A few random thoughts on Detroit

I don’t have anything profound to say about the Detroit bankruptcy, but I’ll throw out one possible factor.  A few years ago I did a post discussing the striking population loss of large American cities after 1950.  The top 4 continued to do pretty well, but almost all of the rest of the top twenty dropped completely off the list.  Detroit barely hung on, dropping from #5 to #18, and Houston and San Francisco also stayed on the list.

Detroit isn’t unique, just bigger.  These cities also lost well over half their population since 1950:

St. Louis:  857,000 to 319,000,  Cleveland:  915,000 to 397,000, Buffalo: 580,000 to 261,000

All three now have fewer people than Mesa.  So why did Detroit go bankrupt?  I’m sure there are lots of reasons, but one might be that Detroit is in Michigan, and Michigan used to be very rich.  When I was young I saw a list showing that Michigan had the highest income in the nation.  I can’t find that list, but this link shows that in 1969 only Alaska, Hawaii, Maryland, Connecticut and New Jersey had higher incomes.  And all five are sort of special cases.  Alaska and Hawaii have a very high cost of living, and the other three picked up wealthy suburban spillover from NYC and DC.  (This was before the gentrification movement.)  So Michigan was the richest “normal state.”  It was richer than New York or California or Illinois or Massachusetts.  (OK you east coast intellectuals, I grew up in Wisconsin.  So I can define normal any way I wish.  A normal state looks like a mitten with a thumb on the right.)

Back in 1969 Michigan was governed by an ultra-liberal.  Someone named Romney.  Matt Yglesias once voted for his son, but the dad was far to the left of Yglesias.  Rich liberal states build up a costly overhang of government spending, taxes, high wages, high costs of unemployment insurance, union featherbedding rules, expensive pensions, etc.  That’s fine if the productive sectors keep churning out the wealth (think Wall Street, Silicon Valley, Harvard/MIT, etc) but if your economy tanks you’d actually have been better off with a more modest government, say Texas, or even an Indiana.  

I’m making a modest claim here.  The expensive overhang in Michigan helps explain why the Detroit area struggles even compared to other rustbelt cities.  I’m sure it’s not the only factor.  Alex Tabarrok has a related post.

It also might explain why upstate New York has done so poorly.  NYC is productive and appealing enough to survive NY state’s high taxes.  Upstate isn’t.  Places like Rochester used to be major centers of high tech—I kid you not.  They have no chance against Austin or Raleigh, or even against low tax cold/rainy states like New Hampshire and Washington.

PS.  Perhaps individual incomes in Michigan were higher in 1969–the link I found had household incomes.

PPS.  I have fond childhood memories of visiting rich Michigan to see my grandparents in the 1960s.

PPPS.  I’m a bit puzzled by the graph on unward mobility that Matt Yglesias and others are discussing.  I don’t have time to study the underlying research, so perhaps someone can help me.  I notice that the strongest upward mobility by far is the Great Plains and Utah areas.  And to a lesser extent the Kentucky/W. Virginia border area.  So that’s what?  Boom and bust farm area with low rainfall, fracking area, and coal country.  Is this really a story about economic opportunity, or is it about a casino-like economy where you can easily go from poor to rich and vice versa?  Or both (my initial view).  Are the studies based on annual income (volatile in commodity areas) or twenty year averages of income (less volatile in commodity regions.)

PPPPS.  As you know I’m a big fan of Yglesias, but unless this was a joke that went over my head, it seems out of line:

Detroit is everything that conservatives hate””labor unions, black people . . .

I hope Yglesias does not adopt the Noah Smith approach to conservatives.

Otherwise I agree with the post.



38 Responses to “A few random thoughts on Detroit”

  1. Gravatar of TravisV TravisV
    27. July 2013 at 06:29

    Prof. Sumner,

    I think Krugman’s latest is addressed directly to you:

    “Milton’s Paradise, Still Lost”


  2. Gravatar of Patrick R. Sullivan Patrick R. Sullivan
    27. July 2013 at 06:32

    I lived in NYC in the mid 70s when it went bankrupt in all but name. It was the same thing as in Detroit, in that massive numbers of people were believing in the tooth fairy; money would be left under your mattress…but you didn’t even have to lose a tooth to get it.

    Same thing in Illinois, California now too.

  3. Gravatar of Mike Sax Mike Sax
    27. July 2013 at 06:43

    I’m sorry but I have to ask: what is it that makes anyone assume conservatives love either unions or black people?

  4. Gravatar of Patrick R. Sullivan Patrick R. Sullivan
    27. July 2013 at 06:45

    I left this comment at Krugman’s blog;


    ‘Paradise’ to Friedman was monetary policy NOT conducted through interest rates. He complained about that even through the Greenspan years, all the while admitting Greenspan’s Fed had ‘found its sweet spot’.

    Seems to me that the recent unpleasantness vindicates Friedman on that score. Now his 1950-ish monetary aggregates scheme was rendered ineffective (if not impossible) by banking and finance innovations, but he couldn’t have been expected to foresee that.

    Were he alive today, he’d probably side with Scott Sumner and the Market Monetarists. That’s the side he was on in the latter years of his life with regard to Japan. Matter of fact, so was Paul Krugman.

    How long it will remain there is anybody’s guess.

  5. Gravatar of Patrick R. Sullivan Patrick R. Sullivan
    27. July 2013 at 06:46

    Mike, that’s known as the fallacy of the false dichotomy. In case you’re keeping track.

  6. Gravatar of TravisV TravisV
    27. July 2013 at 06:54

    New from Krugman:


    “there’s a danger in the Stiglitzian approach, namely that people might conclude that fixing the short-run shortfall in demand must wait until we fix the long-run problem of inequality, which is going to be very hard and a long time coming. We need stimulus, or at least an end to austerity, now, even if restoring a middle-class society isn’t going to happen any time soon.”

    I strongly agree with Krugman about this. I’m just not sure that fiscal stimulus would help much.

  7. Gravatar of TravisV TravisV
    27. July 2013 at 06:56

    Patrick R. Sullivan,

    Good point about Friedman and interest rates. It would be useful to see quotes from Friedman on that subject throughout the years (not just applying to Japan).

  8. Gravatar of Steve Steve
    27. July 2013 at 07:49

    “I’m a bit puzzled by the graph on unward mobility that Matt Yglesias and others are discussing. ”

    I think the key is that they use quintiles based on NATIONAL income, rather than local income. So if a county has 50% of people in the bottom quintile and 5% in the top quintile nationally, even if the county has perfect income mobility locally, only 5% of people make it from bottom to top. If it starts poor, it ends poor, therefore there’s “low income mobility.”

    Here’s a quote from the paper:
    When analyzing local area variation, we continue to rank both children and parents based on their
    positions in the national income distribution. Hence, our statistics measure how well children do relative
    to those in the nation as a whole rather than those in their own particular community.


  9. Gravatar of Steve Steve
    27. July 2013 at 07:54

    If you applied the same methodology globally, you would find that the Persian Gulf is a bastion of income mobility! But those countries with black people (Africa) have no income mobility.

    It’s actually surprising that this study didn’t find more income mobility in San Francisco or New York, given that the cities are wealthy. The odds of making it to the top quintile should be higher than they are, given the relative wealth of those cities.

    Yes, the plains states prospered due to recovery in resources. But since the reference period was 1996-2000 (a resource bust), that shouldn’t be surprising. And I don’t see why those economies are more “casino-like” than, say, the stock option economy of San Francisco.

  10. Gravatar of dtoh dtoh
    27. July 2013 at 08:11

    From the founding of the Republic up until the oil shocks, Michigan was one of the richest if not the richest states/region in the country. Astor’s company, the American Fur Company, was probably at one point the largest company in the world and had many of its operations centered in Michigan. Subsequent growth of the copper, iron and lumber industries continued to fuel the growth. Annual freight volume carried through the Soo Locks was higher than the Suez and Panama canals combined. Much of the capital driving the auto industry came from profits generated in the lumber industry.

    During most of the 1900’s Michigan (not California) was the go to place for ambitious entrepreneurs.

    Michigan’s economic decline had little to do with poor governance and high tax rates, but rather, was a result of over-dependence on the auto industry, which became uncompetitive as a result of high union driven wages. Not only Detroit, but much of the state was dependent on the auto companies and it’s suppliers. Benton Harbor whose economy was dominated by foundries supplying the auto industry now has a per capita income on a PPP basis which is lower than that of China.

  11. Gravatar of ssumner ssumner
    27. July 2013 at 08:40

    Travis, Job growth accelerates in 2013 and he still thinks fiscal policy is more important than monetary policy? Someone is lost.

    Patrick, Good observation. Even NYC had to reform somewhat to get back on its feet in the 1980s, although it’s still an inefficient public sector.

    Mike, That’s an odd question–is it somehow related to something I wrote?

    Patrick, Exactly right about interest rate targeting.

    Travis, Yes, the other Krugman post was very good.

    Steve, Did they really do that? That’s just bizarre. Someone with more time than me should study the paper and do a post.

    dtoh, Thanks, but haven’t the Japanese opened car factories in states like Indiana and Ohio? So why not Michigan? The US. auto industry has not declined, it’s just as robust as ever. The Michigan auto industry has declined.

  12. Gravatar of Liberal Roman Liberal Roman
    27. July 2013 at 08:53

    OT: Sometimes I think Krugman writes blog posts specifically in order to drive Sumner into pure seething rage:


    “Robert Skidelsky has an interesting note asking why quantitative easing has received so much stress in recent economic debates. Its effectiveness is, after all, questionable “” whereas there is overwhelming evidence that fiscal policy works as advertised.”


  13. Gravatar of Liberal Roman Liberal Roman
    27. July 2013 at 08:56

    Actually as I read the rest of this post, this is one of the most direct attacks Krugman has written at Sumner without naming Sumner.

    It of course makes no sense. I can’t wait to see the inevitable Sumner post ripping it to shreds and then the inevitable Krugman non-response.

  14. Gravatar of jknarr jknarr
    27. July 2013 at 09:03

    Common denominator: serial governance by Democrats. More broadly, lack of political competition creates lack of adaptation.

    Serial Republican governance likely carries its own problems, but bankruptcy typically ain’t one of them.

  15. Gravatar of TravisV TravisV
    27. July 2013 at 09:31

    Mark Thoma on Krugman’s new post on QE:


  16. Gravatar of jknarr jknarr
    27. July 2013 at 09:49

    Reading “Milton’s paradise”‘ I have to ask whether PK understands the mechanisms of monetary policy at all. In fact, I assert that he is ignorant.

    The Fed Funds interest rate is only useful in how it guides the formation of base money. Fed funds targets are set near the Wicksellian natural rate – recall that , before targeting, markets had to watch monetary base operations for a sense of where the Fed wanted the rate. If the Fed wants to expand base money, it sets the target rate a bit lower than the Wicksellian – and it finds out via the demand for base money.

    (Very likely, approaching the proper zero bound – not 25bps – the Fed would be forced to provide reserves at a asymptotIcally increasing rate as it closed on absolute zero Fed Funds – note that demand for cash would be a crucial component of this reserve-printing treadmill: demand for currency would drain reserves. Currency demand is a feature, not a bug.)

    PK makes a fetish of interest rates, and ignores the only important intervening variable to monetary policy: the monetary base.

  17. Gravatar of jknarr jknarr
    27. July 2013 at 10:17

    Supporting the asymptotic creation of base money as rates approach zero – the base money move from 25bp to zero would be explosive, and multiples of the move from 100-25bp. PK would get much, much more than a “couldn’t hurt” outcome.


  18. Gravatar of Matt Waters Matt Waters
    27. July 2013 at 10:36

    I always point to Michigan when people make arguments that boil down to “hey, we have a good thing going, don’t ruin it.” These arguments are in response to free trade, immigration, unionization, automation, and occupational barriers to entry, but they’re all basically the same. If somebody happens to be on the right side of a government distortion, they just want the world around them to be set a stone, with no dynamism whatsoever.

    In general, this is great for the people involved. Michigan was one of the richest states in the 50’s and 60’s, after all. But their products were crap and when Japanese competition came in, they responded with lobbying for protective tariffs vs. improving their products. The lack of dynamism ultimately catches up every time and forking standards to improve, things can of be set in stone.

  19. Gravatar of JVM JVM
    27. July 2013 at 10:37

    “Detroit is everything that conservatives hate””labor unions, black people”

    Honestly I’m sorry to say that outside the well-educated blogosphere (and maybe the coastal US) this is fairly mainstream conservative opinion.

    My attitude about Republicans generally is that their labor policy seems to reflect some vague understanding of the supply side, but I can’t really handle their constant veiled racism (I guess in addition to their love of ecologically destructive externalities and un-utilitarian policy preferences, but of course the dems have similar problems).

  20. Gravatar of Matt Waters Matt Waters
    27. July 2013 at 10:42

    Both Krugman’s and Thoma’s post are extremely shallow, just as they always have been on QE. As they get pushed more on specifics, they inevitably confuse the theoretical and political. They say that in THEORY QE couldn’t work at the zero bound because POLITICALLY the Fed would tighten too quickly. This is just like me arguing that fiscal stimulus could never work theoretically because it would be impossible politically, which is of course complete nonsense.

    I mean, I enjoy reading Krugman and his prose can be great, but his avoidance of answering direct questions on monetary policy can only mean that he’s staked his claim so deeply on fiscal stimulus that he cannot change his mind.

  21. Gravatar of Richard Davis Richard Davis
    27. July 2013 at 10:47

    Metro Detroit has never lost population. I mean Wayne county + Oakland county + Windsor. Detroit city limits are arbitrary lines. This was just folks moving from one neighborhood to another one nearby. It also happened to my hometown Macon, GA in the 60’s. Now everyone lives one county to the south.

  22. Gravatar of McArdle on Detroit and unwisdom from a Vermonter | Economic Sophisms McArdle on Detroit and unwisdom from a Vermonter | Economic Sophisms
    27. July 2013 at 10:50

    […] Scott Sumner had some thoughts on Detroit (what!). This reminds me of an article Megan McArdle had on Bloomberg News this week. McArdle’s article is worth reading and I’ll trim down a quote: […]

  23. Gravatar of Steve Steve
    27. July 2013 at 10:58

    Liberal Roman wrote: “this is one of the most direct attacks Krugman has written at Sumner without naming Sumner.”

    Krugman almost never mentions Sumner by name, because he is afraid his readers might choose to educate themselves.

  24. Gravatar of marcus nunes marcus nunes
    27. July 2013 at 11:23

    TravisV & Liberal Roman
    It can easily be shown that Krugman is wrong. If that was an “attack”, it was a pretty lame one

  25. Gravatar of Patrick R. Sullivan Patrick R. Sullivan
    27. July 2013 at 12:26

    ‘Sometimes I think Krugman writes blog posts specifically in order to drive Sumner into pure seething rage’

    If so, he’s been a miserable failure.

  26. Gravatar of ssumner ssumner
    27. July 2013 at 12:32

    Liberal, In the US, in 2013, the view that fiscal stimulus works just seems pathetic.

    JVM, You look young in the picture. I hope as you get older you’ll go out and meet actual conservative people in the middle of the country. Yes, there are plenty of conservative racists, but there are also plenty of liberal racists.

    Matt, Good point.

  27. Gravatar of Patrick R. Sullivan Patrick R. Sullivan
    27. July 2013 at 12:32

    As I predicted, my comment at Krugman’s blog is gone. If you can’t stand the heat, delete the comment.

  28. Gravatar of Steve Steve
    27. July 2013 at 12:39

    “Steve, Did they really do that? That’s just bizarre. Someone with more time than me should study the paper and do a post.”

    Scott, it would be a huge project to figure out what they actually did since the original IRS data isn’t available, and the documentation is hard to decipher.

    However, I took a look at a few of the spreadsheets and noticed a few interesting patterns.

    For one thing there’s a strong negative correlation between commuter zone population (log(CZ population)) and income mobility. (R^2 27%, n=727) The bigger the city, the less the income mobility.

    The effect becomes even stronger if you sort by state and look at income mobility in different cities within a state. For example:

    Milwaukee = low mobility
    Eau Claire = high mobility

    Philadelphia = low mobility
    Altoona = high mobility

    New Mexico:
    Albuquerque = low mobility
    Roswell = high mobility

    The effect breaks down with a handful of “super-cities”, but I think this is a statistical artifact as well. Miami, New York, Boston, Washington, and San Francisco do fairly well in income mobility. But these Commuter Zones were all designed to include the very wealthy suburbs, including Hollywood/Fort Lauderdale, Westchester, Newton/Wellesley, Alexandria and Silver Spring, and basically the whole SF peninsula. So we don’t know if the mobility is being driven by the wealthy suburbs (likely) or the poor urban areas (unlikely).

    The result for Massachusetts is:
    Springfield = low mobility
    Boston (super-metro) = medium mobility
    Nantucket = high mobility

    Also instructive: Dallas has very low mobility, but Fort Worth has average mobility. I wonder how far Miami’s good “mobility” would drop if you split off the suburbs around Fort Lauderdale and Palm Beach.

    They don’t have CZ income data available, but I bet there would be a super-high correlation between average income and “mobility” they way they calculate it.

  29. Gravatar of dtoh dtoh
    27. July 2013 at 14:35

    Well like I said…unions. I had a discussion once with the guy who oversaw all the overseas planning for Toyota. He said, Michigan has the best workers of any state in the U.S. but Toyota would never build a plant in Michigan because of the unions.

  30. Gravatar of Negation of Ideology Negation of Ideology
    27. July 2013 at 17:19

    Mike and JVM –

    “I’m sorry but I have to ask: what is it that makes anyone assume conservatives love either unions or black people?”

    “I can’t really handle their constant veiled racism”

    I think these statements say more about you two than they do about conservatives. The problem with ad hominem attacks like that is they prove nothing. Let’s assume for a minute that all conservatives are against piling on huge debts and unfunded liabilities because they hate black people. Does that mean we should do those things?

    It’s wiser to address actual arguments than to try to divine motivations, especially of large groups.

  31. Gravatar of Bill Ellis Bill Ellis
    27. July 2013 at 19:37

    Is anyone really making the claim that Racism is equally distributed along the liberal /conservative axis?

    Wouldn’t that be odd if it was?

  32. Gravatar of jknarr jknarr
    27. July 2013 at 19:47


  33. Gravatar of Benjamin Cole Benjamin Cole
    27. July 2013 at 21:36

    Detroit used to be the richest big city on Earth, 1950-60s.

    Rarely mentioned today is that the federal government sucked billions upon billions of dollars out of high-income regions and states, and redistributed to rural areas, beginning post WWII and continuing on. This goes on today, as each rural state has two Senators.

    North Dakota enjoys net–net!–federal spending (in excess of revenues) of $20k per capita. Per capita!

    What if Detroit had net per capita spending of $20K? Every year?



  34. Gravatar of Steve Steve
    27. July 2013 at 23:22

    I took another look at the Chetty Saez et al research. Their website is http://www.equality-of-opportunity.org/index.php/data

    They focus on a cohort born in 1980-1981, and use tax return data from 1996 to measure parent income, then tax data from 2010-2011 to measure child income at age 30.

    They appear to characterize parent income based on NATIONAL percentile, and then child income based on COMMUTER ZONE income. Their preferred measure (based on graphics) is the expected commuter zone outcome for children whose parents were in the 25th percentile NATIONALLY. The reason they are doing this is to defend the EITC as well as progressive local income taxation.

    The general pattern that emerges is that the east, especially the south and the rust belt, has low income mobility, while the west has high income mobility.

    The west and the east both have high ABSOLUTE tax expenditure, so they focus instead on PROGRESSIVITY of taxes (where the southeast is low), and they conclude that income mobility is robustly related to progressive local taxation (thanks larges to the mobility of California).

    Their graphics focus on expected percentile outcome for someone starting at 25% nationally. The results range from ~35th percentile in Memphis/Charlotte to ~46th percentile in Salt Lake/Reading. Their graphics show a clear pattern of South and Rust Belt having lower income mobility.

    My thoughts:
    The studies are based on comparing 1996 parent income with 2010 child income. So to some extent, the graphic simply reflects which areas of the country (resources) improved over those 14 years, and which got worse (autos and manufacturing).

    There’s also a clear element of race in the study. There’s definitely some relationship between low mobility and black population, especially in the south.

    But there’s also some cherry picking. It’s interesting that overall tax burden didn’t predict much in the way of mobility, so Chetty, Saez, et al, focused on progressivity. California has progressivity and uncome mobility, so that outlier drove the national results.

    Leonhardt focused on quintile-quintile probabilities, rather than the percentile expectations used by Chetty Saez. That’s probably because the South looks especially bad on quintile-quintile, while the entire east looks bad on percentile expectations.

    Meanwhile, Krugman created a “sprawl” metric that mismatched the Chetty commuter zone metric with the Census metropolitan zone. Krugman’s top ten metropolitan zones included Miami (#11), Atlanta (#12), and Dallas (#14) as top ten, while excluding three of the actual top ten. This enabled Krugman to falsely claim that sprawl is inversely related to mobility.

    NOTE: I’m still a little confused about which Chetty Saez percentiles were national and which were commuter zone. I currently believe parents were national and children were commuter zone (because the EITC is set based on national income, and they wanted to measure local outcome). But they could have done a better job clarifying this.

  35. Gravatar of JLK JLK
    28. July 2013 at 06:47

    Regarding Buffalo/Rochester/upstate NY, Ed Glaeser has a different take, both in a JPE article and this City Journal narrative piece:http://www.city-journal.org/html/17_4_buffalo_ny.html
    His story is one of technological change that favors some endowments over others, bad weather, and above all the way decline begets decline. (Reaching a low-state equilibrium doesn’t need to happen quickly, even when markets are efficient.) So high tax rates of NY state don’t necessarily enter into it.
    I believe the excess-infrastructure-is-expensive story, where a one-time negative shock creates an excess burden on a region that make it less competitive absent other factors. So a Detroit race riot (not necessarily unjustified given then-prevailing conditions) can induce a death spiral; bad weather, maturing baby boomer union populations, Japanese economic reconstruction, etc all pile on.

  36. Gravatar of ssumner ssumner
    28. July 2013 at 08:05

    Thanks Steve. That’s very helpful–I wish I had more time.

    dtoh, That’s part of it.

    Bill, No, but I think it’s much closer than most people assume.

    jknarr, Thanks, Bill really should look at the Tabarrok post you link to–it will open his eyes.

    JLK, I agree that many factors were involved.

  37. Gravatar of chris mahoney chris mahoney
    28. July 2013 at 13:37

    George Romney was a piker compared to Nelson Rockefeller, the worst governor in American history. He also turned the NY GOP into a corrupt enabler of the Welfare State.

  38. Gravatar of Bob Bob
    29. July 2013 at 09:38

    St Louis isn’t bankrupt, but it still has Democrats in the mayor’s office, and a major hollowing out of the city.

    What you get instead is a lot of small municipalities that, using eminent domain and zoning laws, do their best to keep their population homogeneous. Said municipalities are so small you can cross four of them in a 10 minute trip to a major supermarket, and do their best to steal retail from one another to capture the sales tax generated by people living in other municipalities. You get situations like having unincorporated subdivisions that none of the neighboring municipalities would want to take in, because they do not seem profitable enough.

    It’s a model that leads to major segregation and a high chance of financial trouble for municipalities. Remove the stadiums and the night district, and the city is reduced to a few commercial skyscrapers and a lot of poor people who are close enough to work at your local retailer, but can never move within 5 miles of you.

    The city is then stuck trying to bribe companies to move there, but even then, where do the poor go? They have to live somewhere, and that municipality will always get more costs and less taxes.

Leave a Reply