We need a vaccine for COVID-19, and we need it as soon as possible.

May 1 was the disease’s deadliest day in the United States. Cases continue to rise, even in places residents thought would be missed. We still have no real way to find outbreaks before they start.

Still, across the country, to varying degrees, businesses are starting to reopen. People here and there are starting to go back about their lives. Adherence to safety guidelines appears spotty; some are willfully flouting these practices while in large groups. Even though they run contradictory to White House guidelines, the president himself has supported these protests.

In that environment, it will be difficult to hold together any sort of unified response that keeps the virus at bay long enough for the usual vaccine development period.

That’s a shame. In trying to trade off good public health for a rejuvenated economy, we’ll likely get neither — making the off-ramp to this outbreak longer, deadlier and more disruptive.

That means we need a vaccine sooner rather than later. Developing one, however, is a long process — the federal government should do everything in its power to speed it up.

The vaccine-approval process is built for safety, not speed. It typically takes years. Even in a rush, the 12- to 18-month timeline cited by the Trump administration has been called too optimistic.

The costs of delay are enormous. The U.S. response can already be measured in trillions of dollars, and that sum appears to be woefully inadequate. The coronavirus has led to about 70,000 confirmed deaths, and probably many more. The restrictions on movement, though not nearly as restrictive as those in countries that have more successfully fought the outbreak, are starting to fade, all but guaranteeing a rebound in cases.

With that in mind, it is likely impossible for the government to overspend while helping produce a vaccine.

Soon, nearly a dozen vaccines will be in the early stages of human trials, the Associated Press reported Monday.

One candidate, from Pfizer, may be ready for emergency use by September — though not for wider use until much later. The company is testing four variations at the same time in the hopes of finding the right one more quickly, though that’s no guarantee that they will.

Also, some vaccine producers are making millions of doses of promising drugs, even though they know most of them won’t work — if one of drugs works out, then they won’t have to wait to manufacture it, saving significant time.

Because of all the speculation and overlap, these are costly efforts. The federal government is already supporting two drugmakers, and plans to ultimately support three or four more.

It can and should do more. Bill Gates has suggested using public and private money to build factories for the top vaccine candidates. Most would fail, but the investment would save months — and lives and money.

However the government supports vaccine makers, it must be comprehensive, and built for speed. There should be billions of dollars for research and development. The regulatory process should be truncated in ways that recognize the urgency of the problem.

The federal government failed over the last three months to produce the supplies necessary for widespread testing, which is critical to restarting activity safely. The Trump administration failed, and in the rush to reopen, Americans will pay.

We cannot afford to look back six months or a year from now, and say the same thing about a vaccine.