Mandatory Voting Won't Solve Anything

Peter Orszag's column about mandatory voting gives me an excuse to bring up some other issues with voting.

For one, the participation paradox has always bewildered me.  The participation paradox is the argument that my vote has almost zero effect on the outcome, but has a positive cost, therefore if acting rationally, I shouldn't vote.  However, if everyone acted rationally (or if rationality wasn't randomly distributed among voters), then the system would collapse.

I didn't want to accept this.  What I recently decided, however, was that the number of people voting doesn't really make much of a difference as long as it's above a certain threshold.  That's because when the number voting is high enough, randomness takes care of the ignorance of some people, the passion of others.  I now believe that if everyone voted, the outcome would be similar to what it is now.  The extension of this thought is that if one party was successfully able to increase turnout among its adherents, the opposing party would adapt (maybe they'd become more passionate/worried or the party would increase voter-drive efforts), and the final result would return to equilibrium.

Therefore, there's no need to increase turnout to 90% like Orszag wants.