Why Do So Many Households Hold So Few Interest-Bearing Assets?
I'm not sure this post will inspire as many comments as the post below it, but increasing national saving is an important issue both generally and in the Social Security reform debate, so I thought I'd present some empirical evidence on how the accumulation of financial assets by households changes as households become more informed about financial markets. A paper by Casey Mulligan (University of Chicago) and Xavier Sala-i-Martin (Columbia University) appearing in the Journal of Political Economy, Vol. 108, No. 5. (Oct., 2000), pp. 961-991 (JSTOR stable URL – subsc.) reports that according to the Survey of Consumer Finances, 59% of U.S. households hold no interest-bearing financial assets over and above employer held pension funds and IRAs. Why do so many households hold so few assets? The authors argue that the transactions and learning costs of entering financial markets are sufficiently high so as to more than offset the expected earnings for most households. They also suggest that people with retirement fund assets such as employer held pensions and IRAs have lower transactions and learning costs because their exposure to retirement assets may bring additional understanding of how such markets function. They find that “(a) the elasticity of money demand is very small when interest rate is small, (b) the probability that any individual holds any amount of interest-bearing assets is positively related to the level of financial assets, and (c) the cost of adopting financial technologies is negatively related to participation in a pension program.”
I want to focus a bit more on (c) which tells us that participation in a pension program increases the likelihood of holding financial assets at all income levels. The Social Security debate is often centered around the idea of an ownership society for lower and middle income class workers so the focus will be on households with low amounts of financial wealth. A new econometrics textbook that will be out this fall uses data from the study to investigate this issue. The example in the book asks “How likely is an individual with $1,000 in total assets to hold any of it as interest-bearing assets if he or she has no retirement accounts?." At an asset level of $1,000 the probability (from probit estimates) that an individual will hold any of the $1,000 in interest bearing assets is 12%. However, when an individual already has a pension plan of some type, the probability of holding additional financial assets rises to 18%. Also, note that these percentages pertain to a particular asset level, $1,000, and according to result (b) the percentages increase as the asset level increases.
This is evidence that one of the barriers to entering financial asset markets is the cost of learning how they operate. This may also explain why discussions of add-on accounts have noted much higher participation rates with opt-out as opposed to opt-in programs. There is a much larger incentive to learn what you need to know to protect the principal or liquidate the assets than there is to put the assets into a retirement account. That is, if the investments are automatic or if particular funds, etc. must be chosen there is an incentive to make sure the principal is protected and to learn the rules under which the principal can be drawn down if needed. In the process needed knowledge is obtained. When the system is opt-in, the expected return is not sufficient to trigger the learning needed as a prerequisite to participation in financial markets.
Let me be clear. I believe the solvency issue has been oversold and we do not need to radically alter the Social Security program. This is in no way a call for private accounts to solve some imagined hyped-up problem. But decreasing the barriers to participation by lower and middle income households in financial markets is an important goal and this tells us something important about how to do that. As I watch colleagues fret over the very few retirement options available to them (me too), and these are Ph.D. economists, and as intelligent friends ask questions about annuities (like what the heck are they?), etc., it seems to me that these barriers are substantial.
I do not know if it is lack of knowledge of the types of financial assets, their risk-return characteristics, knowing where to go to purchase assets at the lowest fee, and so on that constitutes the biggest barrier to participation. But the change in participation rates from 12% to 18% in the numbers above at relatively low asset levels from simply having a pension account no matter how passive the participation suggests there are potential gains to be made through better education. Less than full information among participants is a known market failure, especially when information is asymmetric (why do annuities come to mind again?). My casual observation, and more to the point empirical results, suggest lack of information is a substantial problem. If we can identify and overcome areas where lack of knowledge is a barrier to participation, perhaps we can increase participation in financial markets at all income levels, particularly among low to middle class households.
What is the most important informational barrier? Is it as simple as knowing where to go to buy an asset like a T-Bill, corporate bond, or index fund? Or is it a lot more than that?
Posted by Mark Thoma on Friday, June 24, 2005 at 02:34 AM in Academic Papers, Economics, Macroeconomics, Reading, Saving, Social Security |
Permalink
TrackBack (0)
Comments (24)
You can follow this conversation by subscribing to the comment feed for this post.