3 ms·
> The issue here is not distribution of wealth, it's consumption. The more we consume now, the less we invest in the future. Do you disagree that BI will increa
by mrow84 12y ago
> The issue here is not distribution of wealth, it's consumption. The more we consume now, the less we invest in the future. Do you disagree that BI will increase consumption?
The point of BI is to redistribute consumption to the lower end of the economic distribution, at the expense of the higher end, not to increase overall consumption, though that might happen as a side effect.
The net increase in consumption minus the net productivity gains will be the cost of BI, paid to increase economic equality. However, I have seen quite a few (in fact I would say most) proponents of BI argue that the "cost" will be negative, that the productivity gains will, overall, exceed the increase in consumption.
A basis for this claim is that the marginal return on consumption for poorer people is higher than for richer people, and so they can more efficiently use the redistributed consumption, leading to higher overall productivity. The issue of consumption is, therefore, only half of the argument, and must always be considered in tandem with productivity.
The answer to your final question in this context is that we should favour BI because it is intended to increase net productivity at the same time as increasing consumption on the low end: we will have our cake and eat it.
- infinite8s 12y agoI believe the gp is saying that BI won't cause a redistribution of consumption from the high-end to the low-end, but instead of redistribution of investment on the high-end (since consumption doesn't scale linearly with income - probably something logrithmic looking), and this loss of investment will have a much greater negative effect on productivity than any minor increase due to consumption on the low end.
- mrow84 12y agoYes, I think at this point it's really just a difference of opinion as to what the effects of BI would be. My view is that the marginal utility of those dollars spent by the high end on investments is lower than the marginal utility of increased consumption at the low end, especially as the actual amounts involved are not that big. I guess the best way to resolve the issue would be experimental validation! ;-)