« Teaching, OLG models, and the phenomenology of perception | Main | Do economically illiterate slobs use duality theory? »


Feed You can follow this conversation by subscribing to the comment feed for this post.

Good post. I like the idea of education as an investment in "human capital".

Branko: "For me to get an income from my financial capital, I do not [need to work]."

Does he not miss the point that in order to acquire financial "capital" (or better IOUs as you pointed out) you need to work first? Only when the real return of that IOU is > 0 then you will get income not derived from your own labor (assuming infinitely lived agents).

Depends what you want to do. Human capital is useful if you want to downplay certain (conflictual) distinctions and emphasize (harmonious) others. It has no transcendent validity and those who pretend it does are exercising pseudo-scientific ideological imperialism.

Exercise for the student - or the professor in this case:

Integrate with Piketty


"No. This misses the point. Some land is more productive than other land, but there are two possible reasons for that difference in productivity; it might be due to greater investment; or it might be due to the land itself. "

As urban dweller, the main thing I see is an increase in the "productivity" of the land due to higher income people moving into the neighborhood. E.g. wealthier people move in and rent prices go way up. Same exact piece of property. A lot of this is due to clustering -- e.g. economic geography -- which is essentially random. No aspect of the land has changed at all, other than the adjoining pieces of land being occupied by higher income individuals. And homeowners are very sensitive to what happens outside of their own lot but in their neighborhood, creating HOA's, for example. This aspect of locality is critical to understanding land.

For Humans, the corresponding attribute might be having a certain social network -- e.g. socially close to higher income people.

Very well done! This is why I say such nice things about you!

Good point rjs. This ties in to Jane Jacobs work on what cities are all about, I think.

Yes financial capital is just rent extraction (rent on ownership of money like ownership of land).
You can make exactly the same argument surprisingly with a temporal labour theory of value (by temporal I mean the version exposed by James Mill and the elder Torrens) where time is discounted by the interest rate, and avoids the ricardo capital problem. In this you trade labour time today for discounted labour time tommorrow - over a full working lifetime, exactly as you would do if you were constructing fix capital which depreciates. Ill do a spreadsheet math example on my blog which shows exactly the trade off between time invested in developing human capital today and future income over a lifetime. Interestingly towards the ZLB it becomes rational for many who can afford it to give up work to study leading to less production.

Financial capital absolutely exists from a private perspective, even in worlds without investment. It just vanishes at a social level by netting out. I don't think the term is unreasonable from the perspective of an individual.

Above comment was not from Gene Callahan but me - worrying that Wordpress can let you comment as someone else - bug

Nick, from Becky Hargrove (who was unable to link it here)

[Marcus: I found you in the spam filter too! But Typepad can learn, and by fishing you out of spam I am training Typepad! NR]

Odie: thanks. I do too, but it's Adam Smith's idea. (Unless someone else thought it up first, which they probably did.)

rjs: agreed. Or, more generally, whether my land or labour is valuable depends a lot on other people's preferences, as much as on my land and labour itself. I always tell my students this example: suppose Wayne Gretzky had been born in England when I was. He had the exact same natural skills, and did the exact same training to develop those skills. People would have said "What a great skater; but what is he going to do for a job?" (Of course, he probably wouldn't have invested so much in improving his hockey skills if he had lived in England.)

But this (wandering slightly off your point) makes me wonder about "social capital". It's clearly important, but is it really "capital"? Maybe, maybe not. In some cases yes, because it needs an up-front investment? Dunno. We do talk about "building reputational capital", which looks like a correct use of words.

Sandwichman: looks like you have nailed that Adam Smith guy!

JKH: Google finds for me a good post by Karl Smith on Piketty and land capital. Not sure if Karl is using the concept in quite the same way though. More like "land wealth".

Gene: thanks!

D: If people care about the timing of consumption (which they do), and if some very long-lived asset like land exists (which it does), we are going to see positive interest rates. And if consumption loans exist, we are going to see what looks like income on those loans, but it isn't really income. If I swap 100 apples for 105 bananas, we don't call those 5 bananas "income".

Alex: OK, but I would say that the concept of "financial capital" rests on a fallacy of composition.

Fedupguy: whatever.

At the very end of his post, Branko Milanovic urges that the word 'capital' retain it's traditional definition "for property distinct from one’s labor."

Of course there is a lot of property types that are distinct from labor but they all have, in common, the property of ownership. All property is owned by someone(s); a fundamental truth that is a result of human capabilities. Ownership is always the outcome of human labor exercised in some fashion.

While capital is "property distinct from one’s labor.", it is also the result of one's labor as you point out. Land must be cleared (which requires labor), fertilized (which requires labor) and harvested (which requires labor). Harvest, the last step, represents a considerable investment of 'capital' which illustrates that while the word 'capital' does not mean labor, 'capital' can be the result of labor .

In the last paragraph of this post, you urge abolition of the term "financial capital". This makes no sense to me. Financial capital is simply the result of labor stored in a financial manner. Bonds are financial capital as is money. The existence of both is the result of labor performed by both the present owner and the seller who originally created the bond or money.

To conclude, you can see that the term "financial capital" has a clear meaning to me.

"Before you can grow wheat on it, you usually need to clear it, or drain it, or fertilise it. There's an up-front investment." (Yes. It's called human capital.)

"The rents on land are as much a return to capital as they are a return to raw land." (Yes. Just as the produce are as much a return to land as they are a return to human capital.)

Abraham Lincoln: "Labor is prior to, and independent of, capital. Capital is only the fruit of labor, and could never have existed if labor had not first existed. Labor is the superior of capital, and deserves much the higher consideration."

Odie and Nick - "in order to acquire financial "capital" (or better IOUs as you pointed out) you need to work first" is what you've noted above.

Endogenous money. ---- According to the Bank Of England the majority of money is created ex nihilo by banking institutions. So you can get money from a bank with a good idea, and neither side, lender or borrower, has to have earned the money first in order to establish the contracted relationship reflected for the borrowing in the lending account established for them.

Look at the current worldwide trading in debt (larger marketplace than for stocks) and the rest of the OTC marketplace (the contract values of derivatives alone rivals the size of the stock market), these grew explosively to enormous size almost in the last 15 years or so. Do you think the financial underwriting for participation in this explosion was all "earned" by participants in the real economy (and where was that happening?)? Since these marketplaces more or less exploded in the period 2000-2007, the financial earnings were all earned in some short period just before 2007? I think everyone needs to think a whole lot more about endogenous money.

Otherwise love the post. I use the term human capital as a communications frame when I speak about the US social security program and the implications of the fact that productivity gains in one generation almost all flowed to somewhere else other than into wages/salaries then, especially when considering that the period just before gains had been equally shared. Social Security draws from the current economy its "earned" income from human capital and should be paid the proper return (again, the labor market was screwed up then and some people took most of the gain, an indication of a rent-seeking result to me (it is true that humankind did not change much somehow in 1979 to cause the change for natural reasons).

Roger: "At the very end of his post, Branko Milanovic urges that the word 'capital' retain it's traditional definition "for property distinct from one’s labor." "

If I own the ores (or the mining rights to those ores) under an area of land, I own property distinct from my labour. But that is not "capital" in the traditional sense of the term. It's "land" in the traditional (Ricardian) sense of the term.

"Financial capital is simply the result of labor stored in a financial manner. Bonds are financial capital as is money. The existence of both is the result of labor performed by both the present owner and the seller who originally created the bond or money."

No. It takes me labour and land to produce apples. But it takes me (almost) no labour or land to write "IOU 100 apples next year signed Nick Rowe" on a bit of paper. Same for the Bank of Canada writing "$20" on a bit of paper (I mean plastic).

Bud: "Abraham Lincoln: "Labor is prior to, and independent of, capital. Capital is only the fruit of labor, and could never have existed if labor had not first existed. Labor is the superior of capital, and deserves much the higher consideration.""

Abraham Lincoln is talking nonsense. Replace "labour" with "land" in that sentence. And there wouldn't be any labour if parents hadn't invested in feeding their children. Chicken and egg. And land was here for billions of years before both, and neither could have existed without land. Therefore land is the superior, and deserves much the higher consideration?

Land prior to investment is like a license. You cannot put a mine on it without a license. You cannot be a farmer without a license. You can buy a house but you cannot put it anywhere without a license (a building lot). The license conveys the right to invest, what we call land purchase costs.

"What we call "land" is as much capital as land. The rents on "land" are as much a return to capital as they are a return to raw land."

I disagree. The outlay to make land productive is capital, even if it's something land-ish like topsoil or fertilizer. The outlay could have been used elsewhere to produce something else. The rent of land is a passive, spontaneous flow thanks solely to the parcel's position relative to other parcels. The literature resists it, but a better term for land rent might be location rent.

Apologies if this isn't germane to the argument on human capital.

john: there's land, and then there's the right to do things with that land.

LSTB: location (of land) matters, but it's not the only thing that matters. For building land it's very important, but for agricultural land it matters less (though it still matters). Some land is more fertile than other land, even with the same investment. Some is flat, and some is too steep to be ploughed.

I feel that the analysis is not deep enough. "Capital" is used in common parlance for anything that reduces the spacetime costs of doing something, which includes making connections, and gathering something. (Improvement or training to make it better is secondary to the definition, not primary.)

This covers a huge variety of things. The continuing confusion of economists is that they tend to think these things should be monetizable for fungibility. But the basic definition already shades-off into things which are sometimes money-fungible and sometimes not. Examples:

"Industrial machinery" speeds the process of manufacture by greater precision in mechanical operations (Babbage).

"Housing" reduces the costs of sheltering from the weather, costs of gaining privacy, costs of storage of personal objects, etc.

"Human capital" is judged by the ability to make quicker useful choices in a line of work through knowledge, skills, talent (e.g labor, entrepreneur).

"Property" reduces the time to verify transactions, make contracts, etc. (John Locke, baby!)

"Scientific theories" reduce the effort of ratiocination to organize the word intellectually, by use of language, equations, etc. (Ernst Mach, "The Economy of Science")

"Farmland" has the condition of making crops more productive, easier to plant, to grow, to handle, etc.

"Financial assets" are ranked by ease of storage, likelihood of valuation through time, ease of fungibility (a.k.a. liquidity). Thus bonds, gold, real estate, etc.

"Government programs", properly understood and executed, reduce the personal and social costs of welfare enhancement (e.g. retirement security), and of correcting for market failures (e.g. security, environmental protection, basic R&D funding).

"Cities" reduce the spacetime costs of transactions for market participants, etc. (Adam Smith Chap. 3) https://www.youtube.com/watch?v=TkXTL92WfFA

And there are others. In general, being a recipient who is placed at, or near, the center of certain formations is also often a form of capital, such as inside: 1. sociopolitical structures, 2. manufacturing industries, 3. geographic clusters, 4. financial markets (e.g. bankers).

The common theme is the greater value of the reduction of the cost, in space or time, of a transaction or a transformation (such as by industrial machinery), under the purview of the thing which reigns as the capital in that circumstance. Whether it is monetizable or not.

Economists could try to draw a line between the ones that are monetized (and thus interfungible) and those that are not, but I imagine that won't be good enough for growth theory.

The analysis needs a paradigm-shift. Much better just to divide the world into two kinds of things:

1. Individuals, ideas, inventions, technology, prime locations, buildings, institutions, "capital". These each reduce spacetime costs of:

2. Transactions, transformations, communications.

Capital is use
not a thing or person
As in a thing used as capital
To be used as capital a person or thing
must exist in a certain social context

Nick: Thanks for mentioning the ideas of David Ricardo. A quick Google search turned up his 'law of rent' found at


I notice that his writing in no-way detracts from the meaning of capital as a reservoir of labor earned. On the contrary, he ties together productivity of land, availability of free land, and wages. A capital value of land will result from those parameters.

I assume that Canada and the United States both treat land as property of the state. It follows that if both 'own' the land, both entities are able to rent the land to private parties. In the United States, we call this rent a 'property tax' and realize that 'owners' must pay this property tax or risk loss of the lease (ownership). The capitalized value of any property will include this property tax as one of the fixed costs of ownership.

So far as I can see, the Ricardian treatment of land rents is completely consistent with treating capital as "property distinct from one’s labor."

Has any economist ever actually studied physical capital? (Machinery, for instance?) Yes, "human capital" and "land capital" are similar, but has any economist since Marx ever seriously studied the means of production?

I'm being provocative. The environmental economists have done so. But my point is, most economists jump directly to financial capital -- they put a dollar value on the "capital", act as if it can simply be purchased, and proceed to ignore its true character.

There is relatively little literature on the role of physical capital in the economy. And isn't that *weird*?

...I see that Lee Arnold has made my point better than I could.

' "Capital" is used in common parlance for anything that reduces the spacetime costs of doing something... The continuing confusion of economists is that they tend to think these things should be monetizable for fungibility. But the basic definition already shades-off into things which are sometimes money-fungible and sometimes not. '

What Lee Arnold said. But also note the assumptions in "There's an up-front investment, which is costly in the "opportunity cost" sense, because it uses resources that could have been used elsewhere to produce something else." That something else would also have been an investment, no? So what's the "cost"? If I am typing this blog comment, I am not being a carpenter, so am "losing" $50 an hour? Or not being a banker, in which case I am making the world better of by $2000 an hour? If we are to deal in counterfactuals, we need some way of deciding which counterfactuals. Second, the alternative to "investment" is this sense is not consumption, it's being dead. No living creature gets to eat without working for it. If I eat apples, I first have to "invest" time in picking apples. There is no alternative in which apples leap into my mouth. So working is not an "investment" in the sense of deferred consumption.

The category of investment only makes sense within the context of a set of social arrangements. It has no applicability to production and consumption considered as wholes.

My argument agaisnt the name "human capital":

"human capital” can’t be sold, loan, rented, mortgaged, etc., without selling, loaning, renting, mortgaging your labour also, unlike occurs with capital in classic use of the word (then “human capital” is more a subtype of the category “labour” than of the category “capital”)

In other words - in the production process, the owner of the "labour" has to be the same person as the owner of the "human capital"; this not happen with "land" and "land capital" (I think that are not rare cases where a land belong to one person but a building in the land belongs to another)

Miguel: I can rent out the services of my labour, my land, and my machines. I can borrow against my labour income, land income, and income from machines. I cannot (nowadays) permanently sell my labour, but I can permanently sell my land and machines. (Though long term contracts, where I promise to play hockey for the Ottawa Senators for the next 4 years, come close, but are hard to enforce.)

Labour is like owning a house in trust, where you can live in it, or rent it out, but cannot sell it. But it's still capital.

@Nick Rowe:

> Labour gets disutility from working (at least at the margin, though some of us actually enjoy our jobs, and get positive total utility from working even if we get negative marginal utility from working an extra hour); land doesn't.

Is this necessarily true? It seems to me like the disutility from working is exactly analogous to the operating costs of working land. My highly productive orchard will still not actually produce anything unless I provide labour (my own or hired) to pick the trees.

Majromax: producing apples requires both land and labour. (It requires capital too, because you need land and labour to plant the trees, and apples only come a few years later, but let's ignore capital). The land has an opportunity cost, because it could be producing pears instead. The labour has an opportunity cost too, because it too could be producing pears instead. The owner of the labour might care whether his labour is producing apples, pears, or is idle, quite apart from the income earned. Does the owner of the land care about whether his land is producing apples, pears, or is idle, apart from the income earned?

(The idle labour might be producing leisure; the idle land might be producing leisure too, because some of us land-owners get pleasure from wandering around our land.)

"I can rent out the services of my labour, my land, and my machines."

But you can't rent your human capital without renting your labour.

An important difference is that the distribution the ownership of the "human capital" is more important than the distribution of "non-human capital" - assuming a finantial and/or a market with no or few transactions costs, it is irrelevant, for the global production of the economy, who owns the capital (a worker can work with the capital of another person, both by the system of the worker hiring capital - renting a shop; taking a loan to buy a machine, etc. - or by the system of the owner of the capital hiring the worker)

In contrast, the distribution of "human capital" is very relevant: 1 person with 10 degress e 9 persons with no degree are very different than 10 persons with one degree each.

Nick: "But it takes me (almost) no labour or land to write "IOU 100 apples next year signed Nick Rowe" on a bit of paper."

I think the point was that presumably you made that promise because the person to whom you gave that IOU gave you, say, 75 bananas in exchange for it.

That is, his asset (the IOU) was earned by his labour expended in producing 75 bananas with which to buy the asset. Or perhaps even more directly, he was employed by you for 40 hours, and you paid him in 100-apple IOU's.

I think the distinction which you want to make between "real" capital and financial capital might be that while financial capital might look like real capital at the micro level, to individuals, from society's perspective, it does not directly increase production.

That is, if I work, and exchange my labour for an IOU that yields 100 apples a year from now, from my perspective that looks like working and building a machine that will produce 100 apples a year from now. If I had actually done the latter, society actually will have 100 more apples next year. My buying an IOU from you, though, does not create 100 more apples.

I actually have another beef with what Branko says. I thought about it when considering Scott Sumner's idea wage tax equals consumption tax. The basic idea is: What is Labor?

Example 1: Everybody knows one example that is hard to explain in "labor theory of value". Suppose you go around and suddenly you find the most expensive diamond yet. Just few seconds to pick it up and you are rich. Theory goes that it is OK because diamond is valuable and to find similar one there is a lot of searching labor required ON AVERAGE. Finding and selling diamonds is labor income, not capital income. So even though you lucked out and found the biggest diamond on the very first day you tried this work, you sold it and now do nothing for the rest of your life it is supposedly a fair return to your labor.

Example:2 Suppose you work as a trader. Maybe you buy and sell houses. You laboriously try to find houses that are cheap and sell them when they are expensive. For average person there is not much income in this. But suppose you really lucked out or you are really talented and you were able to close the deal of your life. You are now golden and do not have to work anymore. That's fine, it is just fruit of your labor. Or is it? Isn't selling house for more than you bought considered capital gains? But what about all that work that you put into finding out which house to buy and to sell?

In reality there is virtually nothing that can be considered as pure "capital income". Even management of one's wealth requires some labor - at least you have to hire or approve somebody who will do it for you. And as many people especially those who won lottery and spend it quickly may attest keeping on wealth can really be hard work. I agree with Nick, capital and labor have a lot in common.

Your basic notion of the similarity of land, labor, and capital was best articulated by Frank Knight in his theory of capital and rejection of the tripartite distinction between land, labor and capital. See his "Ricardian Theory of Production and Distribution," which was published as the very first article in the first issue of the Canadian Journal of Economics and Political Science in 1935.

Of course, few economists have ever agreed with Knight, but among those who were fans was T.W. Schultz, one of the authors of human capital theory.

Ross: interesting connection.

A couple of years back, partly for fun, but partly seriously, I did a post on "Dutch Capital Theory" in which "capital" is just the name we give to land that we created ourselves. But suppose we lost the history books, and couldn't remember which land we had created ourselves and which land God had created. It wouldn't matter. The distinction between capital and land only matters for current decisions looking forward, if we can create new land.

But my "Dutch Capital Theory" was different from Knight in having a more subjectivist than productivity theory of the rate of interest.

I've always thought the proper distinction had to do with the possibility of accumulation (so, for example, the "capital" in say, the Hecksher-Ohlin model is not really capital, it's just land for which we use the notation "K"). You can use machines to make more machines. You can't take wheat and turn it into land. You can accumulate "skillz" so human capital is capital.

notsneaky: You can accumulate "skillz" so human capital is capital.

This to my mind is one of the crucial flaws in the "human capital" concept. There's a limit to how much skill a musician, even a genius, can accumulate. Granted Mozart might have died rich if he'd been a bit more thrifty, but his wealth was never going to rival the aristocrats' no matter what. But there's no meaningful limit to the pile a Rothschild can amass.

notsneaky: "so, for example, the "capital" in say, the Hecksher-Ohlin model is not really capital, it's just land for which we use the notation "K""


But if we forgot that we had made the existing stock of capital goods, and imagined they were given to us by God, and were land, would it make any difference? As long as we still remembered how to make new ones?

And if technology changes every period, so the new machines we make are never the same as the old machines, it's just the same.

Kevin: there is diminishing returns to "capital", if you hold labour constant?

Of course, it depends on how many kids and grandkids you have ;-). I am thinking of writing a post trolling Elizabeth Breunig (she deserves it, for calling me "evil") on Babies as human capital in an OLG model. My initial thought was that saying new-borns are "human capital" is an abuse of the term, because they haven't learned anything yet. But then I remembered reading "Larks Rise to Candelford", the memoir of a poor village in 19C England. Only one couple in the village had any saving, and they were the only childless couple, and they needed a pension. Everyone else invested in children, who were costly to feed and clothe, but would support them in their old age. Their children were their pension plans.

Nowadays we don't expect our own children to support us like that, usually. But in aggregate it's still true. Our kids will be working to produce the goods we cannot store for our retirement. No kids no saving. 3 generation OLG model.

Nick: there is diminishing returns to "capital", if you hold labour constant?

Exactly; and there's just so much labour that the household can supply. But the Duke of Westminster doesn't have a problem finding builders and estate agents to make optimal use of his property portfolio. So human capital differs in an essential way from the kind Thyssen accumulated.

Of course whether this matters depends on whether, like Piketty, you're exploring questions of wealth distribution, or whether you just want to do a growth-accounting exercise. For the latter you're going to need some variable that captures the evolution of skills and you can call it human capital if you want; it sure sounds better than "a blaze of amateur sociology."

The comments to this entry are closed.

Search this site

  • Google

Blog powered by Typepad