The Federal Reserve quarterly Flow of Funds report shows that Household sector net worth (value of assets less liabilities) reached nearly $100 trillion in 2017, more than five times GDP. Economic theory has a difficult time in explaining why households have accumulated this much wealth. According to the Life Cycle Consumption Model, people save during their high income years so as to build up a financial reserve that supports consumption spending in retirement. This model predicts that the profile of wealth against age for a typical household will be an inverted “U” with wealth peaking around retirement age and then declining toward zero. Economist Edward Wolff has conducted intensive analysis of time profiles of wealth for individual families and he reports in his recent book, A Century of Wealth in America, that the Life Cycle Model (LCM) only holds true for a subset of households. In particular, the LCM seems to hold for the third and fourth quintiles of wealth holders (the top or fifth quintile represents the richest 20% of households, the fourth quintile the next 20%, and so on). The model fails for the bottom two quintiles simply because these households have insufficient income to save much of anything. The model also fails for the top quintile. The top 20% of wealth holders own more than 80% of total wealth, with average net worth about $4 million per household. Most of these people are going to leave an estate upon death. According to the LCM, they have saved too much.
In order to explain saving behavior for the top quintile, we need to include other incentives for saving aside from deferred consumption. These incentives might include a bequest motive, a desire for prestige or power, or recognition of great uncertainty about future income, investment return and longevity. Due to this uncertainty, risk averse people might assume “worst case” scenarios and thereby be led to excessive “precautionary” savings. Naturally, precautionary saving can occur at any wealth level, but it is likely to be significant only for the most affluent.
The current Nobel Prize winning economist is Chicago professor Richard Thaler. Professor Thaler is one of the principal founders of behavioral economics. Behavioral economics asserts that human beings do not act like the economic man depicted in traditional economic theory. Instead, borrowing from psychology, Thaler has conducted numerous experiments that demonstrate how humans routinely make simple mistakes or suffer from cognitive biases. One of Professor Thaler’s main applications has been evaluation of savings behavior.
Possibly thinking only of the bottom three or four quintiles, Thaler is convinced that people in their high income years are not saving enough to support themselves in retirement. In one of his well-known popular books Nudge Thaler and co-author Cass Sunstein reveal clever “tricks” they have come up with to induce people to stop making this “mistake” of saving too little. One such trick is to alter the signup for company 401K plans from “opt-in” where the employee chooses to join the plan, to “opt-out” where the employee is automatically enrolled in the savings plan. Apparently, under the opt-out strategy the participation rates are much higher. A second idea is “save more tomorrow” which means to automate increases in the savings rate for any increase in wage. And a third is to increase tax withholding rates so that tax refunds are greater, and may be used to bolster 401K contributions.
In Thaler’s model, people act as if their decision making was driven by two competing forces – the Planner and the Doer. The Planner is a rational, future-oriented calculating machine (an “Econ” in Thaler’s jargon). The Doer (a “Human”) is present oriented, reactive, impetuous. In order to make “smart” decisions, the Planner must plan and somehow contrive to keep the Doer from disrupting the plan. In the case of retirement savings, Thaler assumes that the Planner “knows” that he should be saving more, but cannot get the Doer to comply.
But is Thaler correct in assuming that the majority of people are saving too little? There is considerable controversy among economists regarding this question. Some, like economics professor Laurence Kotlikoff, argue that most people, not just the top quintile, are saving too much. He bases this conclusion on a detailed financial planning model that he has co-authored. This model includes great detail on Social Security rules, private pension plans and patterns of household consumption. According to Kotlikoff’s model, the typical retiree requires less than 70% of income earned during working years in order to preserve a similar standard of living in retirement. Through the combination of Social Security, private pensions and savings (often in the form of home equity), Kotlikoff finds that the majority of seniors are doing just fine. In fact, in many cases they could have spent more during their working years without creating a debacle in retirement. I’m sure there is something to this argument, particularly if in retirement you move from a high cost area like Southern California to a beautiful but much lower cost area in rural Tennessee, like an esteemed colleague did a few years ago.
Thaler’s Planner disagrees with Kotlikoff’s Planner. What I find striking about the optimal consumption problem, perhaps the most fundamental problem in financial planning, is that it is really complicated. Probably too complicated for anyone’s Planner. This is because we are dealing with the multiple sources of uncertainty mentioned above – including future income, mortality and rate of return – and a confusing array of complex financial products that may be used to transfer consumption from one period to another. Kotlikoff’s solution is to take on all the complexity and uncertainty, make a host of “reasonable” assumptions, and grind through to an optimal conclusion.
The approach I have taken in thinking about this problem is to seek simple rules that seem to work fairly well. Instead of an optimization model like Kotlikoff’s, I have built a simulation model that estimates the probability of failure (defined as running out of money) given a spending rule and conservative assumptions about future income, investment return and longevity. Oh wait, the problem is getting complex. Anyway, based on my conservative assumptions (modest future income growth, low future average investment returns, long life expectancy, and positive utility accorded to leaving a family bequest), I come up with recommended savings rates that are a lot higher than what we observe today.
Thus, I tend to agree with the Nobel Prize winner. The bottom line is that every person has to engage a Planner to calculate a reasonable spending plan. This Planner may be a complex optimization model like Kotlikoff’s, a simpler simulation model like mine, a professional advisor (CFP), a trusted family member or friend, or a deep dive into financial education. The best answer for each individual will depend on personal characteristics such as risk tolerance, time preference (willingness to forego consumption today for additional consumption in the future), bequest motives and other factors. If the result of that planning is that an increase in savings is called for, then Thaler’s ideas about getting the Doer on board may be useful.