Paul Graham makes a strong argument that differences in wealth are not inherently bad (indeed, they have positive effects on society as a whole). I have always held that a society must have people at the wealthy end of the bell curve in order to shift the whole curve toward greater prosperity.

However, wealth increases in fits and starts, not monotonically, and modern wealth creation depends on long value chains. Graham does miss one simple fact: At some point, the differences in wealth become so extreme that the fits and starts of the wealthy, amplified along value chains, have devastating effects on large numbers of the poor. The actions of the wealthy are not necessarily evil, but even the well-intentioned ones can have unpredictable negative effects. At this point, the poor are motivated to deprive the wealthy of this leverage by any means.

If society scrutinizes actions with an intensity that is proportional to the concentration of wealth, it may prevent many negative effects. This justifies government regulation. But many negative effects are inherently unpredictable.

This is why I have come to believe (contrary to Ayn Rand) that the rich have a philanthropic obligation. Philanthropy may not prevent the negative effects, but it will compensate for them. In the arithmetic of the aggrieved mob, that may be sufficient to prevent revolution and allow differences in wealth to be preserved.

As a concrete example, when the mob occupied Wall Street, the CEO of Goldman Sachs would have done well to hand out box lunches to the hungry occupants.

I know this idea seems laughable, outrageous, but why? The biggest expense to him would have been delay of various meetings, but the PR return to him, to Goldman Sachs, and to capitalism overall would have been huge. This one act would have defused the occupation instantly, and it would not have spread around the world.