In Michael Mann’s 1995 movie masterpiece Heat, bank robber Neil McCauley (Robert De Niro) explains to panicked bank customers that he has no intention of hurting them: “We’re here for the bank’s money, not your money. Your money is insured by the federal government. You’re not gonna lose a dime.”

As clients at Silicon Valley Bank (including my employer, Vox Media) learned last week, McCauley’s promise isn’t quite true. Deposits at banks and credit unions are indeed insured by the federal government, but only in cases of bank insolvency, and only up to $250,000 per person, per bank. If an individual or business deposits more than $250,000, that amount could vanish if the bank fails.

As someone who enjoys getting paychecks from the business that employs him, this concerned me, until the Fed, Treasury, and Federal Deposit Insurance Corporation (FDIC) eventually stepped in to guarantee all Silicon Valley Bank deposits. But while even those with more than $250,000 deposited were made whole this time, uninsured US depositors do, in practice, lose money with some frequency. A 2020 report from economists at the FDIC, which insures deposits at banks, found that a significant share of bank failures from 1980 to 2013 resulted in uninsured depositors losing some funds. From 1980 to 1987, as the savings and loan crisis began, some 24 percent of bank resolutions resulted in losses for depositors; from 2009 to 2013, after the financial crisis and reforms meant to protect deposits, 6 percent did. In the non-crisis period from 1992 to 2007, a staggering 65 percent of bank failures resulted in deposit losses.

But there is one state in America where this is simply not a problem. Since 1934, not a single depositor at an insured bank in Massachusetts has lost a dime. All deposits above the $250,000 federal limit are insured by a private entity — the Depositors Insurance Fund (DIF). Massachusetts law requires banks and credit unions to pay premiums to the DIF, which in turn guarantees depositors in case of bank failure. Such failures are rare in Massachusetts these days; only one bank, Butler Bank in Lowell, failed in the financial crisis. But in the S&L crisis in the ’80s, 44 banks failed in the state, over a quarter of all banks. In each case, depositors were made whole.

How did this come about? What does it mean for banks and consumers? And is something that came out of Massachusetts worth emulating for the rest of the country?

The Massachusetts deposit insurance experience [fireworks erupt, people hoot and holler in excitement]

Generally, when only one state in the US has a particular policy, it means one of two things. Option one is that it’s the only state that’s ever tried that policy, due to its particular, unique history. The Bank of North Dakota, owned and run by the state government and created during a short stint of quasi-socialist control of the state, is one of the former; no other state has ever set up a government bank, despite many attempts by advocates.

Option two is that a bunch of states have tried the policy, but all but one have since abandoned it. Maryland’s health care system, where all insurers pay the same prices, is one example, and Massachusetts’s banking system is another. A number of states tried deposit insurance systems in the wake of the Panic of 1907, but all of them failed due to either rapid deflation after World War I or the Great Depression a few years later. A few decades later, some states tried again to institute private deposit insurance programs; Ohio was the first, in 1956, while California, Iowa, and Kansas were the last, in 1981. But most of these systems wound up collapsing by the early 1990s, unable to pay what they owed.

Walker F. Todd, a lawyer then at the Cleveland Fed, argued in a 1994 paper that the state failures were largely about regulatory capture. The state funds couldn’t supervise banks enough to discourage risk-taking, or charge them enough in premiums so that the fund could cover them in the event that their inevitable risk-taking inevitably led to big losses. He blamed this on the banks’ influence over state political systems, which meant that state legislatures were inclined to allow risk-taking and disinclined to increase premiums.

So how did Massachusetts avoid this in the 1980s? For one thing, it used the FDIC as a backstop. Stone notes that in 1956, the DIF transitioned to only covering deposits in excess of the FDIC limit. That meant that the bulk of the cost of bank failures still fell on the Feds — unlike in, say, Rhode Island, where banks mostly lacked any federal insurance, relying instead on the state program. In 1985, after other state insurers began collapsing, the Massachusetts commissioner of banks required banks in his state to get FDIC insurance, too. The result was an unusually resilient insurance system that withstood even the spree of 1980s bank closures.

The DIF, which in its current form combines a number of predecessor groups that specialized in insuring different types of banks and credit unions, takes pride in the fact that depositors in its institutions have never lost a cent. “Even during the 1980s, when nineteen DIF member banks failed, the DIF insured $250 million in excess deposits,” Anna-Leigh Stone, an economics professor at Samford University who has studied the Massachusetts system, notes in a paper.

Stone’s research finds that the insurance makes people willing to leave more money in Massachusetts banks: Compared to similar banks elsewhere in New England, she finds that Massachusetts banks held 5 to 6 percent more deposits over the FDIC’s limit. Somewhat surprisingly, she finds that the Massachusetts banks did not use these additional deposits to make more loans, perhaps because the additional scrutiny of the DIF discouraged risk-taking.

Another paper by economists Piotr Danisewicz, Chun Hei Lee, and Klaus Schaeck published last year also examines the Massachusetts system. Unlike Stone’s, this paper finds that both deposits and lending are higher in Massachusetts compared to a control group of non-Massachusetts banks with branches in the state. But the increased loans, they find, tend to be prudent and not unduly risky. Stone told me the difference was due to slightly different control groups in each study, but both papers found fairly responsible behavior on the part of covered banks.

The risks of insuring all deposits

Sounds nice! But deposit insurance is not without its problems and critics. If insufficiently funded, or lacking a formal government promise to rescue the fund if it becomes insolvent, private insurers can fall apart, as happened in many states. But the bigger concern many economists have about deposit insurance is that its existence might make bank failures more common.

Charles Calomiris at Columbia University is the most famous exponent of this view, across a number of papers. Writing in the Wall Street Journal about Silicon Valley Bank this week, he argued that “Virtually every academic study of deposit insurance shows that it promotes, rather than reduces, banking system fragility.”

The key claim here is that people keeping deposits at banks should be watchfully examining those banks, seeing how stable or solvent they appear, and switching between banks based on what they find. This would exert discipline on the banks to behave more responsibly. While it’s unrealistic to expect most people to exert this kind of diligence, Calomiris argues that we could be “free riders on informed discipline” exerted by more watchful depositors, like other big banks.

By contrast, when deposit insurance is generous, clients have little reason to check up on the place where their money is stashed. For instance, I use USAA for checking and savings, my balances are way below $250,000, and I simply do not worry at all about what happens if USAA fails. This problem is known in economics as “moral hazard”: the tendency for actors to take on more risks if they’re insulated from the consequences of those risks. USAA is not a particularly risky bank, but whatever risks come with parking my money there are the FDIC’s problem, not mine. I’m totally insulated from any risk. And because I’m insulated from risks, I’m not going to discipline my bank, which means that bank in turn is going to take bigger risks.

Factors like this are why some systems are moving away from expansive deposit insurance. For instance, Germany, which has long had a byzantine system of voluntary private deposit insurance schemes that amounted to near-full coverage of deposits, is in the process of reducing the coverage those schemes offer, in part to align with other EU members that have less generous systems. The hope is that this will reduce the moral hazard problem going forward.

While Calomiris is right that several studies, including his own, have found evidence that banks generally take more risks when they get deposit insurance, the economists who’ve studied Massachusetts’s system noted that they found little evidence of it in that specific context. “We were kind of surprised that in the Massachusetts setting we don’t find evidence for it,” Schaeck told me. “In fact, what we see in this specific setting is that banks seem to be more prudent.”

The Silicon Valley Bank collapse also presents an interesting challenge to the theory that without insurance, depositors will exercise useful disciplinary oversight over banks. In a way, this is precisely what led to the bank’s failure. The vast majority of its deposits — 89 percent as of the end of last year — were not insured. Depositors had a huge incentive to monitor the bank. And they did. The bank’s failure was precipitated by massive withdrawals, which were in turn precipitated by depositors reading an update on its finances posted on March 8, inferring that it was in trouble, and publicizing this finding on social media. Depositors and tech investors like David Sacks and Jason Calacanis took to Twitter to encourage people to panic:

The news spread so fast that entrepreneur Max Cho told the Wall Street Journal he pulled out of the bank when he noticed fellow passengers on a shuttle bus to the Montana ski resort Big Sky frantically working their phones to pull money out.

Here’s the thing: Calacanis and Sacks were behaving exactly like depositors are supposed to behave in a situation without deposit insurance. They’re supposed to monitor the financial health of the place they’ve put their money and pull out if they sense their money is in danger, in part as a signal to other, less plugged-in depositors to do the same. But this time, that behavior contributed to a run on the bank and its eventual collapse.

The distinction between “responsible depositor oversight” and “starting a massive bank run” turns out to be rather fine in practice.

Should America follow the Massachusetts model?

I asked researchers who’ve looked into the Massachusetts system if they think it could be a potential model for the US as a whole. They were cautiously supportive. “It’s a system that could and probably should be explored,” Stone said. “As we argue in the paper, there are many, many benefits,” Schaeck agreed. He likes that the Massachusetts system is private and effectively managed by a club of bankers, who have an incentive to check up on their competition. But Stone and Schaeck both note that Massachusetts’s banks are rather small, and none of them are large, systemically important institutions like JPMorgan Chase or Bank of America.

Those two, Wells Fargo, and Citigroup each have over $1 trillion in deposits, per last quarter’s FDIC filings. Chase and BoA have over $2 trillion. The FDIC, by contrast, has only $128 billion in its deposit insurance fund. While not all of the big banks’ deposits are insured, it’s safe to say that if JPMorgan Chase failed tomorrow, the FDIC’s deposit insurance fund would be emptied very quickly. Realistically, there’d be a large-scale bailout as in 2008 to prevent further economic fallout, and expectation of those bailouts can and does lead these banks to act recklessly. They’re a true moral hazard.

The FDIC’s fund would have to be significantly bigger if it decided to insure all deposits, especially at those large, systemically important banks. And that could be a political problem. The fund comes from premiums charged to banks; Massachusetts’s DIF charges its own premiums, meaning operating a bank there is more expensive. Banks really hate it when you make them pay more premiums. Just this past October, they cried havoc when the FDIC proposed raising the price of deposit insurance, stating, “banks are in excellent financial condition, so the FDIC’s action is a preemptive strike against a nonexistent threat.” Whoops.

That statement sounds ridiculous now, but the fact remains that banks wield considerable political influence and can often stop things like higher FDIC assessments or push related deregulation. They would argue that an increased premium would be passed on to consumers in the form of lower interest rates, and they might be right about that.

Another option would be to cut out the banks entirely. A proposal known as “FedAccounts,” from three financial regulation specialists — Vanderbilt’s Morgan Ricks, Columbia’s Lev Menand, and UC Law SF’s John Crawford — would let everyday individuals and businesses keep accounts at the Federal Reserve. Unlike a normal bank, the Fed wouldn’t lend out these deposits, so there would be no risk of them being lost in the way Silicon Valley Bank lost its deposits. As Ricks told me back in 2020, when the proposal was going around as a way to deliver stimulus payments, “Virtually every financial crisis in US history and world history has involved runs on money instruments or money substitutes. Runs on this stuff is the preeminent source of acute macroeconomic disasters.”

Indeed, what we just saw at Silicon Valley Bank was a classic run on money. If its clients had been able to keep money at the Fed, earning the normal Fed interest rate, none of this would have happened.

Of course, banks would hate this plan even more than increased FDIC fees. But given the events of the past week, that could be a good reason to try it.