Sunday, December 08, 2013

The Great Labor Dump

One of my favorite teachers at Michigan was a self-effacing but brilliant labor economist named Mike Elsby. Mike has sadly decamped for Scotland, but he continues to turn out excellent papers. One of his latest, a Brookings conference paper, takes on a hugely important question that has been on everyone's minds of late: Why has labor's share of income declined in advanced countries?

There are basically three competing stories. These are:

1. Robots. Technology has made it cheap to replace humans with automation, driving profits up and wages down.

2. Unions. The relentless decline of organized labor in rich countries has robbed workers of their bargaining power, causing owners to scoop up the surpluses.

3. China/India. The end of the Cold War caused a huge, relatively well-educated labor force to be suddenly dumped onto world markets. A glut of labor on the market means the return to labor goes down and the return to capital goes up.

Mike Elsby and his coauthors find support mostly for story #3. In their own words:
U.S. data provide limited support for...explanations based on the substitution of capital for (unskilled) labor to exploit technical change embodied in new capital goods...[I]nstitutional explanations based on the decline in unionization also receive weak support...[W]e provide evidence that highlights the offshoring of the labor intensive component of the U.S. supply chain as a leading potential explanation of the decline in the U.S. labor share over the past 25 years. (emphasis mine)
This finding disagrees with some other recent papers, such as Karabarbounis and Neiman (2013), who support the "robots" story. Read Elsby et al. to see the particulars of the argument. 

Anyway, this is an important piece of research that everyone should know about. Stories about the Rise of the Robots are scary and seductive, but if Elsby is right, then this is more of a concern for the future than the past. And the decline of unions is sad in many ways, but it seems to be a symptom of labor's decline rather than a cause.

So what do we do to bring back labor's share of income? We wait. The Great Labor Dump can only happen once. When it's over - when insanely huge amounts of investment in China have saturated that country with capital, for example - labor's share will bounce back.

...unless, of course, by that time we really are facing the Rise of the Robots.


  1. interesting. I just came across this paper which argues for #2 (well, politics/institutions more generally).

    Their conclusion being mainly based on cross-country comparisons. #1 and #3 don't really explain why it's been much worse for labor in the U.S.

    For my own part, I would find it hard to believe that any of these explanations had no effect.

  2. Seems to me any explanation for labor's declining share has to deal with the fact that real interest rates also seem to have fallen dramatically. (I'll note in passing, one explanation is that real interest rates haven't actually fallen if you choose the dates properly: compared to late 70's, when labor's share was much higher, real interest rates are about the same today. But I don't think that explanation survives the accusation of cherry-picking.) On the face of it, both the "robots" and "China" explanations would suggest rising capital returns and therefore rising interest rates. I'm not sure about the "unions" explanation: arguably it's about how you split up monopoly rents, so it might be easier to reconcile with falling competitive returns. Larry Summers suggests a theory that might make the "robots" explanation compatible with falling rates: capital is getting much cheaper to produce and therefore it takes a lot more investment (if you deflate investment and consumption separately) to use up a given amount of savings. (Equivalently, if you use the same deflator for investment and consumption, the production process now uses a lot less capital than it used to.) Is there an explanation that makes the "China" theory compatible with low interest rates? Maybe increasing costs of adjustment make it difficult to mobilize the necessary investment, so a lot of savings are idle waiting to be invested?

    1. The real interest rate is not the return on capital.

    2. It's the return on a certain kind of capital. Generally we would expect the real interest rate and the overall return on capital to move in the same direction over the space of a few decades. If the two are moving in opposite directions (which may or may not be the case), then we need an explanation.

  3. Would like to add
    The crowding-out effect of housing-price appreciation
    It appears that in reaction to housing-price appreciations, banks increase real estate lending and decrease commercial lending.

    Monetary policy and Chinese investment glut vs. US. investment dearth:
    Low U.S. interest rates found their way into portfolio and foreign direct investments in Emerging Markets and accentuated #3
    China 2000/2010 had yearly money supply increases of 15% and more compared to Germany in 1960/1970 of around 10%.
    Chinese investment glut vs. US. investment dearth during the decade: Chinese were able to adjust to rising labor costs with investments, via point #1, robots

    #6: Digital revolution
    Services can be provided globally, again point #3, in particular Indian programmers....

    #7: Globalization of supply chains
    American, German, Swiss firms can sell under their (quality) brand name and increase profits, but a big part of their supply chain is global.

  4. Anonymous1:59 PM

    This was always such an obvious conclusion that I doubt the honesty of all the economists (and neoliberals, and DLCers, etc) who claimed they had good reasons this wouldn't happen. At the very least, the economic community not only missing this prediction but resisting the conclusion for years after the evidence had been in has to be considered an enormous failure.

    1. Anonymous1:23 PM

      Part of most left-of-center free trade prescriptions was an assumed transfer payment from the many benefited by the freeing of trade to the few harmed by this. It is hardly the fault of the theory that this has, to put it mildly, been opposed rather energetically (with the opposition ironically including a large number of those negatively impacted by the trade liberalization, e.g. southern former clothing mill workers).

    2. " It is hardly the fault of the theory that this has, to put it mildly,...."

      It is indeed the fault of the theory, and the advocates, because it was a huge assumption, and had to come *after* certain groups profited (meaning among other things that their political clout increased).

  5. Ok:
    1) Volker's put raised price of dollar and put enough people out of work to weaken industrial labor unions for decades
    2) US tax policy: no taxes on imports, no taxes on unrepatriated profits, no real enforcement of transfer pricing - vast subsidy of outsourcing
    3) US tax policy - permit management to keep casino profits from massive liquidation of corporate assets.
    4) Ideological hostility of management to labor, far beyond what is motivated by corporate profit
    5) transformation of securities market into one dominated by short term gains and 3rd party management of OPM - with casino winnings for managers who play the game right - see tax policy above
    6) Smart government policies in India and China to protect infant industries and engage in massive import substitution/export development. Note difference between development of industry in India/China versus Mexico where government is too weak to protect local capital.

    And yet, for economists, all this fell from the sky and all we can do is wait. Because, e.g. high dollar and lack of US infrastructure/industrial investment bank is some natural property of laws of economics, not a political choice.

    1. These points are all very, very good, but I'll emphasize point 1 ('Volker's put raised price of dollar and put enough people out of work to weaken industrial labor unions for decades'). This was deliberate, choking the sh*t out of the economy until the bargain power of labor was largely gone.

  6. @root_e
    ad 4) I suggest reading Bawerk's about Okun's law
    Since 2010 employment is rising more quickly than real GDP, but not wages.
    In times of high salary rises e.g. 1996-2000 or 2006/2007, GDP was rising more quickly.
    Since 2010 I see a sort of Japanization, steps towards full employment, but no salary hikes

  7. Anonymous3:30 PM

    Whatever your explanation it has to fit the facts that (1) the fall in the labor share is pretty much global, (2) the US actually has a lesser trend than countries like Norway and Austria where the labor market wasn't deregulated as much, and crucially (3) the trend started in the mid 1980s which is way too early for it to be China and India. The IMF had a paper out in about 2007 suggesting trade openness was a factor but there was also a BIS paper around the same time (on the profit share not the labor share, but reverse the sign if the coefficients and it's the same thing) around the same time that showed global results supporting a technological explanation. As far as I'm aware it's still the only paper that fits all of 1-3 above.

  8. "ad (1) the fall in the labor share is pretty much global"
    is evidently wrong, in developing countries, the labor share is rising
    (3) The trend started in mid 1980s without China. This was the factor #2, the crack down on unions via neoliberal ideas and higher unemployment during that period

  9. IMHO, the problem lies in mis-pricing of capital. As I will describe below, if you mis-price the return on capital, you cause and provide the added incentive to underutilize capital. As the employment of labour is a function of employing capital, then you can under-employ labour over some period of time.

    To explain: in simplest terms say you replace a line of humans building cars with a line of robots. The cars still have to be sold to humans - for now. So theoretically, the human that picks up a gum wrapper from the ground should be credited with "work", even as a re-employed human, and receive some compensation in the form of a car (or share thereof).

    That is unless you "pay" the capital holder to not produce in the event that they don't care to exchange for the value of garbage-picking in the above example. Say they financed the replacement of labour with capital using debt. You can pay them to not produce by relieving some portion of their debt in spite of not producing cars. If they don't have to produce to pay principle and interest, they will expect a different return than something they don't value - like environmental rehabilitation - before they will exchange for a car.

    Now, let's look at a real world example: Heinz announced closing a plant in Leamington, Ontario. The return from that plant obviously wasn't sufficient. Generally, three inputs must be considered, taking value of output as given. Labour and materials are two and if you drop the cost of these inputs the plant value (this being capital, the third input) must rise.

    If instead you lower the price of capital, then the plant must also rise in real value? Not really. You didn't change the real output of the plant. If you were making 1 million bottles of ketchup per year before, this is still the same even with lower capital cost and higher plant value. If I originally purchased the plant for $1 million in an environment of 10% interest rates and you drop the interest rate to 5%, then notionally the plant has (more than?) doubled in money value - thank you very much.

    This is the thing: the "capital" has gone up in value and yet I am no more incented to employ it than I was before. In fact, the money I would make is then less valued. I am less and less compensated to employ the plant even as the value of the plant goes up itself. However, if you raised the interest rate, you would be forcing the value of the plant down to a point where making 1 million bottles of ketchup (and therein covering your variable costs of labour and materials) would be worthwhile. If I was happy to invest $1 million to receive the value of 1 million bottles of ketchup per year before, why should that be any different today, all else the same of course.

    This is nothing new. Adam Smith talked about the difficulty, as interest rates go lower, to get the capital value out. We grow out of high interest rate environments because money is paid to go to work, and we stagnate into low interest rate environments because money is not paid to work. Neither of these is sustainable. Ultimately sometime in the future, it will get to a point and then we'll see how much of a wave the canonballer (fodder?) makes.

  10. Anonymous4:32 PM

    Your also forgetting, aging population. By the mid-80's, the Boomer stopped flowing into the labor force and that "show" is evident.

    1. Less labor supply means lower wage increases/lower labor share?
      Should be rather the opposite..

    2. Anonymous4:44 PM

      I am talking about labor force calculations. Though you have less labor coming in, you have less growth as well. The other reasons have been mentioned including public disinvestment that has been going on since 1973.

  11. Question: couldn't we be waiting for a long time, given the rise of African labor as the cost of Chinese labor goes up? Africa has a lonnnnnng way to go to catch China (meaning the world could lean on African labor for a long time) unless they skip ahead like they already have with personal tech use.

    1. In addition, we see strong signs in most of the developed world[1] that the elites have been using their increased gains to improve their political power. The end result of the waiting game (esp. with Africa and S. America coming 'online' over the next few decades) is a developed world with a multi-generational system, which should be very locked in. The majority of people will have poor education, high debt, high labor market risk, and the laws and systems structured against them.

    2. Oh, in addition 'wait' is what the - well, liars - of neoliberalism have been telling us for four decades now. The rising tide is drowning most people.

  12. Low cost and hardworking immigrants (USA)?

  13. Bill Ellis8:55 PM

    Noah, I think you are not giving number two, the decline of Unions, enough weight. I think you framed it too narrowly: "The relentless decline of organized labor in rich countries has robbed workers of their bargaining power, causing owners to scoop up the surpluses." Strong Unions would have blunted the effects of number three--China/India.

    If increased Inequity in the west is because the owners of the means of production have benefited from owning lower cost means of production in the third world, allowing them to keep a bigger share of the profits without having to share them with the workers---what allows this to happen ? It's Libertarian morality and the economic convention of property rights. And they have gone unquestioned.

    Strong unions would have rejected these conventions.

    I find it interesting that when you and other economists talk of the rise of the machines and how they one day may take our jobs---that then we would be required to reject these conventions and, as you put it in your last jobs and robots post..."resort to some mass redistribution of income"* Why is it OK to reject these hallowed conventions when it is Robots taking jobs but not when it is third world labor? Why is it morally right that American CEO's see 250% income growth off the backs of third world labor while the American worker's wages are near stagnant ?

    * (Your way is in effect, taxing the owners of robots to pay for subsidies, to pay the owners of the robots, to get them to hire superfluous real people. It seems kinda crazy to me. )

    We will also probably need to resort to some mass redistribution of income. The best way to do this is to have the government subsidize wages, i.e. to pay companies to hire humans and pay the humans more. This is like putting our thumb on the scale against the robots.

    1. Nate O8:03 AM

      I agree about unions generally, but it extends to corporate governance as well. Large German firms are required have 1/2 shareholder- and 1/2 employee-elected members of their supervisory boards. As an executive, it's harder to go to your board in Germany and say "I want to outsource production and give myself a raise."

    2. Nathanael4:28 PM

      Why do economists treat mass redistribution of income as some sort of "last resort"? Is it because they're employed by the rich?

      It should be the first resort, and it's called "taxation". It's a good thing.

  14. Melissa makes a really good point -- after China, there's Africa. Nigeria, for instance, now has a population of 180 million, but it's estimated that by century's end, that country's population will rise to 900 million. Maybe, just maybe, the continent will hold down birthrates -- but most of what we'd like to see in Africa -- an end to warlords. modernization, better agriculture ands medicine, industrialization, etc. -- is going to send population skywards. Labor outsourcing lies ahead of us for a century or more.

  15. Production functions are used in the analysis? Really?? CCC debates still have no purchase on Brookings????? WTF.

    Let's not talk about *power* some more!

  16. There's a competing paper arguing that the main driver of the fall in the labor share has been the slower growth in the relative price of investment goods:

    "We demonstrate how the decline of the labor share can be explained by the decline in the relative price of investment goods. Efficiency gains in capital producing sectors, often attributed to advances in information technology and the computer age, induced firms to shift away from labor and toward capital to such a large extent that the labor share of income declined.

    Of the 59 countries with at least 15 years of data between 1975 and 2012, 42 exhibited downward trends in their labor shares. Of the trend estimates that are statistically significant, 37 are negative while only 9 are positive. We complement our analysis with industry-level data and show that six of the ten major industries experienced significant labor share declines while only two experienced the opposite. Most of the global decline in the labor share is attributable to within-industry changes rather than to changes in industrial composition.

    This decomposition, together with the fact that labor-abundant countries such as China, India, and Mexico also experienced significant declines in their labor shares, argues against a simple role for international trade or outsourcing in explaining labor share declines in capital-abundant countries such as the United States."

    1. The decline in the relative price of investment goods has everything to do with the openings of China and India. Both countries had severely outdated production systems, not just cheap labor (Remember the Great Leap Forward and backyard ironmaking?). You can reap large productivity increases inexpensively by upgrading very old production technology to merely old technology. Meanwhile, past high wage rates in the advanced economies justified expensive investments in leading-edge (at the time) labor-saving technology; further productivity improvement would require even more expensive technology. Technology at the leading edge hasn't become cheaper, but it's less needed when wages are low; trailing-edge tech is enough.

  17. So, as noted, the 3 stories can be true at the same time. Then, it's a question for econometricians to tell us whether they think they can tease relative weighting out of the data with any kind of confidence.

    But that's why I am in the distinct minority of economist aficionados not liking free-trade/immigration:

  18. Anonymous8:34 AM

    "Mike has sadly decamped for Scotland,..."

    Was Mike sad when he decamped? Or do you mean, "Sadly, Mike has decamped..."?

    Standards in today's PhD programs are shockingly low.

  19. Anonymous10:38 AM

    We can add another factor:

    #4, dumping the work, unpaid, onto the customers, in so, so many ways.

    -- Shopping online requires people have net access and spend their own time researching, ordering, arranging financing, and so on.
    -- banking and other services have drastically cut teller and other human supports.
    -- self serve checkout lines in stores, where you do the cashier's job but don't get paid for it

    I am sure we can all add to the list. Everywhere I spot this offloading I try my best to circumvent it, but it's an uphill battle.

    Noni Mausa

  20. Anonymous11:13 AM

    Not that it matters much, but #1 and #3 are the same. The rise of robots begins with an interconnected world that allows labor supply to be global.

  21. I haven't read the paper yet but, seriously Noah, you have to be skeptical of any analysis that isn't cross-national. There has been a global trend in the declining labor share, but it has been uneven. A satisfactory explanation has to fit the whole pattern.

    Incidentally, there are two related hypotheses that economists ignore but get a lot of attention from management and employment relations people. One is the role of the finance revolution, private equity, etc. (See for instance The other is the change in the internal organization of firms away from managing how work is done (which requires motivation) and toward precise specification targets for intermediate products (which doesn't), i.e. from input to output based control. (There are different names for this in the mgmt literature.)

    Of course, like other posters have said, these hypotheses are not mutually exclusive, and they probably interact in various ways.

    1. Nathanael4:09 PM

      The tax law changes are, embarassingly, pretty much global. The few countries which kept confiscatory taxes on the rich have NOT had the same economic problems.

  22. Some questions about this article:
    Elsby says "[W]e provide evidence that highlights the offshoring of the labor intensive component of the U.S. supply chain as a leading potential explanation of the decline in the U.S. labor share over the past 25 years."
    Why did this continue so long if the US pays Chinese manufactures in USD and Chinese workers are paid in CNY? That is why has China agreed to take so many USD over so many years? What do they plan to do with all these USD? Can they spend them anywhere else than the US? Should the US allow this trade imbalance to continue given the adverse effects to the US economy and workforce? I.e. should settlement be insisted on in the short term?
    Noah says "So what do we do to bring back labor's share of income? We wait. The Great Labor Dump can only happen once. When it's over - when insanely huge amounts of investment in China have saturated that country with capital, for example - labor's share will bounce back."
    Does this imply that we wait for China to reach full employment at using state-of-the-art technology with productivity that matches the west? If this employment growth has been going on for decades, how much longer will it continue? What are the long term structural effects for the US that occur due to waiting?

    1. You seem dismissive of the "robot" story. There's more to it than:

      "Technology has made it cheap to replace humans with automation, driving profits up and wages down."

      David Autor takes the technical change story quite seriously. His story is that technical change "hollowed out" the labor force, decreasing demand for middle-skill workers - e.g. typists and people doing clerical work. But the demand for high-skilled workers went up. That, combined with scarcity (it's costly to acquire technical skills) increased the wage premium to higher education.

      Here's a random piece of anecdotal evidence about China. About 3 years ago, I was in Shanghai, and talked to an assistant professor who was teaching at Fudan University. Her father owned a textile company in Shanghai, and he was complaining about competition from Thailand where, he argued, they seemed to have an abundance of cheap low-skilled labor.

    2. Also, technology has made offshoring a LOT easier.

    3. Stephen,

      "Technology has made it cheap to replace humans with automation, driving profits up and wages down." I believe this is a tautology, otherwise why would a company implement a technology? I also believe that this has always been true, particularly at the end of WWI when "mass production" really started and people worried about the replacement of manufacturing jobs with machines. I think that the rate of technology replacing humans is the primary metric in terms of timescales of human careers. If technology would to replace all workers tomorrow that shock would be very disruptive to workers and their actions as consumers. However, if workers get replaced at a rate that allows them to change careers along the way, then the change is not so disruptive.

      I think that Autor's comments are really about this rate of change of employment by automation. I agree with his points, but they do not imply high unemployment per se, rather a change in employment from low skilled to high skilled jobs. The challenge is to enable workers to make this change of job category during their working careers. A change of jobs category between generations is much easier as the children simply study something that will allow them to get the jobs more in demand for then than for their parents. We've seen this throughout history. E.g. the father is a toolmaker the son is an engineer.

  23. urban legend5:46 PM

    "And the decline of unions is sad in many ways, but it seems to be a symptom of labor's decline rather than a cause."

    Seems to me there's 100 elephants in the room on this one. What you are talking about is union strength in the declining manufacturing sector. The poorly unionized parts of the non-offshorable services sector are many, many times the number of employees in U.S. manufacturing at its peak. You can't be serious and tell us that determined and coordinated employer opposition has not played a huge role in the failure of organizing in the retail, lodging and healthcare industries, to name only the biggest, to fill in the gaps several-fold caused by off-shoring manufacturing.

    Why are so many liberal economists so anxious to discount the decline of unions -- the decline in the ability of workers to participate in negotiating their share of the revenues their work is substantially and often primarily responsible for generating -- as a major force in growing inequality? Curious? What's up with that?

    1. urban legend5:57 PM

      P.S. It has some earmarks of the "seriousness" thing, much like being anti-Social Security is the only way you can get into the DC "VSP" club. Maybe it's like a necessary mantra for entry into the profession, perhaps because the profession instinctively hates the idea of greasy, uneducated workers interfering with management's vision and decision-making. It's just messy that we have this freedom of association thing mucking up the decisions predicted by pure economics. Or something like that.

  24. You and your mentor are also missing massive tax changes from the late 1970s to Present that privilege non-wage income over wage income, credit market laws that make debt easier (and the parallel issue, easier for bond holders to force payment). Deindustrialization does not begin with China, it begins with Japan in the 80s, and with the migration to the coasts. Also, agriculture as a form of employment went from #1 in 1900 to not-even countable (legally) today, increasing the high cost of the shrinking city phenomenon. It is NOT all about cheaper labor (machines, chinese people), contrary to economic opinion. There is more to history than plugging things into the equation

    1. Nathanael4:08 PM

      Yep. The tax laws are a big issue here, completely ignored by the essay.

  25. Add to that rising costs (health, food, energy, education) amidst 3+ decades of stagnant income and MAYBE you are getting close to the answer. Also, plants on the border with Mexico have been operating for decades.

  26. Robots?? As a computer scientist who has spent a career automating clerical processes, I am still amazed that most people seem unaware of how much automation has replaced white collar workers - not blue. Just think about it. Remember 20 years ago? Airline reservations? Shopping? Newspapers? Your bills? Ordering just about anything? Books? Music? People's memories seem very short.

  27. How do we keep getting into "either/or" choices rather than "both/and"? Multivariate, anyone? Proof that economists can be both vain and boring. Nevertheless US productivity has been improving over 100 years, but persistent trade deficits are relatively new.

    The whole point of the PBOC creating trillions of Yuan to buy trillions of US Treasuries and MBS is to import jobs and demand from the US to China (and as a byproduct exporting inflation too). How can a 2-3% of GDP trade deficit not increase unemployment? Foreign governments own $6 trillion in US securities, how can that not create bubbles and low interest rates? What academic gymnastics could prove otherwise?

    Between the efforts of Asian central banks and the lobbying of US multinationals there is a concerted effort to outsource labor. Free trade must be fair trade, which means that large economies must let their currencies float.

    In the past when machines replaced men, the human labor was directed towards serving each other education, healthcare, and entertainment. But the persistent leak of employment, tax revenue, and viable capital investment is constraining the median American's ability to pay for new services.

    As an aside, why does Apple pay for Chinese rather than robots to manually assemble iPhones? And does that cheap labor retard innovation?

    1. In a word, yes. The New York Times article about iPhone manufacturing missed this entirely; they assumed that if it were assembled in the US Apple would need the same huge army of employees that Foxconn employs. But US labor rates would justify investment in a lot of automation. For that automation to be flexible enough to accommodate the frequent model upgrades, yes, they would have been pushing the envelope.

  28. Anonymous7:29 PM

    interesting that only one comment mentions rent, that too in passing.

    obviously the function of academic economics - to deceive and occult the truth about political economy using differential equations has worked well.

    decline labor share just means rising rents. heck why do you think so many lawyers are there in DC?

  29. Anonymous9:00 PM

    Richard Freeman made the same argument with much the same supporting data back in 2006. See "The Great Doubling," Freeman, R., U.C. Berkeley, 2006.

  30. Hi.......................
    Great post like this must be highly recommended. It is so nice to read such wonderful blog. Thanks.

    Foreign Trade Policy India

  31. Thank you for your post. This is excellent information. It is amazing and wonderful to visit your site. It really gives me an insight on this topic.Here you can find many more tips about Self Improvement and Personal Finance.

  32. Stephen Williamson: "But the demand for high-skilled workers went up. That, combined with scarcity (it's costly to acquire technical skills) increased the wage premium to higher education."

    Which is why salaries for engineers have been soaring in the US for the past few decades?

  33. Nathanael4:01 PM

    Theory #4. Tax rates.

    In the 1950s, the very rich had their income taxed away from them as soon as it arrived. This meant that corporate profits got reinvested in expanding the coproration, increasing employment, raising wages, etc., forcing the CEOs to compete by being "more paternalistic" than each other, rather than "richer".

    Now, they can keep the money. So they do.

  34. Actually I do not find robots displacing that plausible. Or to put it in another way, any technological advance has the same property as "robots". It is not self-explanatory to say "people have lower wages because new technology displaces labor". For instance new technologies in agriculture were displacing farm labor for centuries and the wages were increasing. So why it is different now with robots?

    And I have to second Andy Harless and one Anon commenter on very good hint: Rents. Krugman has a good article about this named "Profit Without Production". Every cent earned due to copyright protection, patents or any other way to prevent competition from entering the market means additional profit from rents. There are tow additional sources of rents Matt Yglesias likes to talk about:

    1) Rising real estate prices cause by which is just another way how to say increasing rents on land is another example of the same phenomenon. The main source of these rents are zoning regulations that protect existing owners of the valuable land.

    2) Occupational licensing that is skewed against low income workers. While factory workers are not only facing competition of global marketplace for manufactured goods and rising imigration, some occupations like Doctors have a lot of monopoly power to ban opening the labor markets to competition.

    The same dynamics goes on top level, where for some reason executives would laugh out anybody opposing relocation of manufaturing due to cultural reasons - but they would be probably surprised if somebody would say they hire cheaper German or French or Chinese executive who will work as hard as them and take fraction of their pay. There is something rotten in corporate governance laws in this regard.

  35. Multi-factoral (with error bars)