Technology v Poverty, and Keynesian Multiplier Problems

This article makes note of two other recent articles and my comments regarding them.

To Cowen’s recent article, How Technology Might Someday Fight Income Inequality, I commented on the page as follows:

I think every good discussion on the way in which technological progress reduces income inequality must mention a few items:

1) Real income equality, not nominal income equality, is important. The economics of abundance, also known as post-scarcity economics. Traditional economics deals with scarcity, but technology generally makes things arbitrarily cheap in real terms as time goes on. In the very long run nothing will have a significant real cost. When everything is free, what does it matter if you have 1 dollar and I have 2?

2) Real quality of life equality, not nominal income equality, is important. Leisure time must be considered in an honest discussion of quality of life because some people prefer to earn less and rest more. In addition, if income has diminishing marginal utility then we can have large discrepancies in income with insignificant discrepancies in quality of life, provided that all incomes are sufficiently large, and it is the case that technology improves real income, or purchasing power, over time.

3) Returning a bit toward the post-scarcity idea, the concept of Open Source and the general low marginal cost associated with information technology is important. Knowledge will have an insignificant real cost in the future. It will basically be free. Free information doesn’t solve all of our problems but it is a game changer. Baseline income, by which I mean the real income of a person making a nominal income of 0, will eventually be greater than average income is today. Absolute poverty will cease to exist, and those living in relative poverty will largely be people who chose to do so because they prefer leisure to additional income.

Don Boudreaux had a neat article on a separate topic. The Science, or ‘Science,’ of the Spending Multiplier introduces, a website dedicated to showing that the Keynesian Multiplier has not been validated using Out-of-Sample tests, creating an important validity problem as explained here. The website includes an exhaustive list of over 120 papers analyzed and welcomes wagers in favor of the spending multiplier. I might be taking up a wager there, we’ll see.

Thought it is currently awaiting approval, I left a comment on the website as follows:

My first thought is that I buy the argument for the government spending multiplier. If the average savings rate is .5 then a sufficiently large spending package should result in overall income in the amount double the amount spent.

Obviously there are some complications. Whether or not banks choose to loan funds is important because lending depends on the existence of profitable loans to fund. To say that the average savings rate is .5 importantly ignores the variation and distribution hidden behind that average. Real-world spending packages are often constructed in ways which disturb the application of theory. There is a probability effect which ensures that the theoretical expectation will deviate probabilistically from reality.

Considering all of this, I still estimate the multiplier to be larger than 1, although I agree that the literature supporting this idea is weak, and I can’t think of any OS tests.

My final point would be that the multiplier hardly matters out of context. There is also a private sector multiplier with which the government spending multiplier must be compared. My suspicion is that they are insignificantly different in theory, and in the real world the private sector multiplier may be better due to malformed spending policies. With the government spending multiplier charitably equal to the private multiplier, there is already know[*] reason to engage in transfers of wealth. Importantly, the transfers themselves incur additional cost. Taxation costs the amount taxed multiplied by the foregone private multiplier. Print-funding costs an inflation tax, also subject to the multiplier. Borrowing has costs and so on.

It is clear to me that even with a spending multiplier well above 1 there is no justification of spending.

* – I meant to say “no,” but the quote is reproduced exactly as it was unfortunately submitted.


Leave a Comment