Does Complexity Have Regressive Social Effects?

I have written about the virtues of simplicity in business before. But when I look around things are getting more and more complex.

Take the tax code for instance. The US tax code is so complex it is almost impossible to deal with for the average tax payer. An example of this is the Flexible Spending Account.

The FSA is supposed to give the tax payer a tax deduction on healthcare expenses that are not covered by their health insurance. It includes deductibles, co-pays and some kinds of health care expenses that are typically not covered by health insurance.

So far there’s not much you can object to. But the implementation is where it gets tricky. At the beginning of the year the employee must elect the amount to be deducted from their pay towards funding their FSA account. They can’t spend more than what is pre-funded in their accounts. If they spend less, they lose the money!

To me this seems hare-brained. If you are expenses go way over say because of some dental surgery expenses where the copays are high, you are out of luck if you didn’t foresee this at the beginning of the year. If your expenses are well under at the end of the year, you either lose the money or in the last month go hog wild buying OTC drugs that you don’t need. Who wants to see their own money go waste?!

To finesse a benefit like the FSA and maximize your gain from it, you have to be well informed, alert and always on top of your spending data. Also, you have to be quite motivated to deal with your benefits provider on a regular basis.

There are many such examples where you have to make active decisions in order to save taxes (IRAs, 529 plans etc.). This is true not just about the tax code but almost everything in American life – from health insurance to frequent flier programs. The complexity of these things – the tax code, health insurance contracts, frequent flier programs – make it difficult for people to make the decisions necessary to get the benefit they were designed to give. Sometimes understanding how the benefit works is difficult. Sometimes taking the action requires too much time and effort. Sometimes its both.

This phenomenon is generally well understood by business and law makers. Take mail-in rebates. Businesses expect a certain redemption rate on mail-in rebates. I think it is something like an x% rebate will result in an x% redemption rate. Which means that if you offer a 10% mail-in rebate you will get a 10% redemption. I can think of other things that should matter, but the point is that the effort of mailing in the rebate form results in a large percentage of buyers not redeeming what is essentially free money.

Similarly, 401(K) plans with opt-in results in about 25% of employee not participating in the plan in spite of an employer contribution. In this case it is the lack of information or guidance that becomes the hurdle, not so much the effort to sign up. Opt-out plans are now becoming more common which on average results in over 90% participation rates.

The big problem here is that the people who tend to miss out on these benefits are the people who could use the benefits the most – people who typically don’t have a college degree and are working too many hours to spend the rest of them on the phone with their benefits provider. They can’t afford professional advice and their friends are all in the same boat.

There are other costs of complexity as well. Complexity hinders change. Healthcare reform will take much longer than it would have if people understood it better. Complexity can sometimes be introduced deliberately to take advantage of the consumer (credit card fine print, for instance). Complexity is expensive for the whole system. If the tax code was a simple progressive tax schedule with a deduction for family size and that’s all, we wouldn’t need tax accountants, most of the IRS and H&R Block and Turbo Tax.

But the biggest reason to simplify is that the burden of complexity in our tax code, health care and consumer finance is borne by lower income households. It is regressive.