Thread by @yungneocon

Really wild to me that people think monetary savings leads to growth--even with the idea that all savings are invested, that nominal monetary amounts, held by financial institutions, should unproblematically translate into investment-induced growth is totally absurd.

The line Keynesians use is that, in a barter system, all savings *would* lead to growth but a monetary one is different. But this also plainly isn't true, because accumulation of bartered goods still can't be evaluated without knowing social distribution & structure.

So, for example, if capital goods & inventory depreciate or deteriorate over time (such as food stocks), their accumulation would be self-defeating. If time, social & space constraints cause satiation effects in consumption, then accumulation is just waste.

On top of this, accumulation requires costs for storing, transporting, monitoring, indexing, displaying, etc., which cause dis-economies of scale & scope. Nonetheless all of the above are substantive (tho intrinsic) issues, but they could be assumed away for a model.

The issue that Sraffa pointed out was that unless capital goods are homogenous, which they aren't--specific things are needed to work, and together--to transform one kind of capital good to another requires:
1. Mixing it with labor to produce the new one
2. Exchanging w. another

But, in both of these cases, there is now the intercession of time, and of an aggregate social system. The worker's labor requires goods of its own, both as inputs and so the worker can survive. Exchange will occur at rates that maintain distributional social structure.

As such, commodities then begin to take on a dynamic character--they only make sense in a system that is reproducing or expanding. Land doesn't contribute to production, and is homogenous, so its rent can be factored out. Labor does contribute but is homogenous as an input.

The point that Sraffa makes isn't that, therefore labor is not the source of value, but instead that because labor & land are universally necessary & homogenous, they can be factored out--information, after all, is the difference that makes a difference.

i.e. between different economic systems & between different stages of the same economic system, that labor & land will be used & homogenous is guaranteed, so, one must instead look at what varies if one wants to model the system on a meta-level.

What varies, therefore, between and within systems, are:
1. The composition of commodities
2. Their technical/material relations
3. The composition of the class system
4. The quantitative distribution thereof

Now, I would add that systems also vary by their divisions of labor in ways that are relevant other than class, that institutional systems vary, that metrics vary, and that monetary type varies, and I'd also argue that money cannot be assumed away, but that's for another day.

What Sraffa showed was that in any system of complex commodity production & non-homogenous capital--i.e. a reproducing economy--with two or more claims on output, whether they be labor & capital, consumers & investors, the state & workers, these results hold.

But even in a barter economy, Sraffa points out that commodities have an 'own rate of interest'--the amount at which you can buy the same commodity later--and that, out of equilibrium, no single 'natural rate' will prevail, for all of the own rates will be out of sync.

Indeed, out of equilibrium, there are as many 'natural rates' as there are commodities, for the natural rate of interest will be different depending on the own rate of the commodity you choose as a numeraire!

Now, these results didn't exist at Sraffa's time, but there are many results that show that:
1. Conditions under equilibrium occur are those under which autarky or planning work as well
2. Incomplete markets & extra-market norms/institutions are necessary for exchange

3. Absent a mix of incomplete information, noise, transaction costs, search costs, disequilibrium, etc., there's no reason for trade to occur in the firt place
4. The information & coordination demands of the double coincidence of wants are incalculable

The reason this matters is that it means that *if* trade is occurring between commodities (or, for that matter, if a central planner is allocating them--both the Sraffian & market design results still hold under command!), it means there is disequilibrium, time, etc.

As such, that commodities are being exchanged, or being materially balanced by a planner implies disequilibrium, which implies variation in own rates, which implies there are as many 'natural rates' as there are commodities, and the choice of numeraire.

In other words, even in a barter or centrally planned economy there is no escape from:
1. Capital heterogeneity
2. Land & labor (semi-)homogeneity
3. Rents as a share in produce
4. The effects of time
5. The dependence on metrics/accounts
6. Class structure

7. Disequilibrium of own rates of interest for commodities
8. The technical properties of commodities, capital, labor & production
9. The distribution of commodities between classes and sectors

i.e. in such a system--whether barter or planned--there can be no growth due to accumulation of commodities without there being a class structure, a set of metrical accounts, a pre-given distribution of resources, & methods of converting heterogeneous commodities into one another

As such, we get back to the 'Capital as Power' issue--no accumulation without capital, capital is a claim on commodities & resources, usually, not exclusively, indexed thru the unit of account of money, a kind of socially instituted power.

Even gains in terms of material balance, either more output, or more efficiently produced output, are not guaranteed by accumulation, unless there is conversion into fixed capital or techniques, reproduction of labor relations, and institutions & accounts to preserve them.

Interestingly, that 'institutions' & power (as well as technical conditions, the need for reproducing labor & the necessity of using land, space, time & information) go all the way down, so to speak, provides us a fulcrum with which to move the economic world.

In different ways this is the point being made by N&B's CasP, by Ajit Sinha's gloss on Sraffa, by Phil Mirowski's wonderful work, by the work of G Ingham, V Zelizer, D. Mackenzie, etc., by the Godley/Lavoie/MMT people, by econ anthros ('Debt'), and many more.

And although not making the exact same point, these facts all fit nicely with the Geoist tradition (H George, S Gessell, M Gaffney, B Hodgkinson), with Balance of Payments constrained growth, post-keynesian & neo-structuralist views (Thirlwall, Felipe, McCombie, E. Hein)

The irony I find sort of delightful is how one doesn't even need to go to heterodox schools to show this stuff (Sraffa, Neo-Ricardo, Post-Keynensian, Marxian, Geoist, Institutionalist, etc) as much of this stuff is provable within neo-classical economics itself.