Salaries at Amazon.com in 2002

A friend of mine worked at Amazon over the summer, and spoke about how it still ran itself kind of like a startup, and expected employees to work long hours for below market wages. While this was fine when the company could credibly promise growth (ie. lots of money and advancement in the future) it doesn't work so well anymore.

An individual's wages sit somewhere between his productivity (on the high end) and his next best opportunity (on the low end). After all, no matter what anyone says, you cannot pay someone more than their marginal productivity -- it means you're losing money for every hour that person is your employee. Similarly, no one willingly needs to work for worse than their next best opportunity because (by definition) they can always change to that.

An employer can gain economic advantage by paying below market wages if he promises growth (ie. lots of money in the future). This is clear in option-laden tech companies, but it's also been used by other growth companies, such as Southwest and Enterprise Rent-A-Car. "But!" I hear you economist argue, "an employee willing to defer wages merely asks for more money in the future. If you discount those future wages back to the present, their net present value is the same as a competitive wage would be now." (This means that since money in the future is worth some (less) money now, people who defer cash now will have to be paid more in the future to agree to do it, and the whoe thing ends up being a wash.)

The real profit (economic rent) is made in the little gap after when a company breaks its promise (stops growing) but before employees bid up their wages. It sonuds like Amazon's hit the stage now where no one is under any illusions about its growth prospects (OK, but now fantastic) so want to be paid market wages now.