With President Trump’s stunning return to power, Congress has the opportunity in 2025 to enact additional business tax cuts. One of their specific goals should be to make “full expensing” of business investment a permanent part of the tax code. By allowing businesses to deduct the full cost of investment in the year the spending occurs, this policy would lead to additional investment, faster productivity and wage growth, and higher living standards for American households.

This would cost around $400 billion over a decade. Congress should make up that lost revenue from other sources. In my latest Project Syndicate column, I discuss three — one of which is to repeal the Inflation Reduction Act (IRA):

The Inflation Reduction Act of 2022 (IRA) created around two dozen tax credits to encourage domestic clean-energy innovation and manufacturing, and provides a $7,500 credit for individual purchases of new battery-powered or hydrogen fuel-cell electric vehicles. The law will likely cost more than $1 trillion in its first decade, and trillions more after that. Congress and Trump should repeal the IRA and use part of the revenue to cut business taxes. Even partial repeal of the IRA – such as the subsidies for vehicle purchases — would provide ample revenue to offset the cost of tax cuts.

The Wall Street Journal reported last week that some congressional Republicans are on the fence about full repeal, though are quite open to partial repeal. Republicans senators and House members should brace themselves for an onslaught of pressure by conservative activists who would view anything short of full IRA repeal as a major policy and political failure.

The IRA is the type of industrial policy that used to be shunned by leaders of both parties. The likelihood of (even partial) IRA repeal illustrates one of many reasons why previous politicians were wise to be wary of industrial policy: When such a policy becomes a partisan football, like the IRA, its odds of success decline substantially.

And its odds of success were never high. As I wrote in a recent paper for the Aspen Economic Strategy Group, it’s hard to even know what success looks like:

How do we know when we have “revitalized” domestic manufacturing? How will taxpayers know if their investment in slowing the pace of climate change has paid off? What would pass the cost-benefit test? What would fail that test?

One thing’s for sure: We can be confident that the IRA will not succeed in revitalizing manufacturing employment:

These subsidies have yet to generate noticeable manufacturing employment gains, and they will not meaningfully change the long trend in manufacturing employment. Even if these subsidies increased manufacturing employment by 50 percent — a huge increase — that would merely return the manufacturing employment share to its level from two decades ago, far from the golden era of manufacturing in the decades following the Second World War.

Moreover:

President Biden’s IRA has also (predictably) provoked an international response. Its subsidies for clean-energy projects have so dramatically tilted the playing field toward the United States that French president Emmanuel Macron warned that the IRA could “fragment the West.” This, at a time when geopolitical challenges with respect to Russia, China, and Iran make Western cooperation and coordination more important than at any time since the end of the Cold War.

The US’s embrace of industrial policy has created a permission structure for other nations to do the same. Unsurprisingly, South Korea and the European Union have responded to the IRA with their own subsidies. A tit-for-tat subsidy war distorts relative prices, reduces economic efficiency by prioritizing political ambition over comparative advantage, and reduces the ability of any one nation’s subsidies to achieve their own self-defined goals.

And more generally:

Industrial policy works better in theory than in practice. It often fails because real-world factors, limits on state capacity, and competing political objectives often prove to be insurmountable obstacles. All the old questions about industrial policy are worth repeating in light of its revival: Why should we expect the government to do a good job of picking winners and losers or to allocate scarce resources better than the market? If the government intervenes in markets, how will it avoid mission creep, cronyism, and corruption?

Instead of industrial policy, Congress should increase funding for basic research. In that AESG paper, I argue that public investment in basic research can pay enormous dividends. And I argue that narrow, specific, achievable industrial policy with clear, nonpartisan goals can succeed, like Operation Warp Speed or DARPA. But government efforts along the lines of the Biden administration will very likely not pass a reasonable cost-benefit test.

And because the IRA will likely not pass a reasonable cost-benefit test, Congress should absolutely raid it for revenue to reduce business taxes without increasing the budget deficit. If that happens, policy-makers should relearn a valuable lesson: Partisan industrial policy faces many obstacles to success, one of which is that, often times, the other party wins the next election.