Undercover Economist

The economics of discrimination

The Undercover Economist – FT Magazine, 19 November 2005


My wife can buy cheaper car insurance than I can because she’s a woman and, fairly obviously, I am not. Is this sexual discrimination? Some people would claim that wrongful discrimination is treating somebody differently because they are a member of an identifiable group – a woman, for example, or a black person. That is a sensitive definition but not a useful one. I spent more time in my 20s pursuing girlfriends than boyfriends, and I must admit that this bias was everything to do with the fact that the girls were all members of an identifiable group.

In any case, it can be discriminatory to treat two people identically. Women are, on average, safer drivers and cheaper customers for insurance companies. If they were offered the same insurance rates as men, that would be discrimination.

This suggests a plausible way of working out whether discrimination is going on: if a company’s management is willing to give up profits to exercise a prejudice, then the discrimination is real. An insurance company that offered equal rates to men and women would be leaving money on the table, because if the company cut rates to female customers it could attract extra, profitable business.

If companies are discriminating against employees or customers from some minority group, only the most profitable members of that group will be tolerated: customers who pay well over the odds and employees who are extraordinarily productive or who agree to work for very low wages.

Once you start thinking like this, you realise that certain kinds of discrimination are likely to cause more damage than others. Corporate discrimination is surprisingly ineffective, because it is so tempting to gain a competitive advantage by swallowing your prejudices. Victims of discrimination make attractive customers and employees, since they are offered high prices and low wages elsewhere. A business with a profit margin of 5 per cent and a wage bill of 20 per cent of revenues could double its profits if it could lower its wage bill by a quarter. A chief executive who staffed his business with cheaper workers from downtrodden minorities would look like a financial genius.

That suggests that prejudice on the part of managers would have to be very severe to have a big impact. Discrimination from the public, on the other hand, can be crippling. If customers don’t want to be served by women or ethnic minorities, you will struggle to make money ignoring them.

All prejudice is shameful, but some prejudices are more harmful than others. Under apartheid, where the minority whites could not simply isolate their economy from the blacks, the average white South African was paying dearly for the deep prejudice in his society. This helps to explain why the system collapsed. In the US, in contrast, whites outnumber blacks nearly 10 to one. The economic cost to whites of discriminating against blacks is tiny. But as a small, poor minority suffering from discrimination, blacks suffer a heavy economic price even from mild prejudice.

Free markets tend to work against discrimination. After Rosa Parks sparked the boycott of public buses in Montgomery, entrepreneurs stepped in to offer cheap rides in cabs, despite the city council’s efforts to fix a price floor. The city pressured local insurers not to cover cars used in car pools. That bullying was ignored by Lloyd’s of London, which no doubt saw a profit opportunity.

Closer to home, car insurers didn’t need training in gender sensitivity to offer women a cheaper deal. All this is encouraging. Less encouraging is that 50 years after the Montgomery boycott, much milder prejudices than those defied by Rosa Parks can continue to wreak such high economic costs on minorities. The economics of discrimination sometimes spread a little poison a long way.