Break up big banks? Why size isn't the real issue

Banks that were "too big to fail" came into public consciousness during the Great Recession, but despite some legislative and regulatory effort to eliminate that problem, they haven't really gone away. And now with the 2016 presidential campaign in full swing, both Bernie Sanders and Hillary Clinton are once again calling for an end to banks that are too big to fail.

And they've found an ally in the Minneapolis Federal Reserve Bank's new president, Neel Kashkari, even though he's a Republican. Kashkari made waves this week by starting his new job with a pledge to urge Congress to tackle what he sees as a major flaw in the U.S. financial system: giant banks that need breaking up.

Still, the overarching question remains: Will breaking up American's biggest banks make the financial system safer? Let's examine some of the issues involved.

Size isn't the most important factor

During the Great Depression, banks were relatively small compared to today, yet that didn't stop a wave a small bank failures from inducing a bank panic and financial meltdown. The failure of a large number of small banks can be just as problematic as the failure of a few large banks.

Interconnectedness is the biggest problem

When banks are highly interconnected, the failure of one institution can cause a wave of subsequent failures. Suppose, for example, that one bank has borrowed a large amount from another bank to finance a big loan to a customer. And suppose the second bank, in turn, owes money to other banks and so on. This type of arrangement is common in the nontraditional banking system, which includes investment banks, hedge funds, mutual funds, mortgage companies and other such institutions.

If the first bank can't make its payments because the loan didn't pay off, the second bank will find itself short of funds and in trouble as well. And that trouble can be passed down the line, inducing a "domino effect" of bank failures.

Thus, regulations that limit the degree of interconnectedness between banks can do more to prevent bank failures than simply breaking big banks into smaller pieces.

Specialization is another stumbling block

When banks specialize in a particular type of loan, they're at risk if that sector of the economy has trouble. Banks specializing in agricultural loans have trouble when farmers have bad years. AIG, a specialist in credit protection, got in trouble during the financial crisis, and so on.

It doesn't matter if one big bank is exclusively handling loans of a particular type or if 100 smaller banks specialize in the same type of loan. If that economic sector goes into a downturn, the one large bank or the 100 small banks are in danger of failing. Therefore, regulations that ensure banks have a diversified portfolio of loans can be very helpful in promoting bank safety.

How big should banks be?

The optimal size of a firm is based on what economists call the "minimum efficient scale." This is the size at which the firm's costs of production are the lowest, hence it's the most efficient point for the production of goods and services.

For some firms, the optimal scale is relatively small, and an industry has room for a large number of firms. But in other cases, the minimum scale can be quite large, leaving room for only a few firms in an industry to operate at this level. The most common reason for this is the presence of a large initial investment that needs to be spread over many customers.

The costs of providing phone service, for example, would be prohibitively large if the costs of the network were spread over just 10 customers. But when spread over millions of people, the cost per person is much, much lower, and the firm can charge lower prices to its customers.

What's the optimal scale for banks? Unfortunately, this is a difficult question to answer, and the empirical results are mixed. But it doesn't appear that scaling banks down would have a large impact on their costs. They would still be able to make the same quantity of loans on the same terms.

Keeping political power in check

However, even if the optimal size of banks is relatively large, there's still a reason to consider breaking them up. The wealthy owners and managers of large, too-big-to-fail banks have considerable political power. That power can be used to tilt legislation in their direction, particularly regulatory interventions intended to make the system safer.

I believe this is a real problem, as the continued and often successful attacks by the financial industry on the Dodd-Frank financial reform legislation passed after the financial crisis demonstrates.

Putting this all together, from my point of view the reason to break up banks isn't to make the system safer. That's better accomplished through regulations that limit interconnectedness, ensure diversification of risks and prevent other behavior that might undermine the financial system's stability.

Instead, breaking them up is the best way to prevent "political capture" by the financial industry. Politicians whose election prospects depend on contributions from financial institutions won't make the choices we need them to make to ensure that the financial system is as safe and sound as it can be.