The Stimulus
I've decided to attempt this post in two parts. First, general economic theory, and second, specific provisions of the current Stimulus plan. Hopefully with some kind of conclusion. I would also like to say that while I have read the majority of the posts on Jon's post I don't want to try to incorporate them here for the sake of my sanity, so if I end up saying things that have already been said, my apologies.
The economic rationale behind the stimulus is, as has already been said, largely Keynesian in nature. Fundamentally this can be boiled down to the idea that in rough economic patches the loss of consumer and corporate demand should be made up for by an increase in government demand as the only actor with pockets sufficiently deep to go largely unaffected through the economic woes. Taken in a larger context the Keynesian idea is basically that government spending can smooth out economic cycles using "countercyclical" fiscal policy. It was this idea that shaped much of the Depression-era and postwar economic policy in both the US and Europe. And indeed, Keynesian economics occupies (or should occupy) the same fundamental pedestal that monetarist thought does: that is it works under certain conditions and to a certain extent, but will not work or even fail under other circumstances. Economics is very much like any other science in this way, every theory has its regime of applicability. The difference is that applications of economic theory directly affect people on a gut level (in their pocketbooks). So while noone debates that a Newtonian description of particle motion works in some areas while a Stat mech description is better in others people are less sanguine about economic theories (or more sanguine, depending on which particular definition of that word you choose, as pointed out by Zoe on Firefly).
I will say this: right now is one of those times when Keynesian theory is applicable. However, there is a continuum of how applicable Keynesian theory is and the current numerical expression of that continuum is what people think the size of the stimulus bill should be. In answering this question let us look at the pros and cons of the application of Keynesian theory. First the pros. With private-side demand supressed the government is the primary actor that can push through large amounts of money in a coordianted fashion. The private side cannot do this becuase to do so would require each of the millions of actors in the system to almost simulatneously come to the same conclusion, and it becomes the classic investment paradox: everyone wants to be the second investor, noone wants to be the first, because the first investor bears all the real risk. With a Keynesian approach the government becomes the first investor in a sense. Now the cons. Here lie the classic problems with government spending: timing and efficient capital allocation. Timing because for the effect of a countercyclical fiscal policy to work the infusion of money into the market has to be timed correctly. For particularly long downturns this matters less because the assuption is that the economy will still be operating below nominal level whenever the money hits (i.e. the Great Depression). But for shorter downturns the inflow of money toward the end of the downturn (or even after the downturn entirely) will only create inflationary effects and in the longer run stagflation (i.e. US econmic policy from 1960 through 1980, starting with some bad advice Kennedy got about unemployment, going through Nixon's retarded price controls, and culminating in Carter's botched attempts to deal with the oil shocks, see Robert Samuelson's book The Great Inflation). As for efficient capital allocation, well there's the trick, isn't it? It should be inherently obvious to even the most casual observer that large-scale capital allocation by the government invariably leads to gross inefficiencies in the economy, as the Soviet Union, China pre-1990, and to a much lesser extent France in the last 50 years have shown. This is largely a result of the nature of politics and politicians not being very good at spotting real opportunities while over stressing marginal or nonexistant opportunities for the sake of political gain.
So if we look at the whole of Keynesian theory and try to put a number on the stimulus what do we get? The unfortunate answer is: not much. I tried (briefly) to do a search for what economist's thought the number should be and got a gigantic range from $0 to Paul Krugman's $1T is not enough. While there does seem to be a broad consensus for some kind of stimulus, the form of that stimulus is debated and recently a full page ad in the NYT was taken out by the Cato Institute and signed by 200 economists, including 3 Noble Prize winners, directly criticising Obama's plan. In the absence of doing full calculations myself and in the absence of specific guidance from actual economists I am going to say that roughly $200B in stimulus is warranted (note that this number includes all forms of stimulus from tax relief to direct spending). The reasoning for this number will be explained below.
Turning to the exact form of the stimulus bill we have the CBO report on the version that passed the House (note that this is the HTML of the Google archive of the summary of the CBO report because the CBO website is getting hammered and I can't get through). First the numbers: $815B total, $170B in FY 2009, $356B in FY 2010. Note that the report breaks down the stimulus into three groups: appropriations (i.e. new government spending on projects), direct spending (i.e. unemployment benefits or tax credits for specific events), and tax reductions. In terms of percentages this represents 20.9% spent in FY 2009, 43.7% in FY 2010, and 35.4% after that. This is the roughly 65% spent in the first 2 years that I quoted in my earlier discussion on Jon's post. The 10% number I quoted (and here the fault is confusion on my part) is the percentage of the total in the appropriations category only that would fall in FY 2009 (and that number is actually closer to 8%).
This is where the cons of the Keynesian approach come into play. Note that the House bill has some expenditures extend to 2019. Fully 10 years from now and 2 years longer than even a successful Pres. Obama could be in office. This is at the absurd end of the spectrum. Why, in a short-term economic stimulus bill, are there appropriations for a decade from now? Looking at a history of US recessions we can see that no recession in the last 60 years has lasted more than 2 years, even including the severe 80-81 recession that Regan and Volckler used to finally tame inflation. Other than the Great Depression, no recession in the last 100 years has lasted longer than 3 years. The current recession started in Dec. 2007. Keping in mind that even once the money is spent by the government it takes, on average, at least 6 months for the full effect to be felt a strong case can be made for a stimulus that kicks in in FY 2009 (the federal FY ends Oct. 1).
Here is the fear: excessive government spending on the tail end of the recession or even once it has passed will push the money supply up further than it should be causing inflation and eroding the value of the dollar. Meanwhile the federal debt gets pushed up ever more getting us nearer and nearer to that inevitible line that the world wil not cross where they will no longer take on US government backed debt. Call it "peak debt" if you like (kudos to Daniel for the term). Now one's response might be along the lines of "Bush spent a ton of money on various crap, so why shouldn't we" but that is in effect saying that 2 wrongs make a right, and I reject that false argument.
There is also the question of the merit of many of the programs in the current package. I am in no position to fully comment on this other than to say that many of them do, in fact, have merit. But do they have enough merit to accept them with no focused debate for spending two years from now? I have to judge that in the negative. Having reviewed the summary of many of the programs in the CBO report above I have to say that I like the idea of many of the projects listed. But I would like to see the ones with no immediate effect properly debated, and preferably coupled with a project that is not working and could be cut to make room for the one under consideration (to make this as fair as possbile across departments one could institute the requirement that the cut be made in the same department as the project being proposed, that way in is interdepartmentally budget-neutral).
So here is what I propose (and where my $200B stimulus number above comes from): 1) The current bill has $170B spent in FY 2009, so I take this to be the politically acceptable pseudo-consensus on actual stimulus. Round that up ward a bit to a nice $200B. 2) Take all of the parts of the stimulus package linked to above and cut all projects where at least 60% of the money is not spent in FY 2009. 3) This should leave you just short of $200B in total outlays (for 2009 and beyond). 4)Start adding back in projects with the highest percentage of funds allocated in FY 2009 until $200B is reached. This last bit would, by political necessity, have to be fudgable a bit, but all of the remaining projects should hit at least 40% for FY 2009. You want to be bipartisan? Then actively court Republicans with the last little bit, or at least include them in the discussion instead of dictating priorities. 5) The remaining projects should be primarily FY 2010 and beyond and therefore should be brought to the floor and debated as any other spending bill would be given that FY 2010 doesn't start for 8 months and the budget isn't due for at least 6.
If Obama wants to meet his deadline this would be the fastest way to do it. I understand his desire to reshape US spending priorities, but trying to sly one in under the radar as a stimulus is just wrong on multiple levels. At the very least it sets a bad precedent for future economic downturns when someone with different priorities may be in power. At its worst it could herald in the economic climate of the late 60s and 70s, which wouldn't do anyone any good.