The latest version of the Biden administration’s Build Back Better reconciliation package reintroduces a policy that has been tried before—and abandoned: a corporate alternative minimum tax (AMT). It would be a mistake to revive this complex and poorly designed policy. Instead, lawmakers should consider directly reducing corporate tax expenditures.

The corporate alternative minimum tax was first introduced in the Tax Reform Act of 1986 (TRA86) in response to arguments that are echoed by some today: publicity around a phenomenon in which very large firms with significant accounting earnings were paying little or no tax. The AMT was thus designed to prevent specific companies from using enough tax breaks to face zero tax liability in a given year while reporting an accounting profit. It was repealed in 2017, and the 30-year experience with a corporate AMT shows it is not a good solution: if tax breaks are poor policy, they should be repealed directly; if they are sound policy, all eligible taxpayers should be able to take full advantage of them.

Instead of taking that direct approach, the AMT required corporations to calculate their income tax liability twice, once using the tax code and once using an alternative calculation for tentative AMT, and pay whichever was largest. The AMT created economic inefficiencies, increased tax burdens and complexity, and saw declining tax revenues over time.

Companies that paid AMT in one year received a credit that could be used to offset future liability under the regular tax code (but not below their tentative AMT liability), which meant that for most firms, the primary effect of the alternative minimum tax was to shift tax liability over time rather than substantially increase it.

As illustrated in the accompanying chart, corporate AMT liabilities peaked in 1990 at just over $8 billion, falling to a low of $1.8 billion in 2001. As firms built up and used tax credits for prior year AMT to offset their regular tax liability, the net revenue raised from the AMT system fell: From 1995 through 2001, credits for prior year AMT exceeded current year AMT liability, meaning the AMT system lost revenue on net in those years.

Because firms moved in and out of the old alternative tax system, they could take advantage of credits for prior year AMT to offset portions of their regular tax liability, significantly reducing the net revenue raised. A similar, but perhaps smaller, effect could emerge under the new proposal, as it also includes a credit for prior year AMT. A firm’s ability to use its credits may be limited, however, if the new proposal is more likely to keep firms in the AMT system. If firms remain paying AMT liability year after year, they would be unable to use their credits, limiting the offsetting effects of the credits.

Another relevant problem with the AMT, particularly the original TRA86 version that incorporated book income, is how it treated depreciation deductions. For tax purposes, it makes sense for firms to deduct the cost of physical investments when the investments are made. For accounting purposes, however, spreading deductions over several years often makes more sense. The AMT partially disallowed accelerated depreciation deductions that let companies take deductions for investments faster under the regular tax code.

As a result, it disproportionately affected firms in industries that rely heavily on physical capital investment, like manufacturing. As a 2005 Treasury Department report noted, mining, warehousing and transportation, and manufacturing firms faced the largest AMT burdens. Such disparate treatment of firms in different industries, and of different sizes, created economic inefficiencies.

Throughout the 1990s and early 2000s, the depreciation rules under the AMT frequently changed but eventually allowed temporary bonus depreciation for both regular tax and AMT—significantly reducing the effect of the corporate AMT by reducing the depreciation differences between the systems on a temporary basis. The same Treasury report noted that “a more efficient tax system would treat all firms equally, leaving investment and other business decisions to be undertaken based on their economic fundamentals rather than based on their tax consequences.”

Ultimately, the corporate alternative minimum tax has either ended up as a heavy burden on investment or as an ineffective revenue raiser. If politicians want to deal with the problem of targeted, non-neutral tax breaks as a way to raise revenue, they should repeal those policies directly, rather than use a complex policy like the alternative minimum tax that also punishes many legitimate deductions in the process.