In graduate school at Northwestern University I got lucky with an idea early on and "milked" it throughout my years there, writing on the idea every time I was required to do a paper in a class (econometrics, macroeconomics, monetary economics, etc.).  The idea is quite simple really.  Kenneth Arrow (who has won the Nobel Prize) believes in something called "the increasing relative risk aversion" (IRRA) hypothesis.  The notion is that "safety," freedom from risk, is a superior good in the economic sense.  In portfolio terms, this hypothesis suggests that richer people would be expected to hold lower risk portfolios other things equal, portfolios with a higher share of safe assets.
    Yet another way of thinking about this is that, under the IRRA hypothesis, you would be less willing to take a bet if both it and your wealth were doubled.  Both implications seemed counter-intuitive to me (stocks are quite risky, yet held predominantly by rich people, while doubling a $5 bet and, say, $50,000 wealth to $10 and $100,000 would, I suspect, cause most people to be more prone to take the bet).  Arrow supported his position with the time series money demand work of Milton Friedman (who has also won the Nobel Prize).  In that work, Friedman found that, as U.S. incomes (a proxy for wealth in Arrow's context) grew over a hundred year period, money holdings as a percentage of income increased.
    I felt that many other things (urbanization, changes in family size, etc.) were changing over such a long time period and that income was mistakenly getting credit for the influences of these other (omitted) variables.  I should point out that both Friedman and Arrow are deservedly famous for a great many contributions apart from the issue under discussion here.  Undaunted by the fact that I was disagreeing with two Nobel Prize winners, I managed to get the following articles published.  Note, ironically, that the recent paper, which employs the most fancy econometric techniques, appears to suggest I am wrong and they are right!  (But, somewhere down deep, I still think I'm right, and there are a number of economists who agree with me).  Such is the nature of scientific advance....  The last article referred to below is much more recent and involves growth and development, with implications for appropriate measurement of that.

ARTICLES:

1. "Wealth and the Cash Asset Proportion." Journal of Money, Credit and Banking, Vol. VIII, No. 4 (November 1976), pp. 487-496.

2. "New Evidence on Income and the Velocity of Money." Economic Inquiry, Vol. XVI, No. 1 (January 1978), pp. 53-68.

3. "Relative Risk Aversion: Increasing or Decreasing?" Journal of Financial and Quantitative Analysis, Vol. XIV, No. 2 (June 1979), pp. 205-214.

4. "The Velocity of Money: Evidence for the U.K. 1911-1966." Economic Inquiry, Vol. XVIII, No. 4 (October 1980), pp. 631-639.

5. "Domestic and International Policy Implications of the Inflation-Unemployment Trade-off." Akron Business and Economic Review, Vol. 5, No. 2 (Summer, 1974), pp. 9-14 (with T. Sandler).

6. "The Income Elasticity of Money Demand in Less Developed Open Economies." Journal of the Southwestern Society of Economists, Vol. 15, No. 1 (1988), pp. 165-71 (with J. Marchand).

7. "Specification of U.S. Money Demand: Sociodemographic Variables Revisited," Applied Economics.Vol 34, No. 5 (March, 2002), pp. 659 - 665  (with J. A. Holman).

8. "Economic Growth and Business Cycles: The Labor Supply Decision with Two Types of Technological Progress" (manuscript under review).