Well, I'll repost a comment I made over at Mark Thoma's site in response to a posting of a New Keynesian account of "crowding out" from Blinder's and Baumol's textbook a while back:

"There seem to be some dubious generalized assumptions in Blinder's textbook account:

1) Public spending does not count as, in some part, investment. To the contrary, basic research, r&d, education spending, infrastructure, etc. constitute additions to the capital stock, broadly construed, of the productive economy, facilitating, supporting, and even directly subsidizing private capital investments, and encouraging high-cost long-run fixed capital investments, which private interests might be otherwise averse to risking, hence amounting to a form of "crowding-in", in a sense other than stimulating short-run demand. (And I think this factor is larger than the standard, narrowly-tailored public goods argument).

2) The (real) rate of interest is a key determinant in incentivizing private capital investment. But not only can investment be financed through retained earnings, which used to be largely the case, but both the horizon of available unrealized technical possibilities and the level and composition of real aggregate demand are key factors, with the distribution of income between profits and wages being a key determinate of the latter, (and a motive for the former). Low real interest rates would just as likely encourage higher consumption (and lower savings), and thus, yes, higher short-run demand, as increased long-run investment. Low real interest rates also tend to encourage leveraging of short-run speculative investment in financial "assets", rather than risky, high-cost fixed capital investment, which last and must be managed and (re-)financed beyond any given interest rate/business cycle, such that low real rates might not be the prime criteria vs. strategic plans and perceptions of increasing returns. Low real rates might also distort capital structures, by encouraging the assumption of unsustainable levels of debt, (as with recent PE buyouts and stock buy-back schemes). So the "crowding out" of some forms of private consumption and investment spending with deficits yielding higher real interest rates might not be entirely unhealthy long-run, (provided the rate of growth in public debt doesn't exceed the rate of long-run growth, which can service debt out of increased fiscal revenues).

3)Capital stocks are supposed to be a quantitative amounts, both financially and physically, and the amount of capital stocks bequeathed to the future will determine the level of future economic growth. Yet the primary sources of real economic growth, increased per capita income, increases in the real distributable surplus product, are technical improvements in the productivity and quality of capital stocks, innovations in processes and products. Not only can such technical possibilities not be guaranteed, nor predicted in advance, but they tend to depreciate and destroy older and competing capital stocks, just as well decreasing as increasing the quantity and value of capital stocks. How well an economy processes such destructions and dislocations would seem to be just as important a factor in long-run real growth, as whether it plans for sufficient savings/investments. Further, successful outcomes to innovations derive from a large component of contingency, and a probable range of failures, while leading on to concentrations of wealth, income and market power. A fair degree of income redistribution, forms of social insurance and public spending of health, education and welfare, to reshuffle the deck of opportunity-structures, would likely increase the likelihood of such favorable contingencies in innovation and their positive externalities vs. allowing them to rigidify beyond a certain point into dominant, self-perpetuating socio-economic structures. So again, in this respect it's not a priori clear whether moderate deficit spending crowds-in or crowds-out private investment and consumption spending, as the future determinant of future aggregate wealth. Put bluntly, a dollar spent today is not necessarily to the detriment of a dollar spent ten or twenty years hence, since both the composition and the value of a set of prices today is not the same as those of a future set of prices, even if adjusted "really" by a chained index.

So it's not clear to me that focusing abstractly on a few aggregated quantitative parameters is sufficient to adequately address the functional efficiency of a dynamically developing economy, let alone the equity and beneficence of social life."

The upshot is that, in my view, Wray is likely right. Public fiscal deficit spending does not merely compensate for shortfalls in private consumption demand, (due to the invalidity of Say's Law), but compensates for shortfalls in private investment, due to aversion to high, long-run fixed capital costs and its associated high uncertainties. In fact, historically, such investment has been often at least implicitly subsidized by the government, as with, e.g., land grants to railways, tax breaks, or government orders bringing markets to scale. Government spending on public goods amounts to a "loss leader", incentivizing further private investment by providing networks of profit opportunities. It's really just a matter of whether it's an explicit policy in the deliberated public interest or a result of the capture of government policy and decision-making by private business interests.

Now can we move on to debating issues in the real productive economy, rather than the text-book economy, such as appropriate forms of industrial policy to re-structure and re-balance the economy away from persistent trade deficits and financial rent-seeking?

Ellen:

No, "Say's Law" states that supply, (i.e. production), creates its own demand, (because wages paid throughout the production supply chain will constitute final demand). But "Say's Law" is simply not per se true, because, over the long-run, increases in productivity will out-strip wage-based demand. You're caught up in the a priori of static equilibrium thinking that reduces the whole economy to a supposedly complete set of competitive markets. That's a logical idealization with little correspondence to economic reality.

Come on you economists out there. Jump into the Wray, er, fray. It's fun for a non-economist like me to read your exchanges. Learning is entertainment. Writing is a adult play. Fun to watch you romp. More interesting than watching morons bounce a ball or run into one another. You have an admirable noggin, Ellen, and I'm not referring to the eyes of your avatar. Some of your moves are not so hot; others are spot on.

Ellen: you've got it: taxes drive money. Govt spends it into existence so we can pay our taxes. All of macro follows on from that recognition.

I won't enter the fray into what Say really meant.

Mr Sandwichman: I support work sharing, fewer work hours, flex time, and whatever else you are selling. I do not buy the argument that this will significantly impact unemp but even if it reduced unemp by 50% or 90% that is not enough. Let us create jobs for the rest. So in short I am not going to argue against your proposal because i support it; but i am not going to argue for it because my heart is not in it. Surely there are at least 100 good ideas and policies out there; i limited my proposal to what I am willing to argue for. That doesn't mean i oppose others.

thanks to all for comments.

This position doesn't seem to make sense over time. On reality-based planets, government surpluses mean reduced government debt, which means both lower taxes into the indefinite future (all other things being equal) and lower transfer payments from median tax payers to holders of government debt. Both of these factors will have generally stimulative effects.

You can argue that the surpluses put households in a poorer position by reducing the apparent savings required to pay future taxes (which rosy scenario was demolished by the Bush bait-and-switch that made either higher taxes or asset devaluation an inevitable part of our economic future), but that's like faulting a bus driver for the fact that the driver on the next shift ran the bus into a brick wall.