So far Congress has produced three COVID-19-related packages: a small bill funding COVID -19 vaccine research, testing, and equipment; a $100 billion bill to fund testing, shore up food aid programs, and fund sick leave for employees in small business; and the recent mammoth $2.2 trillion bill expanding unemployment insurance generosity, providing tax rebates for households, and delivering small business loans and large business support.

These bills, whatever our disagreements with specific provisions, appeared to recognize two truths. First, that this was not an ordinary recession (indeed, there is a clear short-term trade-off between reducing economic activity and getting a handle on the virus). Second, that the virus itself and the lockdowns will cause a lot of economic pain, indiscriminately plunging household and business income in certain sectors and jobs.

A global pandemic was an economic shock very few businesses could have foreseen or prepared effectively for. Using public health justifications, the government was also forcing certain businesses to close – willing a pause in much economic activity. As an insurer of last resort, or even just to provide relief or compensation to those forced to close or lose income, there was thought to be a role for taxpayers to cover lost income or help keep businesses afloat.

One can quibble with the contours of lockdowns, but the rationale for borrowing might have been dubbed “anti-stimulus fiscal expansionism.” Yes, sending out checks to most people looks similar to measures seen in any past “stimulus” bill. But the lion’s share of the extra government borrowing aimed not to encourage economic activity but to replace or discourage it. Businesses were obtaining support for not producing; workers were being paid more for being unemployed or furloughed. Only in health care was government consumption spending rising.

Nothing about this was the traditional Keynesian argument for government borrowing to “create jobs” or “boost aggregate demand.” There was no talk of “infrastructure investment” and little chat about “putting money into people’s pockets to get them spending.” At least until now. In the past week, President Trump has talked up how a fourth package should include a $2 trillion infrastructure program. Pelosi, meanwhile, desiring infrastructure too, inexplicably wants a fourth bill to repeal the cap on state-and-local tax deductions in the federal income tax – a massive tax cut to the rich.

Not just that, but their reasoning now echoes traditional Keynesian talking points. Pelosi, in particular, talks about “creating well-paying jobs” when it comes to infrastructure. Her justification for the pro-rich SALT cap repeal is premised on boosting consumption. She reckons “they’d have more disposable income, which is the lifeblood of our economy, the consumer economy that we are.” It sounds a lot like she’s a convert to the “trickle-down economics” of left-wing caricature.

There are all sorts of reasons to doubt the efficacy of Keynesian demand management in ordinary recessions. But now, such talk is even more dangerous. In the near-term stopping the spread of the virus will result in a contraction in the supply of goods, both from less production and impaired supply chains. Trying to ramp up government infrastructure projects or consumption among the rich, who tend to spend disproportionately on services affected, therefore risks significant price rises and rationing in particular sectors, crowding out vital activity in the health care sector and other key parts of the economy.

But even when, we hope, the economy rebounds, this type of advanced “stimulus” could be dangerous too. As former IMF chief economist Olivier Blanchard has explained, the unprecedented nature of this crisis leaves us unclear about how the economy will “normalize.” It might be that consumer spending rebounds sharply, in which case these extra funds would throw fuel on the fire unnecessarily. Or it may be that the normalization is gradual, with many supply chains continuing to be impaired, firms reluctant to invest and consumers reticent to spend because of the on-going threat of a new spike in re-infection.

Either way, the best action for now, even in a Keynesian sense, would be “wait and see.” A stimulus-focused expansion right now works against the other measures Congress has pursued. And it would be very risky to commit to before knowing the shape of the recovery.

Write by Ryan Bourne, occupies the R. Evan Scharf Chair for the Public Understanding of Economics at the Cato Institute.