Some Background on Capital Controversy
I am going to be putting out stuff in the coming month on why Thomas Piketty’s use of a neoclassical aggregate production function is so dubious–Austrians and Post-Keynesians agree–but if you just can’t bear the suspense, here are three links:
==> I walk through the “reswitching” controversy, focusing on Paul Samuelson’s excellent numerical example. Note that he sold this in a very interesting way: It is actually a concession that his side lost in the Cambridge Capital Controversy (that’s what Galbraith was alluding to in his critique of Piketty), but he shakes it off and portrays it as the death of Bohm-Bawerkian analysis. (Strictly speaking, Bohm-Bawerk said nothing incorrect during the CCC, because he had been long dead when it occurred.)
==> I explain (and critique) Sraffa’s approach to explaining the income flowing to capitalists. (Sraffa rejected the marginal productivity approach to income determination when it came to the capitalists.)
==> If you’re super geeky, and want to see how even-handed I am, in this academic paper I devote one section to showing that I think the Austrians did not fully answer Sraffa’s critique of Hayek.
This could be very interesting, so fire away and good luck with it. But if you want to relate it to Piketty you need to show that he actually needs a neoclassical aggregate production function for his argument. AFAICT he really doesn’t, but admittedly I’ve only got 60 pages into the book. (It’s bloody good, BTW; the man can write.)
It won’t surprise me if Piketty uses neoclassical arguments now and then, but that’s no big deal if he can get the same conclusions from a disaggregated model. Since he’s a good mathematician, by economists’ standards, it’s quite likely he can.
In short: do this thing anyway, but you’ll greatly raise its salience if you can show that Piketty is talking neoclassical without knowing it. I’d say that would be news to him, and also to Brad DeLong, who suggested on Twitter that a simple search-and-replace exercise could eliminate all CCC-type difficulties from the book.
Kevin,
OK I will try to keep an open mind, but it really alarms me that he apparently missed the entire point of the CCC. (Since that was my reaction, I was glad to see JG saying the same thing.)
But what exactly did DeLong suggest he do? Take out every reference to “an argument relying on a simple capital/labor ratio” and replace it with a different argument?
I’m only half kidding when I say it’s a bit hard to argue with people who say, “Yeah sure, we know this is conceptually meaningless, but if we set that objection aside, look at this result…”
On Twitter:
Daniel Davies (dsquared):[when Piketty writes] about “marginal product of capital” is like watching someone repeatedly try to divide by zero..
Brad DeLong: global search for “product of capital” & replace with “income paid to wealth”. Piketty argument goes through fine…
I really think that’s about the size of it. But as I say I’m only up to page 60. I certainly won’t just assume there’s no problem. If Samuelson could come a cropper then so can Piketty.
thanks
Samuelson’s example is interesting but it bothers me that most real business operates as a continuous process, not as a one off. Thus technique A as a continuous processs uses 7 units per period and generates one output per period. While technique B uses 8 units per period and generates one output per period.
I understand that the question is about business startup finance but as an ongoing business A should be more profitable, so it comes down to a question of whether the price of startup finance can justify adopting what is (long term) going to be a more expensive process to run. Seems counter intuitive that this might be the case.
Replying to myself here, I did some spreadsheet models and the rewsitching theory works if the business is getting sufficiently high interest on their bank savings to give the advantage to whichever technique reaches profitability (in the sense of paying back their startup loans) first. In other words it requires that ongoing compound interest on savings is large enough to swamp out the day to day business profit margin.
That seems a bit of a pathological case, for the most part, interest on bank savings is significantly lower than interest on loans. Also, I just find it highly implausible that a real life situation will happen where interest rates are so high that it’s worth adopting what is an operationally less efficient process for the business, just to reduce the necessary startup finance.
I can post the spreadsheet when I get a moment.
Bob, even if you disagree with him, why do you think Krugmsn believes that using marginal productivity theory doesn’t depend on aggregation of capital? Is there some model he’s thinking of with a heterogeneous capital stock but where marginal productivity theory still applies?
consultingbyrpm blog tag piketty