In their book Nudge, Cass Sunstein and Richard Thaler correctly point out that it is irrational for people to overwithhold taxes from their paychecks each pay period. One reason we do this is to ensure that we get a nice big tax refund when we file our taxes the following year.
Now this is silly since the tax refund represents money that you overpaid in taxes – it is your money. Thus you are making an interest free loan to the government over the prior year until you claim your refund. Why would you do this when you could decrease the amount of withholding from each paycheck, and simply put the extra amount into a savings account or other financial product and earn a higher return on it? You could even save this extra money in a tax-preferred way by directing it right into your retirement account.
Yet we do not do it. It seems more irrational when you realize that a dollar today is worth more than a dollar tomorrow. In other words, even if I plan on using my tax refund for a vacation in the summer, if I had the money a year earlier I could still take the vacation next summer and now I also have the option of doing something with the money for an entire year before it. More choices means more value.
So, this method of forced savings appears to be irrational and the authors suggest that some nudging is in order to improve the decisions that people make regarding their withholding. I don’t totally disagree. But the reason I would favor a nudge here is not because we are making bad decisions, but rather because of another behavioral effect. If a majority of taxpayers overwithhold, then a majority of taxpayers are fooling themselves into thinking they get a “gift” from the government every April. In other words, overwithholding would seem to me to be masking the effects of big government.
The behavioral literature suggests that individual overweight the impact of losses on their utility and underweight the impact of gains. More simply, we are more upset by a $100 loss than we are happy from a $100 gain. Experiments seem to indicate that we value gains at about half what we value losses. So, if we win a $200 lottery and at the same time lose five $20 bills, the net impact on utility appears to be zero (let’s just agree with this for now, I don’t necessarily buy it).
But if that is the case, then think of what overwitholding does. Every month you get paid, you lose a small amount of money (say you overwithold $200 per month) and then a year later you get a $2400 windfall. Since we weight gains only half as much as losses, we may still come out ahead here in utility terms. If however we decide to accurately withhold, so that when tax time comes we neither pay nor receive a thing – what is the behavioral effect of this? Each month, we will be realizing $200 more in income than the previous year. But these gains are not given much weight. At the end of the year, the “loss” of the $2,400 windfall would be given substantially more weight, making it entirely possible (according to this behavioral view) that making better financial decisions here will leave you worse off in utility terms.
So I’d perhaps favor a nudge here – it might be people to become more aware of the large tax and spending agenda of their government. Of course, an even better way to do it, without altering the tax code one bit, is to simply end all withholding. Instead of funds being automatically deducted from your check each month – some going to federal and state income taxes, lots going to payroll taxes – you would get the gross amount of pay, and then each month you would pay your tax bill to each parasite with a claim on you, much like you do with the trash company, phone company, etc.
Again, the theoretical economic effect of this is zero – it just raises the costs of tax compliance a little. But in a world of internet banking the incremental costs are not as large as they used to be. However, let’s play in the land of the behavioralists.
Since people weigh gains and losses differently, a dollar is not always a dollar. For example, if your current take home pay is $5,000 and that comes in a one-shot net paycheck, you will get a different level of utility from that outcome than you would for an outcome where you earn $8,000 per month gross, and then fork over a check for $3,000 to various government agencies.
In the classical economic world, the well being of consumers ought to be the same in each case – $5,000 in your pocket is $5,000 in your pocket regardless of how we got there. However, in behavioral world, the latter situation leaves you worse off than the former. This is because the $8,000 gain is weighed by roughly half its value, and the $3,000 loss is weighed on a dollar for dollar basis. In this formulation, the latter situation really only yields $1,000 of net pleasure while the former yields $2,500 of net pleasure.
The upshot of this? Behavioral economics suggests that we should be nudged in the direction of less withholding over the course of the year, but once we do that, we should be made happier by having funds directly pulled from our checks rather than paying it out ourselves.There is no necessary tension between these two.
The thing I am trying to get at is that my belief that many people’s “liking” of big government or lack of concern about government spending is simply the result of a behavioral trick. And if that is the case, then people aren’t “really” as happy as they claim to be. If the nudgers were consistent, then they ought figure out ways to nudge people to get the benefits and costs of their relationship with government right – because in that case I am almost sure we would see smaller and less intrusive government. Of course, who would be empowered to implement these nudges? Ah, the people who stand to lose from the nudge. If the role of government is to work in the best interests of the people, then we would expect this to be no problem. But if the role of government is to work in the best interests of itself, its special friends and the intellectual elite … well, then, you know how likely it is that this kind of a nudge will be forthcoming any time soon.