Andrew Samwick writes about advising politicians about budgets. Listen to him. He's smart:
Vox Baby: How To Advise on Fiscal Policy: The overriding problem in conducting fiscal policy is that politicians, in both the executive and legislative branches, face electoral pressure to please their current constituencies. It is extremely tempting for them to boost spending or lower taxes today, handing out windfalls to today's voters and leaving an unrepresented constituency--future taxpayers--to foot the bill. It takes an enormous amount of energy to resist that temptation. That energy has to come from the politician's advisers.
In the White House, the most senior of them carry the titles Assistant to the President, in most cases preceded by the word "Deputy" or "Special" to connote their place in the hierarchy. During the year I spent at CEA, I worked with a number of them who were excellent. But if you ask them what their objectives are, or if you listen to their arguments during policy meetings, what you find is that they are working extremely hard to advance the President's policies (which they have often had a central role in shaping and hopefully improving). They may have very good intentions at heart, but in my experience they were not usually the ones looking at the larger picture who might help resist the temptation to act in a politically opportunistic manner. There were exceptions, and I'm not looking to impugn anyone here.
There are other advisers with a central role to play in fiscal policy, however, who are Presidential nominees who undergo Senate confirmation and who can be called to testify to Congress on the President's policies. We should expect them to be the ones who look at the larger picture, who take a longer horizon into their policy discussions, and who do the heavy lifting to help the politician resist the temptation to use future generations' resources to buy votes today. On fiscal policy, the three main advisers are the Secretary of the Treasury, the Director of the Office of Management and Budget, and the Chairman of the Council of Economic Advisers.
To address the overriding problem, these three advisers must insist on an explicit budget target. The weakest one that I would accept is that the debt-to-GDP ratio (including debt held by government trust funds) must show no upward trend. (A stronger one would be that the on-budget deficit should be in balance over a full business cycle.) Long-term entitlement programs should be in projected actuarial balance to the extent it is possible to make a projection. (You can see here what happened to one adviser who made reasonable arguments without first getting agreement on a standard.)
The standard can be that simple, and it isn't as austere as I'm trying to make it sound. It doesn't rule out cutting taxes during the weak part of a business cycle, for example, but it does rule out cutting taxes to run a deficit that the Administration has no intention of paying back during the strong part of a business cycle. As a corollary, it does rule out passing tax cuts with explicit sunset provisions and then arguing to extend them with the budget not in balance. There can be occasional exceptions, but their presumed infrequency should immediately cause them to be fully explained. (Think of the President addressing a joint session of Congress.)
Twice a year, the Administration makes an economic forecast to underlie the budget or its mid-session review. I'd be surprised if you ever see such a forecast that predicts an upcoming recession. That means that every budget or mid-session review should be projecting on-budget surpluses or their quick resumption if we are just coming out of a recession. As I've discussed elsewhere, the "cut the deficit in half in five years" policy was at variance with this standard.
I acknowledge that my views on this don't match the usual discussions of the impact of deficits. I think too much of those discussions are about their impact on interest rates and thus the incentive for private investment. This is where economists spend a lot of their time, but that doesn't make it the most relevant question. Suppose for the sake of argument that deficits don't put much upward pressure on interest rates. Even in that case, they still have to be financed at the existing interest rate, and the burden of financing them has to be borne by someone in the future. Taxing someone in 2020 to pay for our spending binge in 2003 violates my notions of fairness, and that is a substantially more salient issue here than any additional concerns about efficiency.