Philadelphia Reflections

A New Definition of Poverty

Ann Carrns has recently written a short column in the New York Times to help banks afford to lend to marginal cases, but I would propose it actually contains a new and better way to define poverty. As some wag suggests, we have currently worked ourselves into the corner where almost no one is eligible for a loan unless he doesn't need a loan.

That slander isn't exactly the case, and it trivializes the problem. The interest a bank must charge to break-even includes the cost of those who default. If the bank underestimates the default rate, it will go broke, or at least must raise interest rates for those who do not default. Therefore, all interest rates are a little higher than they need to be, to protect the bank against mis-predicting the default rate. Furthermore, the bank must assign someone to make that judgment, case by case ( "underwrite the risk"), and that raises the overall cost still greater. It could be more profitable for banks just to skip questionable cases, so this "opportunity cost" must be guessed at and added. It's actually quite a considerable cost.

To guess wrong by underestimating the risk is equally and visibly expensive in reverse. To some extent, errors on the high side cancel out errors on the low side, so combining and pooling the two helps the problem, and enlarging the pool by combining banks is a good thing. So continuous computerized pooling is a big cost saver. The bigger the pool, the narrower the spread. Finally, an optional insurance layer of risk, if large enough, should soon result in a national market with a narrow spread which is considerably lower than present underwriting methods -- steadier, and permitting the acceptance of riskier poor clients within a more clearly defined range. This approach should improve the reputational risk of banks in general, whether they participate or not, consequently lowering interest rates by isolating a narrower layer of uncertainty.

And finally, it should establish a nation-wide definition of poverty. The government can assume that anyone who cannot afford the cost of this system defines himself as below the poverty level. It does this by stating that no amount of internal rebalancing of the private sector is apparently sustainable without raising somebody's rates. That is, someone may or may not be poor by other definitions, but has reached the limit of imagination within the private sector. It is thus a definition of the poverty level, constantly changing with the money supply, the value of the dollar, and the locality. Loans to a suboptimal client are permissible but are dangerous to the bank who issues them. This definition of poverty is the lower border of a band. The upper limit of that band is the level at which private-sector banks ignore the warning and make the decision to lend, in spite of it. The system may uncover non-market manipulation by this approach but, if so it could be useful for the voters to detect it, early.

Originally published: Saturday, February 17, 2018; most-recently modified: Friday, May 10, 2019