My main reaction to the CBO's new review of the House stimulus bill (see the director's blog; a fuller version) is that the package is much smaller than previously supposed. The CBO says it would increase the budget deficit by $816 billion during 2009-2019, but by only $525 billion when it is most needed, in the remainder of fiscal 2009 and 2010. The stimulus for the rest of 2009 is put at just $169 billion. The 2009-2010 stimulus needs to be much bigger, in my view. And the ten-year stimulus should be much smaller: if you are going to look that far ahead, you should see lower spending and higher revenues relative to baseline.
A lot of the front-loading, such as it is, comes from the tax-cut provisions that many Democrats object to. Almost the whole of the ten-year tax-cut stimulus, a little over $200 billion, arrives in 2009 and 2010. If the plan relied on spending alone, barely half of the ten-year stimulus would take effect in 2009-2010.
The report underlines the distinction between infrastructure spending broadly defined and direct payments to individuals, such as unemployment assistance and food stamps. The first kind of spending arrives far more slowly, and the report emphasises the additional delays that are likely when programs are being scaled up so dramatically.
[U]nder H.R. 1, some programs would receive funding that is significantly above (double, triple, or more) the amounts provided for existing or similar programs in recent years. Frequently in the past, in all types of federal programs, a noticeable lag has occurred between sharp increases in budget authority and the resulting increases in outlays. Based on such experiences, CBO expects that federal agencies, along with states and other recipients of that funding, would find it difficult to properly manage and oversee a rapid expansion of existing programs so as to expend the added funds as quickly as they expend the resources provided for their ongoing programs.
Lags in spending stem in part from the need to draft plans, solicit bids, enter into contracts, and conduct regulatory or environmental reviews. Spending can be further delayed because some activities are by their nature seasonal. For example, major school repairs are generally scheduled during the summer to avoid disrupting classes, and construction and highway work are difficult to carry out during the winter months in many parts of the country.
Brand new programs pose additional challenges. Developing procedures and criteria, issuing the necessary regulations, and reviewing plans and proposals would make distributing money quickly even more difficult--as can be seen, for example, in the lack of any disbursements to date under the loan programs established for automakers last summer to invest in producing energy-efficient vehicles. Throughout the federal government, spending for new programs has frequently been slower than expected and rarely been faster.
The much-touted investment in broadband may be good value for money, but its short-term stimulus effects are judged to be poor:
CBO anticipates that funds provided to the National Telecommunications and Information Administration (NTIA) to administer the broadband grant would take longer to spend--eight years--because the new appropriations would far exceed the agency's 2009 funding of $17 million and the legislation would require grant recipients to deploy such services with technological features that are not widely available today.
The main thing is this: the package needs to be made bigger (something Republicans don't want to hear), and in expanding it the focus should be on stronger front-loading (Democrats don't like where that leads). The clear implication of this report is that still more infrastructure spending is not the way to make this plan better. I would take another look at food stamps, unemployment assistance and fast-disbursing budget support for the states; and then (not forgetting Allan Meltzer's proposal) at additional tax cuts.