Charles Calomiris and Stephen Haber on Fragile by Design
What this covers
Charles Calomiris and Stephen Haber, the authors of Fragile by Design, discuss with Russ Roberts why some countries' banking systems remain stable while others—particularly the United States—suffer repeated crises. Their argument centers on the inescapable role of politics in finance. The authors contend that banking fragility is not an accident but a deliberate outcome of political bargains struck between coalitions seeking to use regulation as a hidden tool for redistribution. Rather than treating these arrangements as anomalies or engineering problems, they frame them as rational responses to deep constitutional conflicts of interest: governments both regulate banks and borrow from them, enforce debt contracts on voters, and manage the liquidation of failed banks whose depositors are voters.
The conversation moves through American history to show how unit banking—laws restricting branch expansion—persisted for nearly 150 years because agrarian and small-banker coalitions could win at the state level in a federal system. Canada, by contrast, developed centralized chartering and an insulated Senate that prevented populist coalitions from consolidating power, producing a far more stable system. The authors trace how the same populist pressures that shaped Canadian and American banking rules produced opposite outcomes depending on institutional structure. They examine how post-1970s American policymakers, blocked by checks and balances from direct fiscal redistribution, turned instead to hidden transfers through GSE support and arcane regulation—ultimately destabilizing the system. The discussion includes the 2008 financial crisis, the role of Fannie Mae and Freddie Mac, and the moral hazard embedded in deposit insurance, tracing the bipartisan continuity of mortgage-risk subsidization across presidencies.
Calomiris and Haber argue that banking crises are not accidents but the predictable product of political bargains; countries like the US are 'fragile by design' because populist coalitions repeatedly use the banking system as a hidden tool for redistribution, while Canada's centralized, insulated political institutions produced a stable system.
- Banks and governments have inherent conflicts of interest that make removing politics from regulation impossible
- US Federalism let agrarian/unit-banker coalitions sustain inefficient unit banking for ~150 years
- Canada's centralized chartering and insulated Senate blocked the populist coalitions that won in the US
Political conflicts of interest are inherent to banking regulation because governments regulate, borrow from, and enforce debt through banks.
- Bank regulation enables hidden transfers because the public cannot understand its arcane implicit taxes and subsidies, and the bill isn't paid until a crisis forces bailouts; this lets even ideologically opposed politicians do deals invisibly, making the temptation to redistribute through banking very strong.
“because it occurs off the budget and because it occurs in a way that is very hard for the average person to understand, and because it is not--you don't have to pay for the bill until the banking crisis occurs and everyone needs a bailout, it's not seen. And so there's a very strong temptation to do this.”
- Banking systems are fragile by design because politics cannot be removed from bank regulation: governments both regulate and borrow from banks, governments use police power to enforce debt contracts on voters who default, and governments liquidate failed banks whose depositors are voters—creating three inherent conflicts of interest that keep politics embedded in banking.
“banking systems are fragile by design because it is impossible to take politics out of bank regulation. And it's impossible to do so because there are inherent conflicts of interest between government and banking systems such that banks need governments and governments need banks.”
- Since the 1970s the US has been tempted to use the banking system as a vehicle for off-balance-sheet income redistribution because checks and balances thwart direct fiscal redistribution, leaving hidden subsidies through arcane bank regulation; this distorts incentives for everyone and makes the US crisis-prone, unlike Australia which addresses inequality directly through fiscal policy such as down-payment assistance.
“There's always been a temptation, certainly since the 1970s in the United States, to look to the banking system as a vehicle for income redistribution off balance sheet.”
GSE capital advantages and CRA-driven mandates created incentives to lower underwriting standards, which spread throughout the market once they moved.
- Because GSEs faced far lower capital requirements ($1.60 vs $4 per loan) plus a guarantee fee, they could buy loans, securitize them into MBS, and sell them back to banks more cheaply than banks could hold them—creating a tremendous incentive for Fannie and Freddie to be in the business of buying banks' bad loans.
“You had to put $4 in capital behind that loan while you held it... I, as Fannie or Freddie, only have to $1.60 in capital behind that same loan... This creates tremendous incentive for me as a government GSE to be in the business of buying your bad loans and selling it back”
- Fannie and Freddie were dragged into buying CRA-type loans kicking and screaming after activists pushed Congress (the 1992 GSE Act required them to repurchase Affordable Housing loans), and in return extracted two concessions: regulation by a weak HUD unit rather than the Fed, and capital requirements about 40% below commercial banks—setting up the dynamics that preceded the private investment banks entering a decade later.
“they followed in the sense that they are dragged in kicking and screaming into this deal. They don't want to buy these CRA loans that banks are making. The activists push in Congress to basically make them do it. And the thing that Fannie and Freddie extract in return is two crucial features.”
- The CRA was not itself a bad idea; the problem was the logic of the CRA coupled to the mega-merger movement, which gave Fannie and Freddie incentives to lower underwriting standards, and once they lowered them for everybody the whole society could pile into deals too good to be true.
“It's not that I think the CRA was a bad idea. It's that the logic of the CRA coupled to the mega-merger movement gave rise to incentives for Fannie and Freddie to lower their underwriting standards. And once that happened, they had to lower them for everybody.”
Unit banking laws served political interests of local bankers and farmers, not efficiency, until technology and inflation undermined the coalition.
- The New Deal's deposit insurance, Glass-Steagall, and Regulation Q (banning interest on checking and capping savings interest) were designed to discourage banks from competing and customers from moving savings, propping up the unit banking system and preventing the consolidation that would otherwise have occurred.
“in order to prevent the consolidation of banks, which is what would have happened in the absence of, for example deposit insurance and Regulation Q, which made it illegal to pay interest on checking accounts... It's all done to discourage banks from competing with one another.”
- Unit banking laws barring out-of-state branching and branch banking had no efficiency or stability rationale; they were a political arrangement benefiting local unit bankers with captive monopoly markets and prosperous local farmers who had a captive lender, at the cost of everyone else and system stability.
“this is clearly a political arrangement. There is no efficiency or stability criterion by which you would do this. It did however work quite well for local unit bankers because they had captive local markets. Essentially they had local monopolies.”
- The agrarian-unit-banker coalition fell apart because technology (networked ATMs and computer credit assessment let banks evaluate borrowers at distance and skirt branching laws, upheld when the Supreme Court ruled in 1985 an ATM is not a branch) plus 1970s inflation that made Regulation Q-capped deposit rates strongly negative, driving deposits into money market funds and triggering the S&L Crisis.
“computer technology allowed bankers to assess borrowers at a distance... it also meant that banks could skirt the laws governing branching by simply opening up an ATM anywhere they could rent 6 square feet of space.”
- The Federal Reserve was created as a reaction to the Panic of 1907-1908; after studying Canadian, Mexican, and German branch-banking models, the national monetary commission rejected them and instead retained unit banking while propping it up with 12 Regional Fed Banks that could lend liquidity to unit banks in crises.
“they rejected all those models in favor of retaining unit banking, but propping it up by creating 12 Regional Fed Banks that could essentially lend to unit banks by increasing liquidity in times of crisis.”
Conservative branch banking regulation did not prevent financial stability or economic performance, as Canada demonstrates despite geographic disadvantages.
- Canada achieved a very high GDP by world standards despite scarce arable land (Iowa has more corn/wheat farmland than all of Canada), dispersed natural resources, and exposure to terms-of-trade swings as a primary-product producer—much of it attributable to financial stability—making it hard to argue Canada paid an efficiency price for its conservative banking system.
“Canada is a country with a tremendously high GDP, by world standards. And in fact has accomplished that even though it's mostly a primary-product producer... A lot of that I think is due to the stability of the Canadian financial system.”
- The 19th-century US had tens of thousands of banks unable to open branches, meaning each bank was tied to its local economy, could not move funds during runs, could not spread risk across regions, and could not capture scale economies—making the system inefficient and crisis-prone.
“every bank which you would think of as a branch is a bank unto itself. That means that in the event of a bank run it's very hard to move funds from one branch to another because there aren't branches. It means you can't spread risk across regions”
- Federalism does not always function as a check that prevents bad policy; it can instead enable other types of bad policy, and Canada also outperformed the US in monetary and trade policy, suggesting this dynamic extends beyond finance—an under-researched flip side to market-preserving Federalism.
“In most of the literature about Federalism it's all about market-preserving Federalism. Federalism is always good. There is this flip side of it, and I think it's been under-researched.”
- The public should extract two heuristics: do not count on elected representatives to watch out for your financial interests as a taxpayer, and any time a politician claims to have found a free lunch—a marvelous subsidy nobody pays for, especially through banking—reach for your wallet, because the bill comes due as a crisis.
“any time a politician tells you that he's found a way to create a free lunch and that there's going to be this marvelous subsidy, and nobody is going to pay for it, reach for your wallet. Especially when that subsidy is coming through the banking”