Wednesday, May 11, 2016

Bonds and Prisoners

This is one of my posts linking apparently disparate things:

  • Donald Trump has gotten some press with his statements on the national debt and how he would handle it.  Apparently he's finally settled on the idea that if conditions are right, we should buy back some debt.  The conditions would have to include running a current budgetary surplus, which seems unlikely.
  • Here's a post attempting to explain why, given a big drop in crime over recent years, we still have so many people in prison.  "Most prison sentences in the United States are for more than one year. Thus, even if crime goes down, and the number of new incarcerations goes down, the total prison population can still increase — because most of those incarcerated in previous years are still behind bars."  One graph of the first point.
  • The writer says changes in the rate of incarceration will track closely with the crime rate.

 The hangup is stock and flow. In both cases--the total US debt and the total US prison population--we're talking a "stock",  a reservoir, the level of which changes if the inflow doesn't match the outflow.

With the debt, the Treasury is selling bonds on a regular schedule, and redeeming bonds as they mature.  If taxes aren't enough to pay the bills, it sells more bonds; if taxes pay the bills, it sells fewer bonds.  So there's no surplus which a President Trump could use to buy back debt. 

With the prisoners, assume the justice system is catching, convicting, and incarcerating criminals (and on average the convict has committed the same number of crimes before capture) at a fixed rate.  (Assumptions always wrong--sometimes the jails are full and criminals are diverted from the system.)  Now you have to assume something about length of sentence.  If sentences served get longer, the stock of prisoners will increase.  If sentences get shorter, the stock will decrease (all else being equal, which it won't be).  The writer fails to make this clear.

No comments: