained the center of wealth-creation. For the most part, the wealth generated in US cities is used to subsidize life in rural and suburban communities similar to other countries with large urban centers. The issue I wish were being discussed in the current political campaign is how we as a nation are going to deal with the decline of suburban areas and the increasing blight of abandoned buildings, shopping centers, overbuilt neighborhoods, as well as the increasing crime and drug-use rates in rural and suburban areas. Its obviously not as bad in suburban areas as it got in many urban areas in the late 70’s and early 80’s, but the tide has definitely shifted. The issues in the cities may have been easier to address because of the centralization of people. More police in the city can have greater effect than the same increase in a large county. The recession is causing many rural county and municipal governments to merge operations for effectiveness, which will only mean that rural communities have even less access to services and support.
I would assume that alot of the voting trends that align with the older role of cities have to do with the fact that older citizens vote most reliably and most people stick to their party affiliations for life. It would be interesting to see if retired voters who move away from the city of their old neighborhoods to the Carolinas or Florida tend to change parties.
My own theory is that cities are the place where different income levels interact the most. I live in suburban Michigan, and the income level and education level ‘feels’ fairly level, even if it isn’t real. There is a greater sense of isolation from poverty or crime. Schools are the only places where you really see a mix occurring, but everyone works hard to ‘mask’ any inequities. Its often shocking to look at a map of crime rates or poverty in a suburban area, because its almost always worse than people assumed. When I visit Chicago, Los Angeles, or Boston, the mix of extreme wealth and poverty are all on display. I think seeing those in need, and those who can clearly help fosters greater support for government services and social ‘safety nets’.

Oct 29 2012 at 10:08pm

Brian posts: “I think seeing those in need, and those who can clearly help fosters greater support for government services and social ‘safety nets'”

Why would it not foster greater support for charity instead?

Check out Arthur C Brooks’ book entitled,
Who Really Cares?

Oct 30 2012 at 8:04pm

I’m dismayed about the lack of civic knowledge Russ and (and especially) Jonathan had regarding why states have two US Senators and the reason for the establishment of the Electoral College.

First, states have the same amount of Senators because the USA is a Republic NOT a Democracy. The 9th and 10th Amendments reserve rights to the individual AND the states. If Senators were elected based on the populace, States’ rights (for what little left there currently is) would be forfeit.

Second, at the end of the interview they discuss the weakness of ‘majority rule’ but did not realize the allotment of US Senators was established to address this weakness; for a consistent majority rule would lead to the tyranny of the minority. Thus, the bigger states would rule over the smaller states, or force the smaller states to form alliances. This was not the intention of a Republican government.

Regarding the Electoral College – in addition to the above, the main reason an Electoral College was founded was so the ‘minority’ (both states and individuals) would not be ruled by the ‘majority’. To illustrate, the number of presidential candidates would increase, and those candidates would only need to raise enough money to campaign in the most populated cities, not states. This would lead to candidates only needing to accommodate to the needs of the electorate in those cities. Therefore, the Midwestern and the Northern states and individuals would be shut out of the electoral process.

Lastly, a point that seems unrelated, but is not – by far, the most damage done unto States’ rights and to the Republic was the passage of the 17th Amendment to the Constitution. This passage forfeited States’ rights, which was in direct contradiction to the 10th Amendment and has since paved the way for the elimination of Republican rule and the blossoming of Democratic rule. Unfortunately, there are scarce resources in regards to the intended and unintended consequences to this passage. Once understood, this would clarify to both Russ and Jonathan the necessity for two US Senators and an Electoral College.

Nov 7 2012 at 9:38pm

I agree. I think it should inspire charity. I would be curious to know how charitable giving compares between rural, suburban, and urban areas. But my guess is that seeing others in need inspires charity. Seeing those in need AND the very rich on the same city block brings up questions about the ‘fairness’ of resource allocation in the community. How to address those questions is the challenge. But think about taking a small child to the big city for the first time where the poor and homeless are actually visible (they are often unseen in suburban areas), and at the same time there are tall buildings, and expensive stores, restaurants, and hotels all around that are obviously the result of great wealth and capabilities. This experience definitely caused my child to ask some hard questions about why we allow people to go homeless when there are so many resources all around.

Nov 8 2012 at 6:13pm


“You find that people who believe it’s the government’s job to make incomes more equal, are far less likely to give their money away”

Nov 10 2012 at 4:14pm

I was quite disappointed with the lack of intellectual rigor of this podcast. I have come to expect more from Roberts and his guests. Listening to Roberts and Rodden discuss why cities tend to be bastions of liberalism sounded like archeologists confidently (and somewhat disdainfully) discussing the lives of ancient Egyptians based mostly on documentation from the ancient Greeks.

The responses to the podcast on this page were pretty much what I expected to see: a majority of the listeners indicating that their biases were confirmed about the character of people who live in cities. As with most things, the geography of politics is at once more simple, and way more complicated.

The most glaring omission in Rodden’s analysis of the political differences between cities and suburbs and rural areas is that cities have always been the place for people to go when they can’t fit in or aren’t accepted in their home towns. Gays, lesbians, transgender people, artists, scholars, and intellectuals, adventure-seekers go to the city because that’s where you can be yourself, find others like you, find the great seats of learning and curiosity-quenching opportunities. Cities have also always been the place of economic opportunity. Unfortunately there are almost always fewer jobs than enthusiastic job seekers, so we end up with an underclass of people with bad timing, more hope than practical skills, and sometimes bad judgement. They are humans, after all. City dwellers are perhaps more amenable to government programs not because of some Stockholm Syndrome, but because managing the threats to urban decay are critical to the survival and health of the city. (Oh, and we elect our representatives, just like you do in the suburbs and country. They in turn hire specialists to develop programs that reflect the people’s interests. If we don’t like what they do, we can elect someone else. There is no dictatorship of the elites. Where that idea comes from I do not know. If what the politicians and policy makers do costs to much in taxes, we say enough. It works pretty well in my city, at least.)

The second most glaring omission in the discussion was any talk about the Second Great Migration and subsequent, massive “white flight” to the suburbs. To put it bluntly, the blacks moved in and the whites moved out. We can get all upset and insulted by this, but xenophobia is a part of being human. If you live in a place where everyone is more or less like you, you are going to be a very bad judge of the character of people who aren’t.

I can walk 5 blocks south and see million dollar homes and 5 blocks north and be at a rent-subsidized highrise. And I can see the residents of both at the local Target. There are also three churches, a synagogue and a mosque within a 15 block radius. When you live with this kind of economic and cultural diversity, it is much harder to create straw men and caricatures of the “other” (though we are guilty of occasionally creating caricatures of people who make us out to be a bunch of Volvo-driving elitists in search of swank cafes). Yes there are homeless people, and probably drug users (though I’ve never had to step over either on the sidewalk), but these souls just remind me that life can be unforgiving and difficult; that some people, through bad choices and bad circumstances fail to thrive. When there is too much poverty in the city, the city fails to thrive. When this happens you use whatever tools you have to solve the problem be they private, public or some combination of the two.

If ever there was a good example of “scientism” at work, this podcast was one.

Nov 10 2012 at 4:27pm

But Ralph, I believe these count donations to social clubs and churches with donations to the local self-empowerment agency, or homeless shelter. It could also be argued that the enormity of suffering makes sending a $25 check to Dorothy Day seems like an ineffectual drop in the bucket. You as Brian notes, when you see the vast differences in wealth, you are more inclined to see a systemic problem that requires the use of the levers of power to fix.

I am not saying one is more correct than the other, but I think there isn’t a simple explanation. When people see a problem, they respond according to their sense of proportion.

Nov 11 2012 at 5:39pm

Even excluding church donations, those who want less government intrusion give more. It’s in their ideological system. It’s how they think. And remember, they are also still paying taxes at the same rate as those who give less. And statistically, they earn less than those who want bigger government.

What do you think of the NYC restrictions on donations to food banks and shelters, and the closing of charity “tax loopholes” by the current administration, or the fact that volunteers aren’t allowed to help clean/fix schools because of union restrictions? Some volunteers have even been turned away from assisting in the hurricane cleanup.

Nov 12 2012 at 9:12pm

Ralph: Where is your data to support this? Also, I think it is important not to lump together religious conservatives (who do appear to want bigger government when it comes to their religious laws), and economic liberals who believe in free market principles and may or may not be religious. The charitable giving of religious conservatives may be more indicative of their core religious beliefs than in any free market ideology. A U of Michigan, Bank of America study indicates that as much as 67% of all charitable giving from households earning less than $100,000 went to religious organizations. I have seen separate studies that indicate that those who attend a house of worship regularly give more to both the religious institution and to other charity. But I haven’t seen a correlation between that information and political views. It also doesn’t indicate where the donations are going outside the faith. Are they going to the local food bank or homeless shelter or battered women’s shelter or are they going to one’s alma mater or fraternal order? Studies show that women give more than men (See Schervish et al, 2006 pp 550-551), but polling from the latest election confirmed previous polling that women’s voting patterns indicate that they are more liberal than men.

There just isn’t enough information to state firmly that the giving patterns are indicative of any one thing. Unless you have data showing the results of donors being asked explicitly about why they give and their beliefs in the role of government, we don’t have enough information to make a firm statement.

In the context of this podcast, it is interesting to note that charitable giving tends to be local (Schervish et al 2006, pp551), so a homogenous middle-class suburb is likely see the residents’ money go to charities serving their own community than to those serving the destitute in the core city.

I don’t know much about the charity loopholes that the current administration has implemented. Could you provide a link to the rulings?

As for the union prohibitions against volunteers. I am with you. I worked for a neighborhood organization in the 90s and residents wanted to paint and otherwise spruce up a local, run-down playground. The city barred them from doing this because of the union contract, but there was no money to pay city workers to fix up the park. So we organized. And when enough people heard about this ridiculousness, the proviso was removed in the next contract negotiations. If not enough folks are paying attention, and most of those who are, say “well what do you expect from government”, then we get the government we deserve. Democratic-republican governments are very good at self-fulfilling prophesy.

John Strong
Nov 14 2012 at 8:56am

Inverse Schumpeterianism

I think that you have approach the phenomenon of leftwing cities with a Schumpeterian-style analysis, but standing Schumpeter on his head.

Schumpeter believed that capitalism contained the seeds of its own destruction because it breeded unruly individualism. I think precisely the opposite is true.

People who love cities are those who are most temperamentally inclined to enjoy the fruit of trade and specialization, which after all, is what cities are for! But these leftwing city-dwellers don’t understand the origins of the emergent order they love, and so they draw the mistaken inference that it is the result of planning, hence they are inclined towards socialism.

Capitalism teaches people to have an *expectation* not just of wealth, but also of order, trust of strangers and rule of law. People in pre-liberal societies do not *expect* cooperation from anyone other than their closest family members. Thus socialism is a kind of capitalist auto-immune disease where people come to expect levels of wealth, coordination and trust that are unrealistic then try to remedy this perceived deficiency by state-imposed solutions. If only they knew how their expectations have been shaped by the commercial culture they repudiate!

Nov 19 2012 at 5:12pm

Ralph: I remember seeing that John Stossel report. I am not inclined to trust those kinds of reports because there are usually some key omissions made to make their point. There may be a sense in more populated areas that if ‘I’ don’t give to charities, someone else will. The report also didn’t count the work of non-profit organizations and charitable trusts. But let’s assume that suburban and rural residents give a higher percentage of their income per capita. They also make less and save and invest less. They also benefit more from government regulations of utilities, government subsidizing of mortgages, and nationally, the less-populates states take more support in dollars from the federal government than they contribute. The idea that John Strong suggests, that cities are inclined towards ‘socialism’ (we are ‘socialist’ to some degree. Where is the line that divides socialists from non-socialists? private versus public fire departments? regulated versus unregulated utility costs?), and rural areas are bastions of ‘rugged individuals’ seems like a matter of perception. I suspect its not reality.

Comments are closed.


About this week's guest:

About ideas and people mentioned in this podcast:Books:


Web Pages:

Podcasts and Blogs:



Podcast Episode Highlights
0:33Intro. [Recording date: October 4, 2012.] Russ: We're going to talk about your research on voting and geography, and if we have time we'll explore some other issues you've been working on as well. But I want to start by talking about the book you are working on, which has at least the tentative title, The Long Shadow of the Industrial Revolution: Geography and the Representation of the Left. Let's start with what seems to be true in the United States geographically. People in cities seem to vote differently from people in suburbs and rural areas, with a lot of consistency. What do we know about this pattern in the United States and outside the United States? Guest: It's something that I think anyone who looks at an electoral map in the United States can easily identify. We've been mainly looking at county-level maps over the last few elections, and it's hard not to see that kind of a pattern. When you look at a state like Missouri, for instance, you'll see that St. Louis County and the counties around Kansas City are blue, and the rest of the state is kind of some shade of red or purple. Russ: Where blue means supporting Democrats and red means voting for Republicans. Guest: Unfortunately, those are the colors that American cartographers have picked up. In the rest of the world, of course, red means Left. We have to do things a little bit differently; makes it hard for me when I make the maps to try to explain the argument in the book and elsewhere. But yes, Americans are always referring to the Democrats as blue for some reason. But this is something that shows up in those county-level maps. But what we didn't see, that is something I've found interesting, is I've been investing a lot of time and effort into collecting precinct-level data, and then trying to make some precinct-level maps--and I've been doing this for a variety of other countries as well--is that this kind of correlation between population density and voting behavior seems to hold even at much lower levels of analysis. So that even if we go within some county, even in rural Missouri, and we go from the very sparsely populated periphery of that county and we move into the County Seat, a place that has a little bit of rental housing, maybe clustered along some railroad tracks, we see that correlation between population density and voting behavior. It shows up even at that fine-grained scale. And that's something I found really interesting and started looking around to see where is that kind of more generally. And it's true in lots of places. I'm discovering it's not completely universal. There are conservative cities. There have been, some of them kind of a long history. So, Stockholm is a good example of a place where there is a clustering of conservatives in the city center, and there has been for some time. There are well-known neighborhoods in Madrid that are kind of, everyone can point to, that are densely populated and are conservative. In the United States, it's really quite striking--I think the only example of a place that is very densely populated that votes consistently for Republicans is little Havana in Miami. Russ: Kind of the exception that proves the rule. Guest: Exactly. So that correlation is especially strong in the United States, and what I've found is that it's very strong in the other countries that use these winner-take-all, majoritarian fusions. Which, really I'm referring to places colonized by Great Britain and Britain itself. So we see that correlation between density and voting in Britain; in New Zealand, especially before they changed their electoral institutions; in Canada, it's very striking as well--so, I've made some precinct-level maps of Canada where you look at the Great Lakes states in the United States and you look at Southern Ontario and that part of Canada, and the electoral maps look identical. You can't even tell where one country stops and the other one begins. You have that clustering of Left votes in the cities. So, it's a very common pattern. And one of the things I'm still trying to understand is under what conditions do we really see, is that pattern most clear. And this is where--the book ultimately doesn't have in the title anything about urbanization or cities. It says something about industrialization. And I think that is really the story. That when I mention that Stockholm has these dense conservative neighborhoods, I think that has something to do with the fact that the process of industrialization was different in Sweden than in the United States. Russ: Now, Jon, before you get to that, just explain for a minute: When you use the phrase "winner take all," you are referring to the fact that once you get more votes than the other person, you get the seat. It doesn't matter whether you win 80, 90%, or 50.1%; or if there are multiple candidates, if you win by one vote. Right? Guest: Right. And this is the way elections were conducted in most of the world up until a period around WWI. You divide up a country into districts, and there are candidates running in those districts, and the person with the most votes wins. It could be a four or five party split and someone could win with 30% of the vote. And in fact, that happened quite frequently in the late 1800s and early 1900s. So, most European countries had a system much like the United States with winner-take-all districts. But then there was a big transformation between the turn of the century and the tail end of WWI, where a lot of continental European countries switched to a system of proportional representation, where the number of seats you received in the legislature is proportionate to the number of votes received by a party. Which has the effect of inviting, of creating more political parties. So, when I teach a course on political institutions, one of the things we think about for the United States is that if the United States had proportional representation, what kind of a party system would we have? And most students, when they kind of think about it, believe there would be a Libertarian Party that would be more successful; and perhaps there would be a party that was more focused on African Americans. Those are at least a couple of the possibilities. Finer grained distinctions on the two. Russ: Green Party. Guest: Yeah, Green Party. And maybe a party that is economically liberal but socially conservative; and vice versa. That's the kind of thing that seems to emerge. So these electoral rules can have a really important impact on who gets represented and what politics looks like in a country. And what I'm doing in this project is thinking about that from a perspective that's a little different from the way we usually think about it. I'm linking it to geography and I think that's part of the reason why these countries look very different. When you combine the geography of preferences with these electoral rules you get really different kinds of results.
7:56Russ: Which countries have proportional representation right now? What are some countries that do it that way rather than districts with winner-take-all, one seat per district or a number of seats per district? Guest: Most of continental Europe uses some type of proportional representation. So, maybe the country that people tend to think of as the purest form of proportional representation is the Netherlands, where there is a very high proportionality between the percentage of the votes and the seats. And there's really just one single national district. The country isn't even divided up into various electoral districts. The same is true of Israel. So, these are kind of thought of as the purest forms of proportional representation. But there are also proportional electoral systems in most of continental Europe. One country after another made that switch. To the point where it's really the wealthy developed countries that still use single member districts are limited to France and the former British colonies that I mentioned earlier, including the United States. Russ: What about the United Kingdom? Guest: The United Kingdom is still the classic case of a winner take all, majoritarian system. And it's one that is still--many of the other countries that were colonized by Great Britain also use these institutions. So, once we move beyond the wealthy countries--India uses a single-member, district, winner take all system, very much like Great Britain. And a lot of the islands in the Caribbean that were colonized by England , and several African countries. Russ: A lot of people have looked at--I know it is real and speculated whether it's relative instability of party leaders and prime ministers is a result of its fractious party system, where there are lots and lots of parties, along the lines that you talked about. Do other countries have that? I mean, in the United States, we only have two strong major parties. That has benefits, but lots of negatives. The result of that is that the president usually stays in power for 4 years, unless, God forbid, there is an accident, or a crime. It's very rare--the House can turn over every two years but often stays in the hands of one party for a relatively long period of time. And so people claim that the United States, because of that and because of the Constitution, tends to be a little bit inertial. Doesn't have a lot of ability to change dynamically and quickly. Whereas these other systems change more quickly but they are less "stable." Which obviously there are costs and benefits of either system. But is that true about proportional representation? Is it less stable, more "fractious," whatever that means? Guest: Well, I think the problem with that claim is that proportional representation contains such a wide variety of different types of systems. And so Israel might be kind of at one extreme. There are ways of making a proportional representation system a little bit less chaotic. So, kind of an earlier example of exactly that complaint--some people blamed the rise of the Nazis and the breakup of Weimar Germany to an excessively proportional system that created lots of parties and chaos, and constantly coalitions falling and new ones forming, and elections being called. So in the German, post-War, constitution, they created a system that is far less fractious. And they have a thing called a constructive vote of no confidence: so you can't take apart a government unless you have another one prepared, already in place. They also have a 5% hurdle--so, you have to have 5% of the vote to get representation in Parliament. So, lots of proportional countries have institutions like these that make them a lot more stable. And it's also the case, and voters in many Scandinavian countries for instance, have a pretty clear sense of which parties will coalesce. And so they go to the polls almost thinking, even though there are 5, 6, 7 parties in some instances, there are often two big blocs of parties; and it's pretty clear which parties will work together. It doesn't always create as much chaos and kind of blackmail potential for the small parties as one might expect. But certainly there are some countries when that happens. The proportional system in Italy in the past has been blamed for some of its difficulties. And Belgium as well in the past has again been a case where a lot of fiscal indiscipline that people have blamed on multi-party coalitions. Russ: Well, it gives minority parties more power than they might otherwise have, not just in terms of number of seats, but in ability to be part--or as you say, blackmail, or veto a coalition-creation. That's something that as Americans we don't have that experience via that mechanism. There are, within each party and sometimes within the legislature, minority viewpoints that can be decisive in similar ways. But it's not the same structure. Guest: Yes. All that diversity gets reflected within the two major parties in the United States. And a lot of the negotiations take place in a kind of a different way. Whereas they might take place between parties in a proportional system. They take place between, say, the Tea Party Caucus and the rest of the Republicans. Or the Congressional Black Caucus and the rest of the Democrats.
13:40Russ: Going back to this point about density--so, we just had a little digression on forms of government. Let's go back to your point about geographical density and the implications for political views. So, there's this strong correlation between population density and political preferences of the voting population. When did that start? When could you first point to that as an empirical regularity? How old is it? Guest: I think that the answer is it's a little different in Europe than in North America. It seemed to correspond to the Industrial Revolution in Europe. Once peasants left farms and moved to cities and started working as wage laborers in factories, they became targets for the mobilization efforts of Socialists and Social Democrats and workers' parties. The very fact of this density and the creation of labor unions is what generated the kind of Left/Right politics that we know, in many developed countries. And urbanization became almost a prerequisite for the generation of the Left as we know it. In the United States it happened a little bit later. The Democrats didn't clearly become the party of urban workers until around the New Deal. So, one of the things, some graphs that I've made that haven't quite made it into the manuscript yet--if you look at the county-level correlation between population density and voting, you just see a lot of nothing up until the 1920s. And then toward the end of the 1920s and into the 1930s all of a sudden this relationship emerges. And it just gets stronger and stronger over time. And that's one of the things that I've puzzled over, is that it seems to have a lot to do with the location of factories and railroads and shipping routes and things like this, warehouses. But the relationship has become even stronger, even in the last 15 or 20 years. So, there seems to be something even more to this relationship in the United States than the mere fact that we have 19th century housing and sort of dense working class housing in the city centers. There seems to be a bit more to it. Maybe it has something to do with--because some of the neighborhoods that are voting for the Left that are very dense are not proletarian kinds of places at all. They are downtown San Francisco and the Upper East Side in Manhattan; downtown Seattle; places like this that are very high-income. Russ: Not a lot of folks working the assembly line. Guest: Exactly. And this is something kind of fascinating that's happened in a lot of cities, not just in the United States. There is this moment at which the working class housing of the early 20th century empties out; there are no longer workers there; and sort of a very new type of person moves in, who is mainly interested in buying, being able to walk out the door and get a certain type of latte, or certain style of sushi. And that's what the city brings them. It's not proximity to a place where they are sharing tools with a bunch of other workers. It's something about consumption opportunities. And so from Sydney to London to San Francisco and New York, the vote share of the Left remains remarkably consistent as the working class moves out and this new class of urban, for lack of a better word, yuppie, moves in. The labor vote share in Australia, in England, stays the same. It stays at 85% or something. Russ: And as you point out, and this is one of the most striking correlations--it's hard to imagine that it's causal, but you make an interesting case--you point out is that voting patterns in U.S. Presidential elections don't mirror the location of manufacturing. Tell me if I got this right. Do not mirror the location of current manufacturing jobs, but do mirror the location of manufacturing jobs decades ago. Guest: Exactly. This is one of the things that really hits you over the head. If you go into the current, 2000 or 2010 Census and you get some data on manufacturing employment as a share of total employment, and you plot that against the Obama vote share or the Kerry vote share or something like that, there's no relationship at all. But if you go back in the 1880s Census or the 1910 Census and you do the same thing, manufacturing as a share of total employment, you see this striking positive relationship between historical manufacturing employment and current Democratic vote share. And so what my argument is about is that these places that are voting, you know, 85, 90% for the Democrats tend to be places like downtown Cleveland, where there's no manufacturing to speak of any more. But all of the housing that was built for the workers in the early part of the century is still there. And so all these places that industrialized before the rise of the automobile have these dense kind of proletarian neighborhoods with affordable dense apartments. These places kind of attract the type of person who tends to vote for the Left, and that happens in lots of different countries. That's kind of, I think, the thing that links all this together in these various countries. And this is where that density correlation comes from. Russ: Now, you just said that in a particular way. You said it attracts people who tend to vote for parties on the Left. I think when the average person--well, let's not say the average person. As an economist, when I think about this phenomenon, which fascinates me: Why are cities so--it's not, as you point out, 55-45, Democrat-Republican. It's 85-12, it's 90-to-7. Your natural thought as an economist is--there are kind of two natural ways to think about it. One is the way you just describe it: that people who are attracted to these types of living arrangements of high density happen to be people of the Left. The other alternative way to think about it is when you live in close proximity to other people you have a taste for larger government, say, or you are going to want more government services; you are going to rely more on government. Another argument would be: in a dense urban area people are a little more scary, strangers are more prevalent, you don't have the ties you'd have in a rural environment with your neighbors because you are not building a barn together. And so you've got to have something to substitute for community. And these sort of--these are kind of cheesy, I don't know what you'd want to call them, armchair explanations. Have people written about this? I assume they have, in Political Science. Is there any argument there? Guest: Yeah, there's kind of less than you would think, but these arguments, they are all kind of there in one way or another. To me it all kind of boils down to: the different stories you are telling right now, is it a selection effect or a treatment effect? Russ: Correct. Guest: So, do cities actually--if we could randomly assign individuals to suburban, rural, and urban environments, would we see that the city actually changes your preferences? Because you understand something about the value of public transportation, for example, if you are an urban resident. You are tripping over drug addicts on the way to work and you want something done about that. Or, is it the case that people with Leftist preferences are more comfortable-- Russ: Want to live near drug addicts. Guest: Exactly. High income people who have the choice. So that's another further refinement. There are people who have the choice of whether or not to live in cities. There are also those--a lot about the housing--I think the combination of the housing availability and actually there is some good work by Ed Glaeser suggesting that one of the reasons why the poor cluster in cities in the United States is because of the transportation infrastructure and the lack of automobile ownership among the poor. So that a combination of this 19th century housing with a bus and train network makes it so that--it could be income that's doing most of the work in the story. And I think that probably still is the case. The stories about the Upper East Side and the Gold Coast of Chicago and San Francisco--those are a little bit unique. For the most part we are talking about people at the bottom of the economic spectrum, the clustering of poor people in cities. That has various explanations. Russ: When we talk about density, right now, most of the time we've been talking about urban density--that is at one end of the extreme, very dense. How does Left/Right vary with suburbs, and then rural? Guest: So, if you think about a graph that on the x-axis, the horizontal axis, is just the origin is the center of some U.S. city, and as you travel along the horizontal axis you are getting more and more miles from the city center, the Republican vote share is pretty much an increasing function of that distance. So, as you move out into the early developed suburbs of the 1930s, 1940s, 1950s, you end up with places that are sort of modestly Democrat. And then you get into the middle suburbs and you find these really purple places. And then you get out to the exurbs and it's pretty Republican. And then you get to rural areas and it's very Republican. But rural areas are never--the precincts are never as Republican as the urban precincts are Democratic. So, you don't really find 85, 90% Republican precincts. Really almost anywhere. Which is sort of an interesting fact. I'm not quite sure what explains it. Russ: It's fascinating.
24:18Russ: So, let's get back to this role of industrialization. What does that have to do with it? Is it just this strange, how this housing stock that's left over from the fact that workers once lived very close to where they worked--they didn't have cars to commute; there was a big factory that hired a lot of people so a lot of people lived near it, and so they were going to be densely housed? And whether it's treatment or selection, it's that housing stock today that's still affecting political outcomes? Is that your story? Guest: Yeah, that's my best explanation for what we see. Because housing stock is so resilient. Once you put the bricks and mortar into place, they don't go away. Housing doesn't really just disappear. I mean, we have our efforts at urban renewal that we occasionally go after. Europe, much of Europe was bombed in WWII; but the striking thing about that was that they often built on the same footprint that the bombed out building--they built a pretty similar style of building, similar height, similar kind of number of units and for a similar type of worker. So you didn't really see the change in, say, German residential neighborhoods after WWII that you might have guessed. So, my point is that there is something about the built environment--that once a city gets built up, its basic structure, while there's plenty of suburbanization that takes place, but that urban core generally something about it is very resilient. And it kind of makes these patterns last a very long time. It's not even just the cities. Russ: So, when Europe moved away from winner take all to a proportional representation system, you tell a story in the book about the role that, as you said earlier, the factories start up because of industrialization; workers become unionized; they are more sympathetic to Socialism; the far Left party emerges as a Socialist Party. And of course that's a threat to existing parties on the Left and to some extent to parties on the Right. How did that urbanization and industrialization affect the evolution of political systems in Europe? And why didn't that happen here in the United States? Guest: That's a good question. Kind of more broadly I'm interested in trying to understand why does Britain still have a single-member district system? Why did it also not switch to proportional representation? Because it came very close. Now, we've never really had a moment in the United States where we were really on the verge of choosing proportional representation nationally. But they did have a moment like that in the United Kingdom. And in fact they have made the transition in the 1990s in New Zealand. As I try to understand it, the way I think these proportional representation systems ended up coming into being is there were actors, there were people who had safe seats in cities. Often it was in Europe it was liberals versus conservatives were the main parties in the early part of the century. And as the working class gained the franchise and was able to vote, these Leftist parties started to win in the cities. And so what that was doing was it was squeezing out the existing parties, who saw that they could only win 30% of the vote, whereas they used to be able to win, say, 60% in some district. And so they still had this sizeable share of the vote, but it was spread out around the population, across the districts. And so it became obvious to these parties that they were going to be squeezed out of existence if they didn't make a change in the electoral system before it was too late. And so those pre-existing parties often tried to make a deal; the various parties that were going to get squeezed out tried to make a deal that led to the adoption of proportional representation. But there was never really a far Left party in the United States that started this squeeze. The Socialists and the Progressives, the Socialist platform was adopted, taken over by the Democrats in the 1930s. And so there wasn't really this moment when an insurgent party was going to push the Democrats out of existence. And perhaps at that moment, the Democrats, they if were smart, would have pushed for proportional representation. But that never happened. Yet in Great Britain, the Liberals were the party of the Left before the Labour Party came around. The Labour Party started to quickly gain support. And the Liberals, had they seen where things were going, they would have insisted on proportional representation. In fact, they eventually tried. And to this day, the Liberals are still pushing for proportional representation in the United Kingdom. But they didn't see that they would be squeezed out and that the Labour Party would become the main party of the Left. And so they kind of missed their opportunity. And they've been ruing that ever since. And they've been squeezed into a kind of marginal, third-party status. It's another version of that classic question about the United States: why did the United States not develop a strong socialist party that squeezed out the center-left party? It's probably something that lies a little bit beyond the story that I'm telling about geography. There's probably a lot more to it. Russ: So, the story that you are telling--see if I have the intuition of it. The intuition is: If I used to get 55% of the vote in my district, and now I get 45%, or 30%, I get nothing. Because I have to win. Guest: Right. Russ: So proportional representation lets some of my team stay in power. It reminds me of the way you hear some people talk about redistricting in the United States. So, you take a guy who has got a 60% vote share, a 70% vote share; and you know he's of the other party. And you know you are never going to win. So you give him a piece of somebody else's district that makes him win by 90%. But that doesn't cost you anything, because he still gets his one seat. You take the other people and you redistrict and you get your share up above 50%. It's somewhat similar in terms of the intuition. Guest: Yeah. The intuition is somewhat similar. There's an underlying distribution of political preferences in space. But then when you map onto that some district boundaries and you make them winner take all, and then you have the legislature that is made of those representatives, you can end up with a very different distribution in that legislature than you saw in the society. And it matters a lot how you draw those lines. And that's what redistricting is all about in the United States. Russ: And you suggest in the book that because of the dominance of left-oriented voters in urban areas, in winner take all systems such as in the United States, the Left in the United States does not get the power it might otherwise get in a proportional representation system. Because it is sort of dissipated in this surplus of extra votes it doesn't get any extra credit for in these cities. Is that a fair statement? Guest: Yes. So I think for a while now, people have noticed that there are urban districts in the United States that are extremely Democratic; and they end up--if it's true that at the precinct level you have a bunch of precincts clustered together that are 80% Democratic, it stands to reason that you draw a winner-take-all district around them and you will have an 80% or 75% Democratic district. So what we end up with in a lot of states, certainly not all states but a lot of the more industrialized states, larger states, we end up with one or several districts, say, like the Cook County, the Chicago districts. The way in Missouri around St. Louis or Kansas City; or in Indiana, it's Indianapolis and Gary--where there will be a Congressional district that is extremely Democratic but then all the other Democrats who are spread throughout the state, in these smaller cities and in the rural areas, there are not enough of them to win districts in those places. And there are too many of them, so they win by large majorities in the cities. And so you end up with an asymmetry in the transformation of votes to seats.
33:00Russ: And you get what economists call infra-marginal and other types of words that describe the competition. So if you are in a very safe seat, you don't have to work very hard to please your constituents, because you've got a big margin of error--or of corruption, depending on your preference. How you describe it. Guest: Exactly. Russ: So all the competition takes place at the Primary level, at the jockeying for who is going to get the blessing of the party's powerful. Because once you get that nomination, you are in. There is almost no chance--I live in Montgomery County, Maryland. Very hard to get a Republican on the County Council, or a Republican in my Congressional district. So, you'd think that would change the incentives of the people in power to please their constituents. Guest: Yes. In urban America, the only election is the Primary. That's a pretty basic and interesting fact I don't think we've fully understood. The only thing you have to worry about is if you somehow screw up and attract a certain type of Primary challenger. But the general election is a foregone conclusion. Russ: Right. Guest: And so you can think about the various ways in which that might affect incentives of incumbents. It's great for the incumbents. It's not good for the Democratic Party. In a state like Florida, you end up with the incumbents in the party are people from Miami-Dade who represent very left-wing constituencies. And then when the voters in the rest of Florida look at the Democratic Party in Florida and try to assess what they are all about, that's what they see. Russ: Sure. Guest: And that's the kind of conclusions that they draw. They look at the kind of representation they think they receive from those individuals. And it makes it hard for the Democrats to compete in state politics. A state where the Democrats do very well in Senate elections and gubernatorial elections and presidential elections of course are notoriously close. But in state districted elections, the Republicans are dominant. Russ: We've been talking about this--the implications--at the national level. Think about urban national politics and policy outcomes. When you think about it at the urban level, we have a lot of core cities in the United States that are not doing very well. There are a few that are--New York being one of them. San Francisco has done pretty well. But I used to live in St. Louis--the urban core there is very dysfunctional. Kansas City is not very healthy. Detroit--an obvious failure. Cleveland, not so good. Part of my first thought as to why those cities have done so poorly and why their suburbs have done so well is because of this lack of competition for urban policies and urban choices. We have what you could call one-party rule in the cities. And that's not because they are Democrats. There's nothing bad about Democrats' policies per se. It's that any time you have a monopoly--essentially what feels like a monopoly--even though there's an election every four years for the mayor, but if the party that the mayor represents wins every time, it can't be very good for serving the customer. Guest: Yeah, I think that's right. It seems pretty obvious that competition is good. And there's lots of evidence to that effect--that the politicians who are forced to contend with very close and difficult re-election battles are much more in tune to the preferences of their constituents. They put forth more effort. They just try harder. It's a very intuitive thing to claim, and I think there's a lot of evidence to back that up. How much of the kind of bad governance in urban America can we attribute to this? How much of urban decline, what caused what--that's a harder question. I don't think I have all the answers to that. Russ: For sure. Guest: Pretty complex one. Russ: Fair enough. Guest: But I think that basic claim is a good one. And I think we see similar things are true in lots of places. It's better for the voter to have competition.
37:22Russ: Now, we're recording this in October of 2012. So we are just about a month from a Presidential election. And we have this weird thing in the United States called the Electoral College. We have this other weird thing, which is not that different from it, which is: Wyoming gets 2 Senators. Wyoming gets 1/50th of the Senate even though they are nothing close to 1/50th of the population. So, on the surface it's an historical accident. When the Constitution was voted on and decided, there were all these political forces, urban and at the agricultural versus cities in terms of jockeying for political power. And one of the ways that competition emerged was through this strange system we have, where every state gets 2 Senators; but being a Representative is more proportional to population. Although every state gets 1. You can't get a third of one. So, as a result, we have a policy that you could argue over-represents that sparsely populated, rural, Republican-leaning Wyoming and doesn't fairly represent New York City or Los Angeles or Chicago at the national level. Is that an accurate description? Guest: Yes, it is, and I think the aspect of that that I've looked at most carefully is just looking at the flows of intergovernmental transfers and other resources that are transferred from the Federal government to the states. There's a pretty striking correlation between the states's Senate representation per capita, legislative representation per capita, and the amount of Federal funds it receives, so that the states like Wyoming are at the very top; and they've received far more funds per capita than the large states. And as for the implications for national policy and whether it pulls policies to the Right, it certainly would seem to--it's not a stretch to imagine that if there's an urban/rural dimension to politics that's going to push things a bit in the rural dimension. I mean, it's not quite as dramatic as we might think, because some of the overrepresented states are actually densely populated, like Rhode Island and Delaware, the initial states that actually created the rule in order to protect themselves. But eventually, it's kind of interesting, the story of how some of the sparsely populated Western states came to be states. You know, there was once a Dakota Territory, and it became North Dakota and South Dakota, explicitly because it was useful at the time in bolstering a political majority, to create four Senators rather than two. And that certainly continues. One of the interesting things about this is that within Senate delegations--so within a state like Ohio, or Tennessee, or one of these ones that has a really asymmetric distribution of partisans within the state, sometimes you see the Senate delegation is actually a little bit to the Left of the House delegation. Because the Senate is just a kind of winner take all for the whole state; and so, these urban votes that don't matter so much in the House elections, they matter just as much in the Senate election as any other vote. So you sometimes see that a kind of Left-leaning Democrat can actually win as a Senator in one of these states. And so in some of these moments you have a Senate delegation that is a little bit to the Left of the House delegation. But in general it's correct that these Western, sparsely populated states are overrepresented and it kind of pulls the Senate as a whole to the Right. Russ: You'd think that would bother some folks, just the way the Electoral College does as well. I think a lot of people resent--of course, it depends who you are rooting for and how you feel about it may change depending on the nature of the Electoral College, but you always get these appeals that we need to abolish the Electoral College and it's such a weird thing and it should just be the popular vote that elects the President. We started off our conversation before we were recording--I think it was before we were recording: you emphasized your interest in positive outcomes. By positive you meant merely--not the everyday use of that word--but rather, in social science positive means the way things are. Not whether they are good or bad. That's what we call normative outcomes. Normative outcomes, we judge whether the outcomes are good or bad. So right now, we are describing positive effects, meaning: these are the natural implications of these kinds of systems of representation. But of course, many people have normative feelings about them. They think: Well, if this pushes politics to the Right, we ought to go to a better system. Those who are on the Right would say: No, this is a great system; this is the way it should be. What are your thoughts on the Electoral College? Is there anything normative to say about it? It obviously encourages certain kinds of campaigning which affect how the national vote turns out. If it were a majority rule election people would campaign very differently than they do now. Guest: Yeah. My basic take on these things is I'm such a positivist on them--as a social scientist I'm always trying to understand how these differences affect outcomes. And of course every once in a while you see something that really strikes you as just unfair. I think Senate representation in the United States has that quality. I think it's hard to come up with a really good normative justification for Wyoming having two Senators and California having two Senators. And the Electoral College as well seems to be something that emerged from a set of negotiations that took place a long time ago, with issues in mind that are very different from the issues people have today. So, I can certainly see why the Electoral College might emerge as something that people are offended by. But it seems--these things are so difficult to reform. Mainly because someone always thinks they might benefit from the status quo. Russ: Yeah. Some do. Guest: But the interesting thing about it is that these things change over time. I told a story earlier about how the Liberals in the United Kingdom should have latched on to proportional representation earlier, and they could have been much better off. And I think interesting in the U.S. context, who latches on to what kind of electoral reform--these things change over time depending on who thinks they might have an advantage. For the Electoral College, lots of analysis has been done on this. It is just like we were describing in the case of the House and the Senate. It is a majoritarian, winner take all way of transforming preferences for a party into a legislative seat. In this it's just one seat--it's the executive. So the question is: is this thing biased in favor of one party or the other. And the answer seems to be: Not really. It changes. There are scenarios. If you go to the website, Nate Silver has situations in which Romney can lose the Electoral College but win the popular vote, and also some scenarios in which the same can happen for Obama. The same is true when we go back historically. In this urban geography that I'm describing, the states are large enough units that it doesn't really seem to affect the Electoral College. So, it's not systematically biased, every election clearly in one direction. As far as I can tell. So at least that aspect of it is not as troubling as one might make it out to be. Russ: And, like many things that the Founders did--which I think is a blessing, not a curse--they made it hard to change. I think it's not just that people are worried whether they would benefit in the future from the Electoral College. It's just that it's very costly to change it; and it gives you that inertia which I think is often a good thing, not a bad thing.
45:55Russ: As I said, we're in October in 2012 and we are going to be hearing a lot about red states and blue states--red America, blue America. And you have a paper called "Purple America." Is there anything in there you'd like to talk about in terms of generalizing the state of the electorate in the United States? Guest: Well, there's lots of work where we just talk at the individual level about voters. Because of the Electoral College and the importance of the Senate and the states, we talk a lot about states. What I've been doing in this project is trying to get lower than that and not just counties but even down to precincts, and really trying to understand how people are arranged in space, and try to understand the polarization that we hear so much about at the elite level, to try to understand what it looks like at the individual level. And so, my colleague has done a lot of work to try to show that individuals are not really very polarized; that a lot of people are kind of in the middle. And this kind of bimodal distribution that we see in the legislature is not really what we'd see in society. So, I've been trying to understand how that works in metro areas, in cities: where are the moderates? And as we discussed earlier, the suburbs are most of the purple places. The geography isn't always what it seems. But certainly the campaigns have a couple of different strategies that they can--we are accustomed to thinking about them as going after Independents in places where there are a lot of Independents. So, one way to try to win an election is to go to these purple suburbs and try to swing, change the minds of Independents. But of course another way is to go to places where there are lots and lots of your supporters and try to make sure as many of them as possible show up to vote. Russ: At least in a swing state. Guest: Exactly. Russ: In a safe state you don't really care so much. Guest: Right. So your strategy varies a lot depending on what kind of state it is. And so, using these kind of fine-grained, precinct-level maps you can start to learn about what kinds of neighborhoods--if you look at what the candidates are doing, what kinds of ads they are purchasing, you can kind of figure out what type of a strategy they are employing. Both parties seem to be working pretty hard to mobilize their base. But at the same time you've got to try to--it's a complicated set of tradeoffs to decide how much to work on the base and how much to work on these suburban moderates.
48:52Russ: I want to close with a conversation about, sort of about the normative side, but maybe that's the wrong way to describe it. I think a lot of people have a romance about majority rule. Certainly one way that small groups of people settle disputes is they say: Well, let's take a vote. And whatever gets the most votes wins. And I think to a lot of people that's obviously the fairest, best way to decide stuff. And so all of these things that we've been talking about that mitigate that--whether it's the Electoral College, winner take all districts--a lot of people say that's just not the right way to do things. Everything should be decided by a majority vote. And yet, as we know from work by Kenneth Arrow and others, majority vote in the normative sense, meaning leading to outcomes we like, isn't so strong as it seems. On the surface, nothing could be fairer than majority rule. And yet when you look a little closer you start to see that majority rule's got some very deep flaws in it. Talk about that argument. Why is it that majority rule is not the best system? Even though I think most people have that as a starting place; that's their default. Guest: Yeah. This is one of the things that when I teach courses to undergraduates on institutions, we do this in the first or second week. It's a very easy think you can do to have the students give their rank ordering of their preferences for what type of pizza that they would like; you have each student rank three and then you put them together. And it's very easy to find groups of students who have what in the social choice literature is called cycling majorities, where you can show that there is no such thing as the majority will. If I set up the institutions in such a way that there's first a round robin tournament of pepperoni versus vegetarian and then the winner of that is paired off against sausage, I can get a different outcome than if I do the initial pairings in another way. And so I can show that whoever controls the agenda controls what kind of pizza the students are having. It's kind of something that we've known since Condorcet and Arrow, the classics of social choice theory: it's simply nonsensical to say that the majority has some kind of will that we will then translate into policy. And so the students are always sort of surprised by this. We like to believe that there is such a thing as the collective will. And I think one of the basic lessons of politics and institutions is, unfortunately, it's possible to aggregate those preferences in very different ways in different institutions and get different outcomes. So we should attribute so much importance to something that we believe was the outcome of some kind of majority choice. Often the truth is much more complicated. Agenda control and political power are often used in getting us to the outcomes we see. It leads us to kind of think in a different way about how we interpret the decisions that are made by legislatures and what they actually mean. Russ: The other problem I have with "will of the people" is majority election. Whether it's 55-45, or 90-10, the loser obviously felt differently. So it's not the will of the people. It's will of those who won that election, whether it's a majority or whether it's proportional or whether it's this weird system we have in the United States. We don't have referenda on every item. It's this weird thing called the Legislature, Congress, Senate; we have committees; all this baggage, this incredible superstructure and infrastructure around the way political outcomes are coming out of our preferences. It's not just a majority rule referendum. And I think most people--there's a lot of problems with our political system--but I think most people think: Well, the best way to do it would obviously be a referendum because that would reflect the will of the people. And it doesn't, for all kinds of reasons. One of which is, as you said, the order of the voting can be manipulated. Information can be manipulated. There are a thousand things along the way. But the most important thing to me is that we all have different preferences. And so once you put it into a political process you are basically saying: We are going to get one outcome, and you are stuck with it--because it was the result of a vote. And I don't see that as necessarily fair at all. Guest: Well, right. So one other way to think about democracy: instead of thinking there is some will of the majority and we are trying to use democratic institutions to aggregate that and then turn it into policy, a very different way of thinking about democracy is: We're going to put some people in charge; and they are going to do some stuff. And after four years we are going to look backwards, retrospectively at what they've done, and decide whether we like it or not. And if we like it, we'll give them another chance. And if we don't, we'll dip into the pool of candidates and we'll pull out another one and see if they can do better. This is really a view of democracy that is more about accountability than about representativeness of some underlying will. But, as you already kind of described, in the United States it's becoming very difficult to do what I just described because we have a Senate, a House, lots of committees, a filibuster. We have the Executive, we have the Judiciary, and we have state governments which then have two chambers and an executive. And it's very hard for us to figure out who is responsible for the policies we see and then somehow hold them accountable. And so that's why, to some minds, kind of a very clear British style of accountability system would even be better. Again, that's a normative question. There's tradeoffs going both directions. Russ: As a--I'll make a normative comment. Because political decisions will struggle to reflect anything remotely like the will of the people, I want as few decisions as possible put into that sandbox. I'd rather have the competition of free association and free choice make those decisions and allow for the diversity of outcomes that private markets and private decisions have rather than political decisions, which are inevitably coercive. But that's my normative preference. Guest: Yep. Thinking about how institutions reflect the will of individuals--these are basic questions that people don't often think about. But they really should shape the way you answer these bigger normative questions about, say, how large should the state be? What kinds of things should the state do? Russ: Exactly. Well, we're almost out of time. Do you want to say anything about this upcoming election as a positive political scientist? Either at the national level--the Presidential, the House, or the Senate? Guest: Well, it seems likely that we will have more of the same in that we'll have divided government of some kind. That seems the most likely outcome. So, if we are hoping for a big resolution to all the uncertainty that we think characterizes the current moment, I suspect there won't be much uncertainty resolution. That said, it is something that will be exciting and interesting to watch, and there's a lot that still has to happen. It's certainly not a done deal in either direction.

More EconTalk Episodes