Behavioral Economics and Forced Savings

Some curious implications

I have a kooky idea.

Let’s start by writing down a production function in the style of Solow, where output Y is a function of capital and labor, K and L. The exponents sum to one, so there are constant returns to scale, but we will say that A (which is technological progress and multiplies labor L) is a function of the amount saved for capital. Capital is, incidentally, the portion saved out of consumption, and let’s also suppose there are decreasing returns to ideas found as a function of capital.

This seems like a not unreasonable description of at least a part of technological discovery. We try out new, capital intensive ideas, and shake out discoveries at random without really trying to find an idea per se. Some of this saving is explicitly for research and development, but we’d actually be happy with any sort of deferred consumption.

This also implies that we save too little. The ideas which are discovered spill over onto everyone, so we just save whatever is personally optimal and not what is socially optimal. There is a role for forced savings increasing output.

Suppose that there is a distribution of people by propensity to consume versus save out of income. These propensities to consume are not driven by liquidity constraints or income, but are instead innate characteristics of the people. This, too, is a realistic feature – there are some people who just spend windfalls, and others who don’t, and it’s not due to anything other than who they are as people. Maybe this is for as simple a reason as people being inconsistent over time? We think that people who engage in the consumption of drugs, for example, are behaving in a way that does not maximize their welfare. Doesn’t seem unreasonable to think that the same stuff implies that we will save too little.

How might we increase consumption? Well, we could have the government tax people and save more. That is one possibility. Another possibility is redistribution from high to low marginal propensity to consume (MPC) people. This could happen through the government, but needn’t. We could also have redistribution because low MPC people sell goods that high MPC buy. We can ensure that these goods are sold only by low MPC types to high MPC types because they appeal to the same time inconsistency which causes both drug consumption and insufficient saving. This implies that markups on those goods are good – the ideal good for redistributive purposes would be one which is produced at 0 marginal cost, and sold at a substantial markup.

What are goods that fit these conditions? Perhaps opium derivatives. Perhaps gambling. In short, all of the vices will tend to increase savings. My dad used to remark that the lottery had enormous margins, and was essentially a tax on the poor to give to the middle class. If you believe that we do not save enough – and that seems very plausible in two different ways – then you may have to consider this good.

How true is this? I don’t know. But it seems worth thinking about.