Say you are an American businessman. You are spending your evenings this week recalibrating your winter prices to reflect the possibility of a deep market slump. Or say you are a member of the Federal Reserve Board. Prospects of an oil price rise lead to your altering the price of money - short-term interest rates. The future of the economy as a whole is unknowable but everyone constantly resets prices to adjust to that future as best they can.

Everyone, that is, except the US Congress. Congress sets its prices - tax rates - in a vacuum. It quantifies predicted revenue via a formula based on the past. Last year a certain tax of 2 per cent brought in $2bn. Doubling that rate will bring in $4bn. Halving it will bring in $1bn. Simple arithmetic.

That is the rule, although the federal government is such a Big Foot operator in the economy that every fiscal step it makes alters the growth landscape and therefore revenues. The forecasting method - known as static analysis - is dangerous for two reasons. The first is that it leads to surpluses today and deficits tomorrow, fuelling that famous challenge to western democracies: political disillusionment. The second is that it limits the tools Congress can use to help the economy.

In static analysis, tax increases are always good and tax cuts are always either "expensive" or "too expensive". The proposed tax cut for dividends that recently generated so much excitement is a typical casualty: the White House is quietly giving up the idea as too costly. Now various Washington players are waging a nasty little battle over static analysis. The Republican House, led by Bill Thomas, chairman of the ways and means committee, has set up a commission to work out how to add dynamic analysis - tax analysis that reflects the macroeconomy - to the government's forecasting toolkit. The Democratic Senate, led by Max Baucus of the finance committee, wants to pack the commission's membership with "static only" supporters.

Democrats oppose dynamic analysis because they think doing so makes them look like good fiscal conservatives, which they think will win them votes. Republicans back dynamic analysis to sell their traditional big issue - tax cuts. But this partisan struggle is a shame, since dynamic models could help both sides.

Consider what happened five years ago, when the staff of the joint committee on taxation sought to quantify the effect of President Bill Clinton's new tax law. The law cut the capital gains rate to 20 per cent from 28 per cent. They might have predicted that the rate cut would generate extra revenue. That was what happened after a capital gains cut in the early 1980s.

The opposite had also occurred: in 1986 a large rate increase had produced disappointing revenues. The reason was straightforward: lowering rates encourages citizens to invest more. By contrast, rate rises encourage them to postpone transactions. But the joint committee used static analysis and predicted that the Clinton cuts would cost $74bn of revenue over four years.

In practice, federal revenues were higher than predicted. Indeed, a Heritage Foundation study found that total federal revenues in the late 1990s were $1,000bn higher than Uncle Sam had forecast. Several hundred billions of those mystery dollars were revenues from the capital gains tax. The increase in revenues was due not only to increasing investment but also to growth triggered in part by the tax cut.

With hindsight, the Democrats could claim credit for their positive step; Al Gore benefited from the party's reputation as a solid economic manager in the presidential campaign of 2000. But the Democrats' reputation would have been stronger had they been able to show that they recognised the potential of their actions from the beginning.

Or consider more recent legislation, which would have reduced inheritance and gift taxes to zero over a decade. The government believed the cut in the estate and gift tax would have cost $306bn over that period. It did not take into account that ending the two taxes would generate growth. This was despite the fact that estate taxes are classically inefficient, diverting billions that could be invested in industry to trust lawyers. The outcome was bad for both parties. Because the paper cost of the tax cut was too high for budget rules, Congress made the cut temporary. Estate and gift taxes will return in 2011. Now voters are angry at politicians over a tax-cut-that-isn't.

The sceptical reader may still wonder whether dynamic analysis is just an excuse for Republican tax-cutting (which the reader may deem irresponsible). But the suspicion is disproved by the fact that states led by both Democrats and Republicans have, for years, successfully used macroeconomic models to determine the effect of changes in state tax policy.

The ability to consider various models matters most in the new discussion of the federal deficit. One can argue, from either a Keynesian or a supply-side perspective, that tax cuts can help the US economy grow fast enough to render deficits unimportant. Yet the static rule means that argument is hard to make on the legislative level. This is sad, since the issue here is not Thomas versus Baucus, or even Republican versus Democrat. It is more growth versus less.

© Copyright 2002 Financial Times