Capital-biased technical change vs low interest rates?

Paul Krugman says that recent technical change has been capital-biased. That robot story sounds plausible to me too. But if so, why are real interest rates so low? (Yes I know there's a global recession on, but real interest rates were falling even before the recession). Maybe we are forgetting a third factor, land, and land rents are rising?

Suppose there were a new technology that caused both wages and interest rates to fall. Why wouldn't capitalists and workers say "Stuff that, let's go back to the old technology!"? Any individual firm, or group of firms, that went back to the old technology, while borrowing at the new low interest rates and paying the new low wages, would make super-normal profits. And all the other firms would eventually choose to, or have to, follow it back.

Unless there were a third factor, and the new technology paid higher rents to that third factor.

We used to think that that third factor was skilled labour, that knew how to work with the new "skill-biased" technology, was very productive at the margin, and so earned high rents. But if, as Paul says, the college premium has stopped rising, and so that third factor is not skilled labour, what is it?

Much longer ago, in the days of Malthus and Ricardo, we used to think that that third factor was land. Maybe it is again. Maybe some of those increasing "profits" are really increasing resource rents?

(I have found good data showing strongly rising rents on agricultural land in the UK, but I can't find a Canadian or US equivalent, only anecdotal evidence of rising rents. And there are more natural resources than agricultural land, of course. Rising land prices are consistent with my story, but don't really support it, because if real interest rates fall land prices will rise even if rents stay the same.)

Or maybe, if not land, it's some other fourth factor that is earning higher rents?

[PS. Historians of thought may recognise some very old questions here. Like how could Marx get both increasing immiserisation of the proletariat and a falling rate of profit? Unless land rents were rising, like In Ricardo. I can't remember if they ever solved that one.]

Comments

[Anon]

The explanation could also fall outside the realm of monetarism: rates could be low artificially, due to high unemployment.

And if unemployment is low not due to not enough money (explained by monetarism), but due to technology globally replacing hundreds of millions of workers permanently (which falls outside monetarism), then the Fed's lowering of rates due to unemployment addresses the wrong problem: no amount of money will make those workers useful again, in a capitalist, free market economy.

"Freedom" includes the right to marginalise, abuse and ultimately abandon other people.

[Luis Enrique]

I'm not confident I understand your thinking, but i don't see how thismnew tech can cause return to labor and returns to captial to fall whilst still being something that somebody would wish to adopt, unless you suppose a third input like land, and it is, directly or indirectly, the landowners who are driving adoption of this tech. And presumably this tech is being adopted because it out competes its rivals, so captial owners can't revert to the old tech because they'd be unable to compete with the new tech implementers.

Are real interest rates, by which I presume you mean something like Fed or bank savings account rates minus inflation, a good indicator of the rations being enjoyed by owners of physical and intangible captial?

I've often puzzled about this. Could we be in a situation where returns to installed captial are high and may have even been increasing as wages fall, yet the returns to new investment are low?

[Luis Enrique]

Returns, not rations

[Hunter Pritchett]

What about the return on Graduate level studies? How has that fared? A Ph.D. is the new B.S. Seems to me people with an MBA from a prestigious university are doing quite well these days...

[Nick Rowe]

Luis Enrique: "I'm not confident I understand your thinking, but i don't see how this new tech can cause return to labor and returns to captial to fall whilst still being something that somebody would wish to adopt,.."

I don't see it either. I think they wouldn't adopt it. My argument is that if they did adopt it (for some unknown reason) they would soon unadopt it.

[Luis Enrique]

What is the mechanism that is supposed to tie real interest rates in financial markets to returns in the real economy? I presume the basic idea is that if real returns are high, then demand for loans to invest in real projects would be high, driving up interest rates whilst the investment thus funded drives down available returns in the real economy, eventually equating financial real rates with real real rates.

But what if high return investment projects are somehow rationed (intentional spelling this time) and financial rates are set by policy makers? Then the mechanism is bust. Some lucky duckies have access to investment projects involving robots, their cost of capital is low, and they make out like bandits. But there is no boom in demand for loans to move the financial markets.

I expect somebody to blow a hole in this idea.

[Nick Rowe]

Hunter: good point. It might be just a signalling war, where as more people invest in a BA the signal isn't worth as much, so you now need to signal your abilities with an MA or PhD. And nothing to do with any bias in technical change.

[Nick Rowe]

Luis Enrique: That idea can make sense too. Top managers grabbing control and skimming off the rents for themselves. But I think there would need to be some sort of barriers to entry. Or top managers might be the fourth factor, with technology being biased towards their skills.

[K]

Nick,

This is right (and krugman is right). Asymptotically, when robots do all the work including the development and manufacturing of robots, all rents will flow to the owners of land just like a few hundred years ago. At that point the prescriptions of Henry George will be self evidently true. Until then a great deal of rents will flow to the owners of non-land capital. Watch for ever broadening, ever lengthening IP protections, and continuing income polarization.

[Anon]

Nick: US corporate board membership is a pretty closed club with high barriers of entry.

The rents are high, so there are little incentives to change.

That's 1 trillion dollars per year of corporate profits looted - a 6% rate.

[Andy Harless]

I think one thing you're missing is the international savings glut, which predates the recession. To put it in economistic terms, there seems to have been an exogenous decrease in the subjective rate of time preference somewhere around the turn of the millenium. (Or one could say that the Chinese government just decided to dump a boatload of capital on the world in the decade-plus following the Asian financial crisis.) This would explain both why interest rates are low and why businesses are substituting capital for labor. It's noteworthy that, at least in Asia, the savings glut seems to have reversed (although now we have a different savings glut in Europe and North America, reflected in the zero interest rate bound rather than in high observed savings rates, but presumably that will change as balance sheets get repaired and as confidence improves). So maybe the labor-to-capital substitution is not a trend that is going to continue.

[Nick Rowe]

K, Anon, Andy: three good comments. I have nothing to add.

[Vlad]

I would suggest the third factor is "brands" . Brands are of course supported by additional intellectual property such as patents and copyrights in addition to trademarks. The high rents go to brands owners that can charge a premium price even in markets that are essentially competitive.

[Nick Rowe]

Vlad: not a bad idea. But brands can be seen as capital. If you invest, you create a brand that makes your goods more desirable in the eyes of consumers.

But here's a slightly different interpretation: brands may make markets less competitive, in the sense of reducing elasticity of demand for a firm's product. Which means firms will be operating on the downward-sloping part of their ATC curves, and do not fully exploit economies of scale. (It's hard for me to explain that clearly without diagrams.)

[Vladimir]

Nick:

Much longer ago, in the days of Malthus and Ricardo, we used to think that that third factor was land. Maybe it is again. Maybe some of those increasing "profits" are really increasing resource rents?

To make the hypothesis even more pessimistic, maybe it's not just land rent of agriculture and resource extraction, but also the rent of habitable land near the centers of economic activity -- which is rising because economic activity is increasingly concentrated.

We do seem to be going towards a world where there just isn't much to do outside of a few large metropolitan areas and resource extraction centers. What can you do in a small town these days? Some agriculture, some government work, and some services that emerge around these, but that's pretty much it. So masses of people flock to the big cities where there's a lot of productive activity and hence jobs, which pushes the land rent in these areas through the roof.

This of course leads to the question why the capital doesn't move out to low-rent, low-wage places. This is indeed happening with the mechanical capital for the physical assembly of stuff. However, if the physical assembly is becoming less important relative to the organizational capital that's necessary to run the rest of the business, and where you must have a concentration of people who are able to meet face to face, there's no way to escape the land rent of big population centers.

[Nick Rowe]

Vladimir: very good point. I wonder how big that implied land rent is, and where it shows up in the national income accounts? I ought to know the answer to that question. It's maybe different for owner-occupied vs rental housing. And different again if a business rents or owns the land.

Let me ballpark it. Suppose we spend 20% of our income on housing rents, or implied rents. And that a quarter to half of that is land rent (the value of the lot is that fraction of the total value of the house). We're at 5% to 10% of GDP.

[K]

I'm with vlad. Some people thing marketing adds consumer value. Seems to me that competing to wear the coolest shoes is a zero sum game. The monopoly rents enabled by competition for social rank make it negative sum, a totally inefficient equilibrium. Brands are yet another form of IP not deserving of government protection. Why do we subsidize increasing returns to scale? If we need to protect producers from deceptively inferior knockoffs we could issue all producers an exclusive producer number that nobody else can use. But protecting the Rolex logo - what a waste.

[K]

Vladimir, Nick,

Good points.

"there's no way to escape the land rent of big population centers."

No, but you can tax them at 100% without loss of output. That's 5-10% of GDP we can use to cut bad taxes like sales or income. How awesome is that???

[Anon]

Don't know - space is still available in urban areas, if really needed: vertically. High rise buildings are one of the greatest inventions of humanity, ever.

If land rent was our big story, wouldn't we see something dramatic in California's vacancy rates, and in other fast growing population centers?

Yet we don't.

[K]

Anon,

What is the postulated relationship between land rents and vacancy rates?

[Anon]

K: assuming a functioning market, wouldn't we see historically low vacancy rates in fast-growing urban areas, driving up rents/rates?

Yet the vacancy rate follows (and in the latest recession, leads) the business cycle, well within historic ranges.

If the story of a long term land squeeze in these areas was true then I'd expect something unprecedented: the vacancy rates ignoring the business cycle or at least going outside the usual range.

[Jacques René Giguère]

Land rents in urban (New York, London) and playground (Aspen, Davos, Gstaad)for the über-class and limited membership in said class ( part of it going to tuition fees at Harvard. Their income is deducted from corporate profits and thus lower return on K while we ascribe their income to their MBA and so manufacture a non-existent return to education myth.
Vacancy rates in überland are essentially zero and we do manufacture more in the form of skyscrapers. Not fast enough to lower the rent and anyway, as they extract more and more, they have the income to push back.

[Determinant]

I think one thing you're missing is the international savings glut, which predates the recession. To put it in economistic terms, there seems to have been an exogenous decrease in the subjective rate of time preference somewhere around the turn of the millenium. (Or one could say that the Chinese government just decided to dump a boatload of capital on the world in the decade-plus following the Asian financial crisis.) This would explain both why interest rates are low and why businesses are substituting capital for labor. It's noteworthy that, at least in Asia, the savings glut seems to have reversed (although now we have a different savings glut in Europe and North America, reflected in the zero interest rate bound rather than in high observed savings rates, but presumably that will change as balance sheets get repaired and as confidence improves). So maybe the labor-to-capital substitution is not a trend that is going to continue.

To follow on from this, firms have engaged in labour arbitrage for the last 25 years, substituting cheaper Asian labour for expensive North American labour as a strategy to boost profits. It impacted the bargaining power of a great swath of North American labour. But the consumption market for that substitution labour, that is where that labour's goods were going was still North America.

Now that trends *seems* to be ending. Asian labour wants higher wages and North American consumers are tapped out and restraining their spending. So the trend is being attacked from both ends. Asia, particularly China seems to be approaching the 1890's - 1920's era in North America where labour began to flex its muscle and demand higher wages and improved conditions.

Thus the last available market for labour arbitrage is Africa, but Africa still has government barriers (lack of stable governments) and it is smaller if you try to arbitrage labour in Asia, Europe and North America so it will have less of an effect.

[Ritwik]

But the rate of profit isn't falling Nick. Corp profits are at an all time high.

And risky real rates are not low.

The theory of interest rates without risk or uncertainty is so incomplete that it is completely misleading.

[K]

Anon,

I don't think so. I would guess that vacancy is almost entirely driven by the business cycle rather than the equilibrium level of rents. At equilibrium I don't see why vacancy would be lower in a small town than in a mega city. Vacancy *and* rents can be high in big cities during a temporary downturn and housing can be very difficult to find in a booming rural region. So long as a booming city is growing at a predictable rate I would expect supply to keep up with demand and vacancy to remain stable.

[Tim Worstall]

"I have found good data showing strongly rising rents on agricultural land in the UK,"

I would be very careful of interpreting that particular set of numbers.

Farm support (through the EU) changed from per capita (ie, crops produced, cattle kept etc) to simple ownership of the right to farm a piece of land. Not to the land owner, but to the tenant (or the landowner if no tenant). Quite substantial sums, £100 an acre pa or more seems to stick in the mind.

As Ricardo would immediately have pointed out this is going to feed through into the capital/rental value of the land soon enough.

Whether that's the whole story is another matter: but it's certainly part of it.

[Nick Rowe]

Tim: good point. The CAP will mess with UK data. I really need Canadian or US data, but all I can find is things like newspaper reports that say "Yes, land rents have been going up a lot around here recently" (I'm not very good at finding things on the web.)

Ritwik: The *level* of corporate "profits" isn't the same as the *rate* of return on *investment*. Interest rates, stock dividend yields, earnings/price ratios, would give a better measure.

Anon: if there were a sudden increase in demand to rent land, we might see vacancy rates fall in the short run, because rents are sticky. But I'm talking about a longer run trend of growing demand, so we would expect to see rising rents and vacancy rates staying roughly constant.

[Makrointelligenz]

Strategic savings in emerging countries and changes in demography based savings could overcompensate the increasing demand for capital and thus explain increasing capital and and real interest rates. With the share of middle-aged saving people increasing and the dissaving rentier class still relatively small, demography is pointing to high savings more than ever. Does that make any sense to you?

[Nick Rowe]

Makro: that makes sense, but I don't think it's right. If there are only 2 factors, how can increased savings make both wages and interest rates fall?

[Ritwik]

Corp profits as % of GDP are on a secular rise. Investment as a % of GDP is on a secular decline. I wasn't talking about the *level* of corporate profits.

Of course, the expected rate of return on fresh investment is low, which is why we're in a recession. There's a dynamic argument to be made here where current high profit rates maybe in response to an expected decline. But there's no escaping the fact that in recessions past, including the GD, corp profits took a hit, even as % of GDP. That's no longer the case now.

[Becky Hargrove]

There are ways for local areas to adapt as returns of scale will sometimes reverse in production, such as with 3D printing which has wide applications. Because of this growing technological trend, by no means should such manufacture have to be "bad" in terms of capital structure or potential land use, because a wide variety of local 3D and related manufacture can become part of community inclusive production choices. What that means in practical terms is that people can invest in such equipment not so much for profit ("external" economic activity on the part of what they invest in) as the knowing that they can benefit as a consumer over time in what the machines are capable of producing locally, and any extra product that other communities want from this largesse is like a dividend. Such production would be beneficial in that it could also focus on local resource options along with recyclables. Plus, the knowledge required for such production could become a part of local education.

[Dan Kervick]

Suppose there were a new technology that caused both wages and interest rates to fall. Why wouldn't capitalists and workers say "Stuff that, let's go back to the old technology!"? Any individual firm, or group of firms, that went back to the old technology, while borrowing at the new low interest rates and paying the new low wages, would make super-normal profits. And all the other firms would eventually choose to, or have to, follow it back.

Nick, if you have time, could you explain the reasoning here a little more? Once the investments have been made to replace the old technology with the new technology, then I assume going back to the old technology is a further cost. So even if the return to capital had been higher under the old technology, the cost of going back to it might exceed the gains to be made by going back to it. I'm assuming that the daily investment of capital can be divided up into something like these three time periods:

1. Daily capital investment - old technological regime
2. Daily transitional period capital investment
3. Daily capital investment - new technological regime

If the capitalists undertake the expensive period 2 investment because they anticipate that the rate of return during period 3 will be higher than during period 1, but they turn out to be wrong, it could be that the cost of re-installing the old technology is higher than the accumulated expected incremental daily return from going back to the old regime.

But also, couldn't it be that the higher return under the old technology was only possible when all of the competitors in the industry were using that old technology, and can't be recovered by going back to it? I'm thinking for example of a shift from a technology that permitted to 10-day order turnaround to one that permits a 3-day turnaround. If in a given industry all of the firms mutually anticipate that their competitors will all shift to a new technology, and that the shift is going to reduce prices of the industry's product, make the production of the product more efficient in terms of both labor and capital, and result in a lower return to each, might they not still conclude that they have to make the shift anyway because their only choices are between a small reduction in the ROI and a massive reduction in the ROI.

Maybe this is the similar to K's point?

Finally, what if the shift to the new technology is accompanied by an increase in the number of capitalists, so that you end up with more capitalists, each making a smaller return on investment, but the total return to capital is much higher than before, and also the share of returns going to capital v. labor is higher as well.

[Johannes Yohighness]

"
international savings glut, which predates the recession. To put it in economistic terms, there seems to have been an exogenous decrease
"
~~Andy Harless~

Believe it! We are overcapitalized in this country with an excess of savings but with increasing level of unused capacity. Why no opportunities for proper investment? Because our rulers have taxed out of business most of the small business start-ups. As a favour to lobbyist representing the excessively large dinosaur companies Congressional Critters have taxed small business straight into bankruptcy court.

And we vote for those gals/guys
?

[Makrointelligenz]

"Makro: that makes sense, but I don't think it's right. If there are only 2 factors, how can increased savings make both wages and interest rates fall?"

I feel this does not have to be a contradiction. While the interest rate is falling, interest income can still stay high when high net investment compensates.
I think the question in your op is misleading: "Suppose there were a new technology that caused both wages and interest rates to fall. Why wouldn't capitalists and workers say "Stuff that, let's go back to the old technology!"?"

I think it is not a new technology they can choose from that makes the difference. it is competition due to high supply of savings which capitalists would have to control to be able to keep the status quo. The new technology might even lead into the direction of higher interest rates but can be overcompensated.

[Robert]

I think the data in Krugman's post are just wrong.

Cleveland Fed Gomme and Rupert
St Louis Fed Pakko

Both dispel the kind of data Prof K uses. Its a non- issue.

[Becky Hargrove]

per Dan Kervick's last paragraph: greater dispersal of capital among economic actors also allows for the potential of greater knowledge use dispersal among economic actors creating overall gain...compromise, anyone?

[Brent Buckner]

I do think you're being overly restrictive in demanding that the technological change lowered real interest rates. I think you'd do well to allow for multiple factors.

Freer trade increased wage competition to U.S. domestic workers and enabled increased productivity of developing economies which may have resulted in greater global savings (and so lowered real interest rates). Technological change may be especially evident in the declining relative price of investment goods (e.g. robots building robots) while not so evident in time-evolving cost of capital (i.e. risk-adjusted real interest rate).

[Anon]

Nick, K:

So if (business) housing supply keeps up with demand, how can prices be so high as to amount to a rent?

There's no natural monopoly here: the third dimension has space up to a mile and more. Let's assume a functional market, I.e. no collusion between sellers and no monopolization of supply. (This, except in a handful of critical areas like housing close to stock exchange computers, is largely the case.)

So why shouldn't prices approximate the cost of building (and thus be one-off, not a perpetual rent), why should they amount to a rent?

[judabomber]

Agricultral land data by state and land type for the United States can be found here:

http://quickstats.nass.usda.gov/

Survey>economics>expenses>rent>cash expense for crop or pastureland in $/acre>

[Stephen Gordon]

Not sure if this says anything about the conjecture, but the labour share in Canada - which profits from resource scarcity - doesn't seem to have the same downward trend you see in the US:

Labour’s share of income: why are trends in the U.S. and Canada so different?

[Jacques René Giguère]

Anon: "a mile or more"?! .No. Tecnologically yes but above 70 stories, no rent is worth construction costs and what you need in extra servitude space ( water conduits and elevators) takes back on the lower levels what you gain on the upper floors. You can try double-deck elevators and sky lobbies but building above 70 is today as it was for the Empire State Building and the belfries of medieval cathedrals: signs of youthfulboasting. Let's not go into trivial Freudism but sometimes a tower is not just a tower...

[Nick Rowe]

judabomber: Good find! thanks.

So, US land rents were rising slowly, roughly in line with inflation, until 2007, at $77.5 per acre. Then they increased more quickly, to 2012, at $125 per acre.

That is some sort of confirmation of my hypothesis/guess.

Stephen: good point. Not sure if that confirms or disconfirms, what should be a global hypothesis. Maybe it's because Canadian labour is used more in the resource sector?? Ours being a more resource-intensive economy?

[Anon]

Jacques:

Let's assume an even more conservative limit: 50 stories max, with current technology.

We are nowhere near that limit in most of the fast growing economic centers, except a few square miles of the hottest business centers and special geographic/political regions like Hong Kong.

And the vacancy rates I cited appear to support that view: there's still healthy supply & demand forces at work, with only the occasional scarcity.

(You can also check rental/office space spending of some of the largest U.S. corporations in their SEC filings - it's not a significant sum in terms of GDP impact: less than 1% GDP I'd estimate.)

[Nick Rowe]

Dan: "Nick, if you have time, could you explain the reasoning here a little more? Once the investments have been made to replace the old technology with the new technology, then I assume going back to the old technology is a further cost."

If the new technology meant lower wages *and* lower rates of return (interest) why would they switch to it in the first place? Even more so if there are adjustment costs of changing technologies.

Ritwik: "Corp profits as % of GDP are on a secular rise. Investment as a % of GDP is on a secular decline. I wasn't talking about the *level* of corporate profits."

Corporate profits are not the same as total returns to capital. They include land, if the corporation owns land. They exclude the returns to those who hold corporate bonds. They exclude non-incorporated capitalists.

Capital =/= corporations.

And, I'm not talking about profits as a % *of GDP*. I'm talking about the % *per year* rate of return on investment.

[Anon]

One argument I haven't seen mentioned is somewhat of a taboo in economist circles: the long term effects of systematic union busting in the U.S.

Unions, in the abstract, increase negotiation strength - pricing - on labor supply side. Their counterparts on the business side are business associations and business lobbying groups.

Union membership is at historic lows in the U.S. and the business lobby is at a historic high.

Is it really a surprise that disorganized workers are getting a worse deal than decades ago and that most of the profit remains with business owners?

[Nick Rowe]

Robert: good point.

Paper by Paul Gomme and Peter Rupert here (pdf)

But I can't find which Pakko paper you are referring to on this list.

[K]

Anon: Ricardo pretty well closed the book on this topic some 200 years ago. Wiki Law of Rent: "the rent of a land site is equal to the economic advantage obtained by using the site in its most productive use, relative to the advantage obtained by using marginal (i.e., the best rent-free) land for the same purpose, given the same inputs of labor and capital"

Vacant land in a big city has enormous value, the PV of all future rents. The owner never did a thing to create that value; it's entirely a windfall positive externality produced by other peoples economic activity (which is why it should be taxed at 100%).

[Greg Ransom]

"Suppose there were a new technology that caused both wages and interest rates to fall."

How does this work?

I'm particularly interested in the logic of choice which says new technology lowers the rate,of interest.

[Nick Rowe]

Greg: short answer; it doesn't work (assuming no third factor). My argument was a form of RAA. Suppose it did work. Then they would reject it, and go back to the old technology. Therefore it wouldn't work (unless they made a mistake in choosing it originally).

Old technology: 1 unit of labour produces 1 unit of output 1 year later.

"New" technology: 1 unit of labour produces 0.5 units of output 1 year later.

Paul Krugman has a new post, responding to this one (saying I might be right, but not sure if the land effect is big enough, more research needed -- fair enough), and saying it could instead be an increase in monopoly power, similar to Vlad's/my argument in comments above.

[Pragmatic Liberal]

Hmmm. A big chunk of the variation in the cost of owning or renting land is property taxes, so I plotted property tax collections as a share of GDP: http://research.stlouisfed.org/fred2/graph/?g=dEx There doesn't seem to be much of a correlation.

[Nick Rowe]

Pragmatic: but don't they simply adjust the mill rate to get the same revenue if property prices go up or down?

[Nick Rowe]

I'm surprised nobody has mentioned oil yet. Oil (in the ground) is land. So is copper ore, etc.

Being a farmboy, "land" to me is farmland. Some of you city kids quite rightly say that "land" is city centre land too. But oil and copper and iron and hydroelectric etc. are all land.

[Nick Rowe]

Rant warning on.

You know what the biggest problem with the aggregate production function really is? Why Y=F(K,L) = K^a.L^b where a+b=1, is so very wrong? It's not that it aggregates L and K. It's that it leaves out land. You can't double capital and labour with the existing technology and double output. Not without doubling land too. And look at how we divide up national income: return to labour plus return to capital. What about returns to land?

Blame Manchester. Blame Marx/Engels. Blame the Manchester Guardian in particular. It's all about Kapital and Labour. What about LAND!?

End of rant.

[Dan Kervick]

If the new technology meant lower wages *and* lower rates of return (interest) why would they switch to it in the first place? Even more so if there are adjustment costs of changing technologies.

I suggested two answers to this question in the previous comment.

1. They have imperfect knowledge and are imperfect calculators. Things that look good sometimes turn out not to be.

2. They have to switch for competitive reasons just to avoid greater losses. The whole industry can't be modeled as a single agent seeking its best aggregate outcome. It is a collection of competitors seeking their best individual outcome. They could be in an N-competitor situation in which standing pat and innovating are the only two options, and where the relevant outcomes for each competitor A are:

a. no change if all competitors stand pat
b. small losses for A if all competitors innovate
c. large losses for A if even one of the other competitors innovates

Suppose at least one competitor plans to innovate and all the other competitors know this. Then the rational choice is to innovate.

[Dan Kervick]

Sorry about the italics.

[No worries. I fixed it. NR]

[Nick Rowe]

Dan:

1. Mistakes happen. But then firms which stuck by the old technology should make higher profits if they can now pay lower wages and lower rates of interest.

2. We know that a competitive equilibrium (with no externalities) is not a Prisoner's Dilemma. We know that with monopoly power the equilibrium is like PD in some dimensions (if all firms in the whole economy increased output all would gain but any individual firm that increases output loses). If an economist tries hard enough he can eventually come up with a model showing almost anything is possible. I have played around a bit with macro models where all firms have monopoly power. But I can't immediately see a way to rig (let alone plausibly rig) a model so that it would be individually optimal but in aggregate suboptimal for firms to adopt a new technology that is worse on both dimensions -- paying lower rates of return to capital and to labour.

[K]

Nick,

"What about LAND!?"

That was Henry George's point, and the reason why he didn't join the Marxists. As for oil, etc, I thought that went without saying. The classicals right from Adam Smith always included in the meaning of land all scarce things not made by humans.

[Nick Rowe]

K; Yep. And you were right about Ricardo. On land rents, Ricardo is da man.

More ballparking:

There's about 1 billion acres of farmland in the US. Say $100 per acre rent. That's $100 billion. US NGDP is about $16 trillion. Hmmm. Farmland alone isn't big enough. Farm rents are less than 1% of GDP.

Canada has 167 million acres of farmland. So farmland rents are about 1% of GDP.

My guess is that city centre rents are a bigger deal than farmland.

[PlanMaestro]

Have you analyzed the largest share in the economy of regulated sectors? The rise of banking, healthcare and for-profit education are high ROE sectors thanks to protected profits. Also consumer goods have managed to protect margins while becoming a larger share of the economy.

I think that would be an interesting analysis: a comparison of the industries that composed the S&P in the 1970 to today's. You would find a much higher percentage of competitive sectors like manufacturing, energy, and other commodities.

The transition to services has been much on the back of this high ROE industries with barriers to entry, many thanks to regulation. So Krugman's "monopoly" story has something going for it: monopolistic profits thanks to barriers to entry… specially in regulated industries.

[Frank Restly]

Nick,

"Historians of thought may recognise some very old questions here. Like how could Marx get both increasing immiserisation of the proletariat and a falling rate of profit? Unless land rents were rising, like In Ricardo. I can't remember if they ever solved that one."

There are two issues here:
1. Market segmentation - You can get changes in relative prices through barriers to entry. They can be market based (competition, collusion, etc.) or law based (patent rights, tax policy, etc.).
2. Credit expansion / contraction - You can get changes in absolute prices through currencies that can both expand and contract.

And so you can get a falling rate of profit through credit contraction, and falling real wages through labor contract abrogation. All that land rents do is add a third good for money flows to be directed towards / pulled from. I believe this is related to your earlier article about Cantillon effects.

[David]

Hi Nick-

I'm a first time commenter. Isn't the lack of return on investment--I'm assuming that is what you meant by rents--directly a result of the exponentially growing technology curve? I mean were do the worlds best and brightest mathematicians and scientists go upon graduation? Answer: Wall St. We now have physicist that determine the rate of return on capital with black boxes, as it happens in light speed. This seems to be squeezing the margins for your average investor.

[Will]

I'm always surprised when Nick comes out for Henry George. I've also wondered for some time why Land stopped making the list of "factors of production." I'd suggest that Nick's blaming Marx for it is not fair. Part 6 of Capital, vol. 3, is titled "Transformation of Surplus Value into Ground Rent", and rehashes at length Ricardo's argument (actually made earlier by Malthus, Torrens, and others, but never mind, I like Ricardo too). I think the man to blame for hushing up Land is John Bates Clark.

And just as oil is land, things like taxi medallions, bandwidth, railroad right-of-way, rights to air strips, should, I think, be considered land. Legal restrictions can create land.

[Anon]

K:

But the rate of rents can stay higher than the average return on capital only as long as construction of new space catches up. (With the exception of physically limited areas where the building of new space is not possible anymore.)

So a significantly higher rate from rents is only possible if growth is much faster than construction. But this would immediately show up in sharply lower vacancy rates, as buyers rush for free space, driving up prices and driving down the vacancy rate.

Keeping free space vacant only makes sense if you cannot find a buyer, or if you speculate on growth in the future not yet priced into rents.

Yet we are seeing only a modest drop in vacancies, in some of the fastest growing urban areas of the U.S. - which seems inconsistent with the theory that rents are a significant drag on the economy.

[Richard H. Serlin]

Doesn't seem like real expected stock returns have gone down, just the risk premium (or perhaps falsely perceived amount of risk) has shot up.

Stock investment may be much better for funding productive advance than bond investment. I base a theory explaining the equity premium puzzle on this:

All of the explanations for the equity premium puzzle I have seen in the literature are based on
the demand side; trying to find utility functions for a representative investor and ex-ante
probability distributions for returns that would explain investors demanding such high average
returns for stocks relative to bonds, rather than bidding those returns down. But I suggest a
supply based explanation: The long run supply curve for corporate stock may simply be
extremely long and flat, and consistently about 5 ½ percentage points in return higher than the
premium bonds supply curve, even at stock quantities as high as the entire national savings rate.

Why would this be? I posit that stock might simply allow a firm to create more wealth with an
investment dollar than bonds, and this is because of the flexibility of stock. Firms are able to
invest in high return long run projects when they raise money with stock that they sometimes
cannot when money is raised from bonds due to the short run constraints of having to make
interest payments and satisfy bond covenants.

at: http://works.bepress.com/richard_serlin/18/

[K]

Nick,

Here's an interesting NY Fed paper about land value in the New York Metropolitan area. Prices for unimproved land in the vicinity of the empire state building run at several thousand $/ft^2. In the whole 11842 square mile area they claim the average price per ft^2 was $366 in 2006. That works out to a total value of $120 Tn. Since that's like double the total value of public stocks and bonds in the US it stretches credulity. But then, apparently the value went up by a factor of 6.5 over the preceding 6 years. Comments anyone?

[K]

Anon,

We are talking about *land* rents right? Not real estate. You don't seem to distinguish between vacant land and vacant apartments and offices. I am not following you.

[Jane Walerud]

If a company sitting on cash (invested at 1% or less return ) can see an investment where they can substitute capital for labor( which definitely costs something, it's not easy to predict how much it will cost over the life of an investment, and is a fair amount of trouble to boot), I think the company just might make a capital intensive investment rather than a labor intensive one.

Rather than looking at actual rates of return wouldn't you look at alternative rates of return?

/Jane

[K]

Nick: "You know what the biggest problem with the aggregate production function really is? ... It's that it leaves out land. You can't double capital and labour with the existing technology and double output. Not without doubling land too...Blame Manchester. Blame Marx/Engels."

Will: "I think the man to blame for hushing up Land is John Bates Clark."

I blame both Marx and the neoclassicals. But in the end it was the neoclassicals who wrote the economics texts, and the expunging of finite resources from the long run equilibrium strikes me as a far more egregious omission than the expunging of class. It also achieved the expunging of George who in all likelihood was a far more serious threat to rentier interests than was Marx.

[Ronald]

Isn't it possible that this is a red herring?

GDP(I) ignoring errors and taxes less subsidies on production etc. is equal to: COE + GOS. GOS consists of dwellings, private and public non-financial, financial, general government and gross mixed income.

The time series refer to post 1970, but if you look at a longer time series COE is relatively stable at around 50%. It rises in the late 60's to 70s and then reverts.

Surely isn't the question why did it spike in the 70s? As opposed to why is it falling now?

Both US and Australian data have the same trend: up in the 70s and falling back thereafter.

Maybe it's because of stagflation? COE rising due to inflation and expected inflation and unions having greater bargaining power, and stagnating GDP growth.

[Nathanael]

Ah! Mr. Rowe, I have an idea!

Infrastructure rents. Consider what your Internet Service Provider charges, or what your electricity distributor charges.

While these are not natural resources -- they are, in fact, capital goods -- they are not conventional capital goods and increasing returns there would direct high returns to a tiny tiny segment of the economy. And one which is hard to expand. The people controlling the *existing* capital investment would reap higher returns, but anyone who tried to duplicate it with *new* capital investment would be at a massive disadvantage -- since this sort of stuff is natural-monopoly material.

So, the incumbent has a huge advantage against any potential competition -- the telecoms installation was mostly done 50 years ago and/or funded by previous company's bankruptcies -- but can charge rents which correspond to the cost of building an entire duplicate system from scratch, plus more due to the natural-monopoly effect.

[Nathanael]

Richard Serlin, I always love your comments at every blog: you are highly empirical.

You wrote: "Doesn't seem like real expected stock returns have gone down, just the risk premium (or perhaps falsely perceived amount of risk) has shot up."

What's going on here, I can explain, being an individual investor. The fraud rate -- specifically, frauds perpetrated by executives against stockholders -- has shot up, and the government has abdicated its role of policing for fraud. Therefore stocks are MUCH riskier.

The company may have built a giant scam, as Enron did, and as most of the major banks did. Even if the company's underlying business does well, that's no guarantee the stockholders will get anything at all, because the entire thing could be looted by the execs, as happens routinely.

Expected stock returns *when accounting for the percentage of companies which are frauds on the stockholders* have dropped. Before accounting for that, they've stayed the same. Is that clear?

This is the general consensus view of actual investors.

[Nathanael]

I like Will's point that legal restrictions can create other things with the same characteristics as land. For instance, cabling to people's houses: it is practically impossible to assemble the right-of-way without government assistance. Like railroad right-of-ways, it's land, more or less.

[Greg Ransom]

Thanks Nick.

On 'capital', "land" and "labor" -- I can't imagine what someone using the relational logic of choice at the margin can be thinking when attempting to salvage the these gross backward viewed, objectively/physically measured categories which as categories have no role in the forward oriented choices of anyone.

Many of the economically relevant characteristics of land are improvable or used up like other production inputs. And whereas many of the economically relavant properties of land are essentially permanent, the logic of the valuation of oil is different because that is a an inherently non-permanent consumable resource -- it's completely used up.

In a fully worked out logic of relational valuation (marginamsim) extended to production inputs, the backward looking and objectively/physically measured aggregate categories of "land", "capital" and "labor" are not observable as units of valuation and input and choice.

Completing the marginalist revolution by extending it to heterogeneous production goods which are either permanent or used up is hard -- but an ertzats partial return to Ricardo and a non-marginalist logic of valuation using objectively measured valuations with no relation to future anyone's future oriented plans is, well, incoherent.

Nick writes,

"I'm surprised nobody has mentioned oil yet. Oil (in the ground) is land. So is copper ore, etc.

Being a farmboy, "land" to me is farmland. Some of you city kids quite rightly say that "land" is city centre land too. But oil and copper and iron and hydroelectric etc. are all land."

[Ritwik]

K

The paper you cited quotes this :

"For example, as we observe later, the value of land
is influenced significantly by its level of preparation for building.
If the earlier years of our sample were dominated by sales of raw
land and the later ones by more finished parcels, then the figures
would overstate the “true” growth in land values over the period
by reflecting in part the value of site preparation."

It goes on to claim that land for immediate development fetches much higher prices than that for investment, and so on and so forth.

Which raises the age-old question for Georgism - what is the unimproved value of land? And what part of land's value itself can be said to be improvements, or capital? I don't think it's enough to say that ownership of land benefits from increased economic activity which is solely the creation of others' economic activities and thus should be taxed at 100%, until we have a good way of delineating 'improvements to land'.

[Ritwik]

K

India has about 15 trillion sqft of agricultural land. This yields about Rs. 15 trilion, or $300 billion of agricultural income per annum. So, about 2 cents/ sqft. The land itself retails at, on an average, about 40 cents/ sqft. As you can imagine, the average Indian farmer is ridiculously poor, even those who actually own some land aren't doing much better.

The govt has eminent domain, and pvt individuals can't sell agri land for non-agri purposes without the government consenting. While initially this was done to prevent landlords from selling off their land to industrialists, leaving the actual tillers of the land in the lurch, the law has had the usual unintended consequences.

In one of the more celebrated agri to commercial land sell-off deals where the farmers actually got their due, a pvt firm acquired agri land to set up a car factory at nearly $4/sqft, about 10 times the ongoing rate of agri land. The state govt played a largely benevolent role, getting a good deal on behalf of the farmers. Needless to say, many previously poor or poor-ish farmers became rich overnight.

From a Georgist perspective

1) what would you say is the unimproved value of the land? $0.4/sqft? $4/sqft? Something in between?

2) How would you tax the gains made by the farmers? At what rate, and on how much of the deal value?

[J.V. Dubois]

Nick: ""I'm surprised nobody has mentioned oil yet. Oil (in the ground) is land. So is copper ore, etc."

This is very true. For instance Saudi Arabia produces approximatelly 12 million barrels of oil a day with extraction costs of approximatelly $2.80 a barrel. Let's say that the price of the oil is $102.80 It is easy too see that Saudis extract rent from a world economy on the scale of 440 billion dollars a year. Just Oil from Saudis can capture 3% of US GDP

There is also a whole world of rents that hide as regular labor or capital. If we are speaking about labor, there are professions that outright limit an entry into the market by licensing. Then there are "star" professions - there is only so many top baseball players so in a sense that tiny group as whole group will have it all. The same goes for other professions. For instance there is only so many people that can have personal ties to important people in government with power, those who happen to have that access will be greatly rewarded. You may object that there is competition in these positions for superstars, and lobbyists - but in that sense there was a competition among medieval aristocrats. It is when these closed groups start to capture more and more of resources when you start calling them nobility.

If we are talking about corporate rents, then you have vast companies that are plainly regulated - that are negotiating the price of their goods with government and they call the rent as "adequate profit". Then you have companies that were first in some important areas and they already have large market share and market power. If the goods they sell are veblen goods, or if (as is the case of branding) they sell "an experience" - a psychological need, status etc - or if they misuse information asymetry - then I have a hard time calling their investment in marketing as "capital". In the very same sense you cannot call a costly military coup in Saudi Arabia as "investment".

[Nick Rowe]

K: good find. Hmmm. That $120 Trillion for total NYC land values does sound very high! But we need to convert it into a rental. And that depends on r and g (the growth rate of rents). Gordon equation says $120T = annual rent/(r-g). Suppose r-g=1%. That's $1.2 T rents. Man that's big! Around 8% of US GDP??? Just for NYC?? That must be too big.

[Tim Worstall]

It's also worth considering that there's something wrong with an assumption here.

I don't know enough about US GDP statistics to even find the information. But I have done this with UK such.

Yes, the labour share has been falling. But it is not true that the inverse of the labour share of the profit share. The way the EU measures these things there are four pieces.

Labour share (within which is wage share plus employer paid taxes on employment)

Profit share

Mixed income (sole proprietors and self employed)

Taxes minus subsidies.

When we look at the UK figures

http://uneconomical.wordpress.com/2012/04/23/uk-gdp-by-income/

We find that profit share has indeed risen from the mid 70s (but arguably that was entirely unsustainable) but not since 1980 or so.

What has risen is mixed income. Well, we've more self employed now than then so not a huge surprise.

And taxes minus subsidies. Well, VAT came in in 1973 or so, at 8%. It's now 20%.

Within the labour share, the wage share has fallen as national insurance (think social security) paid by the employer has risen substantially.

The labour share has indeed fallen. But it seems to have mostly been because of the effect of higher taxes.

Whether this is true of the US or not I don't know. But I do think it's something that someone should check. For the profit share just ain't the inverse of the labour share. And it worries me when everyone is showing us the labour share in order to talk about the profit share.

[K]

Nick: "That must be too big."

Apparently it was! It was the peak of the "housing" bubble.

From the table on page 4, the price in 2000 was $55/ft^2 giving a total value of about $18Tn, or about $180Bn in rents using your figures, or about 1.8% of year 2000 GDP. Off the top of my head, that seems conceivable as an equilibrium value given the importance of that region.

But I'm less skeptical than you about the 2006 value (though it obviously didn't reflect the PV of future rents). Apparently Japan in 1991 was worth $20Tn, or 20% of the world's total wealth. The land of the emperor's palace alone was worth as much as all of Canada. It's not like anybody is actually earning all those imputed rents. It's just paper value in our collectively deluded imaginations.

[OGT]

Great post. My first thought when I saw the Krug post was the coordinated twenty year fall in real interest rates across the developed world. But, that's mainly because I have a bug about that process. It seems really interesting and potentially pretty important, but, there haven't been many papers I have seen that try to explain it.

Could it be twenty years of opportunistic disinflation? Savings glut from rising developing countries without developed financial sectors? (Brad Setser once wrote the US has a comparative in creating debt). Or a secular technology driven change?

[Mike S]

Why wouldn't capitalists and workers say "Stuff that, let's go back to the old technology!"?

Isn't this exactly what happened? China did this and ate our lunch. The use 1980's technology, low interest rates, and cheap labor to extract demand from the United States.

[dwb]

no, there is another factor which i have direct experience with: taxes. Take the example of friend A, who works at a small start-up service based business. Friend A works very hard and travels a lot to build the business. When the time comes to get paid, the company, which would rather plow cash back into the business, defers the "bonus" and instead pays her in securities with a very low tax basis, whose gains will eventually be taxed at capital gains or dividend rates. There are many ways tax-avoidance scholars can effectively game the tax code to pay compensation as dividends and capital gains to lower the tax bill (any wonder "compensation share" dropped precipitously after the Bush tax cuts on dividends etc.). If I start a consulting company, I can reclassify my wages as "revenues," deduct all sorts of things (and avoid a bunch of payroll taxes) and then dividend out some of the "profits" after a suitable amount of time. Small businesses account for a significant fraction of gdp in the US, most are service based with little capital, and if you look into it you will find a lot of entrepreneurs getting "compensated" for their hard work in tax-favorable ways.

[K]

Nick,

"Suppose there were a new technology that caused both wages and interest rates to fall. Why wouldn't capitalists and workers say "Stuff that, let's go back to the old technology!"?"

That doesn't seem right to me. The old technology doesn't just continue to generate returns. The excess returns are monopoly rents that accrue temporarily to innovators by virtue of trade secrets, legal IP protections or just first mover advantages. Only new technology can generate returns *from here*.

[K]

and only new technology can produce productivity gains from here!

[OGT]

Great companion piece from FT Alphaville, worth a read:

http://ftalphaville.ft.com/2012/12/10/1303512/the-robot-economy-and-the-new-rentier-class/

[libertaer]

"Like how could Marx get both increasing immiserisation of the proletariat and a falling rate of profit? Unless land rents were rising, like In Ricardo."

Maybe this is all wrong, but I always thought that immiserisation in Marx was due to unemployment caused by a lack of aggregate demand. (I read Marx as a pre-keynesian based on his "Theories of surplus value". You can already find there the insight that "recession are always and everywhere a monetary phenomenon".)

Raising demand by monetary policy doesn't work for Marx, because he had no place for a central bank. Money in Marx is supplied by private banks which use the assets of borrowers as collateral for their loans. If profit expectations fall because of uncertainty or secular stagnation or whatever, the value of assets as collateral for loans falls too, so the only way of raising the money supply is closed. Nobody can borrow, everybody has to pay back their loans. Money will reflux to the banks (law of reflux). Today central banks can intervene, pushing asset prices up again. Using fiscal policy (helicopter drops) we could even have an monetary economy without any capital goods at all, just workers producing and exchanging consumption goods. For Marx this was impossible, capitalism is all about capital goods because money only gets created by using capital goods as collateral. No demand for investments = no money, people can start to barter again.

The creation of a national central bank was one point of the ten points of the communist manifesto. So if today a lack of aggregate demand isn't a problem anymore, that's because everybody is already communist :-)
On the other hand, as a European I'm not so sure, the ECB seems to keep capitalism alive and kicking.

[Robert]

Pakko's paper is really a memo. But it makes the total compensation point.

Michael R. Pakko, 2004. "Labor's share," National Economic Trends, Federal Reserve Bank of St. Louis, issue Aug.

[jt]

Would China be a good example of this? Cheap financially-repressed Chinese labour are the "robots" and the Chinese government subsidizes its own infrastructure/housing and developed countries' housing (over)investment (irrespective of ROI). The resulting increase in commodity prices and the associates rents are collected by corrupt Chinese officials/businessmen and commodity companies?

[Wonks Anonymous]

Anon, where did that 1 trillion dollar figure come from?

It is true that buildings can expand into the third dimension. Physically that is, not necessarily legally. There are often zoning rules which restrict the heights of buildings, or density more generally.

[Sandwichman]

I'm with you on the "oil is land, too" point, Nick. Before blaming Marx for the disappearance of land, you should have a close look at "The Trinity Formula" in Volume 3 of Capital (part 7, chapter 48). Note that the whole of part 6 is on "the transformation of surplus profit into ground rent." Note also that Marx does not deny but affirms the classical theory of ground rent and attributes it to its rightful originator, James Anderson (he accuses Malthus of plagiarizing Anderson). But Marx also supplements Anderson's theory with the proposition that surplus profits are also converted into ground rent. That is to say, not all ground rent is derived from the inherent productivity of the land -- some of it (and at times much of it) is due strictly to the exclusivity of the land and its consequent ability to act as a lever for extracting surplus value from labour.

Some interesting avenues of analysis open up on the basis of "oil is land, too", Marx's critique of what he called the Trinity Formula, his critique, in volume one of Capital of the "Theory of Compensation as Regards the Workpeople Displaced by Machinery," Stanley Jevons's "paradox" of fuel efficiency leading to increased consumption of fuel and Maynard Keynes disavowal of the notion of the "self-adjusting economic system." Throw in Robert Costanza's analysis of embodied energy (1980) and I think we're getting somewhere.

[Sandwichman]

By the way, there was a significant error in Engels's presentation of the chapter on the Trinity Formula, which became obvious in retrospect after Marx's notebooks were published. Engels had mixed up the sequence of three sections of the chapter making it appear there was something else missing. So when you read the chapter, it would be important to track down the commentary on what the right order should be.

[Paul Friesen]

Shouldn't we expect a low return to capital combined with an increasing share paid to capital as businesses move "offshore"? Labour-intensive manufacturing moves first, leaving more capital-intensive industry behind. The increasing share just follows from the nature of the activity left behind. The low returns are because competition and potential competition from abroad force firms to keep prices down.

[Jacques René Giguère]

Sandwichman: exactly. Central NYC, LOndon and Geneva plus Aspen, Gstaad and St-Moritz are where monopoly profits go to die and rise to the 55th floor of corporate headquarter and condo heaven ( along with a couple of russian nieces).

[Michael Hollander]

FYI, there's been a significant technology rollback in finance - securitization is completely dead.

[Lord]

Land in small towns is perhaps 10% of property value, while in a metropolis, say one million, it exceeds 50%, and in a megaopolis exceeds 80 or 90% of value.

[K]

Michael,

Technology, in economics, means something that *raises* productivity. CPDOs don't count!

[K]

Ritwik,

You could impose a capital gains tax on land value. Henry George proposed to continually confiscate the land rents (Adam Smith also pointed out the efficiency of a land rent tax). Charging the short rate on the assessed unimproved land value would do the trick. Also, long term land leases from the state, like in Hong Kong, would work. Maybe 30 or 50 years?

"How would you tax the gains made by the farmers? At what rate, and on how much of the deal value?"

All of it! I know it sounds cold hearted when applied to Indian farmers, but there are lots of people in India who are worse off than the farmers. The point is, those farmers had *nothing* to do with the value of that land. They were the beneficiaries of the monopoly use of it for a period of time and then, suddenly, the economic activities of *other people* made that land very valuable. So who should get the windfall from that positive externality? I say the public. Pay a citizen's dividend, spend it on something useful, share it!

[Frank Restly]

Nick and K,

The article K references indicates that the survey group looked at about 6100 land deals between 1999 and 2006. What the article does not mention is how much property was involved in those 6100 land deals. Suppose K is right on his 11,842 square miles of available land. But suppose only a couple of hundred square miles of land was bought and sold during that time. Can you honestly compute the value of 11,000 square miles of land based upon 200 square miles of sales (less than 2% of total land)?

"You could impose a capital gains tax on land value."
Unless you are implying that the government should take a chunk of the land itself as a tax, I am not sure how you would reasonably accomplish this. Taxes are payable in the government's money and so to establish the value of the land with respect to the government's money, that land must be bought and sold for the government's money. If that land is highly illiquid (is not bought and sold very often) then determining that value is a shot in the dark.

[K]

Frank, Nick,

I had a closer look at the definition of the NY Metropolitan area and it seems the area in the paper is only about 35% of the area as defined by Wikipedia. So more like 4000 square miles. So only about $40Tn at the peak and $6Tn in Y2K. $400Bn in imputed rents is a more reasonable 3% of the US economy.

As far as the statistical significance goes, there were over 6000 transactions in the sample, which is generally plenty to rule out random error. I don't think the fraction of land area sampled is statistically relevant. It's really all about the number of independent samples.

[Will]

K:

Experiments with Georgist land-value taxes have not gone well in the Pennsylvania cities where they were tried. The landowners soon captured the assessment office and persuaded it to keep assessment values lower than actual values, according to a recent Harper's article about the origins of the Monopoly board game in George's thought: http://harpers.org/blog/2012/10/monopoly-is-theft/