Macroeconomists seem overwhelmingly worried that the COVID-19 shock could cause a significant recession, if unaddressed by policy. This message has still yet to get fully through to most policymakers, it seems.

Much of the policy discussion so far has focused, admirably enough, on targeting aid to workers likely to be directly affected by the virus itself and to reduced work caused by the “social distancing.” The responses to these issue have been proper: boosting the capacity of the health system, mandating emergency paid sick leave, reforms to unemployment insurance (UI), and providing free testing.

These measures are smart and well-meaning. However, they need to be supplemented by large-scale stimulus. Simply put, we need to do more to buffer the wider economy against the fallout of the COVID-19 shock.

As workers are laid off from directly affected industries (like restaurants and travel), their incomes will fall and so they will spend less money in even nonaffected industries. Because the industries directly affected by COVID-19 disproportionately employ low-wage workers with little wealth (and therefore little to no savings to turn to to maintain spending when they are laid off), the reduction in spending that will accompany wage losses will be even faster and sharper than in typical recessions. These spending cutbacks will then cause work reductions and income losses in nonaffected industries, and the vicious cycle will deepen.

One prime propagating mechanism that will make this vicious cycle worse if left unchecked is the response of state and local government spending. A negative economic shock causes tax revenues in these governments to fall. Balanced budget rules at the state and local levels will cause spending to contract, putting further downward pressure on economic growth.

In short, the economic shock from COVID-19 will come extraordinarily fast and be very broad and will have a large effect on the economy. This means that even after targeted interventions are undertaken, quick-acting and large economic stimulus will be needed.

What specifically needs to be done? Send cash payments to households and have the federal government take on states’ Medicaid spending for a year. Crucially, each of these should also include triggers to keep them going if economic conditions warrant.

Some have questioned the need or the efficacy of broad-based stimulus like this, arguing that the unique nature of the COVID-19 economic shock will render such stimulus largely ineffective. The argument is essentially that giving people more money in the coming months will not let them spend it at restaurants or for travel because the spending reductions in those sectors are due to “social distancing” and not lack of income.

Those arguments don’t take into account the other critical ways such funds would be used to mitigate an economic shock. People can spend more money in nonaffected sectors in the meantime. Also, getting resources to people sooner means that households can avoid missing payments on mortgages, rent, or cars. This matters not only in the short run, but also means they will emerge from the near-inevitable economic contraction of the next few months in less-distressed financial straits.

Similarly, the contractionary effect of state and local government spending cutbacks tends to come with a bit of lag and will pull down growth even after the initial shock passes. Having the federal government transfer large resources to state and local governments via Medicaid could help keep this spending contraction from happening. All of these measures would hence encourage a much faster bounce back of economic activity once the “all clear” is sounded on the virus’s spread and social distancing measures are relaxed.

This sharp bounce back is important—if the economy is hit by a negative shock and no large stimulus measures are instituted to prime this bounce back, the situation could fester, with joblessness and economic activity lingering. Policymakers really should be thinking about how to make sure the recovery from the shock is “V-shaped.” The last three recessions (the early 1990s, early 2000s and the 2008–2009 recession) all lingered on far too long and did serious damage well past their official end. We need to avoid this dynamic this time.

Finally, and perhaps most importantly, stimulus needs to have provisions that carry on if economic conditions warrant. Passing a stimulus that gets us to the end of the year and then creates a sharp “fiscal cliff” would be a disaster. Right now the political incentives are for incumbent policymakers to move quickly to avoid bearing responsibility for economic distress. In January 2021, with elections two, four, or six years away, these political incentives will be much more blunted.

In short, policymakers need to pass substantial stimulus that is quick-acting but also doesn’t threaten to peter out before recovery is ensured. Time is of the essence.