Tuesday, November 11, 2014

My Comment To Peter Orszag's Bloomberg Article About CBO's Non-Partisan Analysis

From BloombergView, "A Party Hack Would Ruin the CBO" by Peter R. Orszag:
The Congressional Budget Office should be able to celebrate its 40th anniversary this coming February with pride. The agency plays a crucial role in the nation's policy-making, providing rigorous and nonpartisan analysis of the budgetary and economic issues Congress considers, and objectively estimating the cost of legislative proposals.

The occasion will be ruined, however, if the new Republican Congress breaks its long tradition of naming an objective economist/policy analyst as CBO director, when the position becomes vacant next year, and instead appoints a party hack.
My posted comment to BloombergView, "A Party Hack Would Ruin the CBO" by Peter R. Orszag:
CBO's existing methodology can be greatly improved. The current methodology is a political compromise of arbitrary analytical policies that are subject to gamesmanship and do not follow accepted financial analysis protocol. CBO's analytical policies are biased towards bigger government and tax increases over smaller government and tax reductions.

CBO uses a 10-year window for analysis and does not time-value (present value) revenue and expenses. Any costs or revenues beyond the 10th year are not considered. For example, If CBO analyzed a government annuity program that collected premiums for 10 years and did not start payouts until the 11th year, CBO would count the premiums as revenue and say the program has no payout expenses. Likewise, if all the payouts occurred in year one and an equal amount of new taxes were collected in year 10, CBO would say the program is revenue neutral, where revenue equals costs. The correct analytical way to deal with mismatched timing of revenue/ expense, and to have an analytical period that coincides with the programs expected duration, is to use a present value framework where a dollar today is worth more that a dollar tomorrow. Present value allows CBO to analyze the financial impact of a program over its expected duration with mismatched revenue and expenses beyond 10 years.

CBO uses a static economic model instead of a dynamic model. CBO assumes government programs and policies will not change projected GDP, employment, hours worked, business formations, household formations, capital investment, etc. This policy favors bigger government and tax increases since CBO does not model negative or positive effects on GDP growth or employment from tax increases/decreases or new government legislation.

CBO assumes all enacted legislation will remain unchanged. Congress regularly enacts cost savings measures that have a future start date. CBO will assume those cost savings will occur and reduce government expenses, even if Congress has a history of delaying that start date of program reductions and cost savings. Similarly, some programs have expiration provisions, but Congress regularly extends the life of program. CBO assumes the expenses will end at the legislated expiration date.

CBO does not use expected values or a probabilistic approach to outcomes. Nor does CBO discuss a range of values for modeling outcomes of a government program. Suppose a new government program has a 70 percent chance of costing $40 billion and a 30 percent chance of costing $70 billion. CBO will use the $40 billion figure to analyze the program. A proper analysis would use the expected value (average) of the cost, which is $49 billion (.7x$40b + .3x$70b).

No comments:

Post a Comment