Applications for the ideas of elementary statistics are everywhere. Generally speaking we should all know about the basics: means, medians, averages, standard deviations, probability, general analysis of risk, and so on. Real variability is often critical to simple business or personal economic decisions. These and other assessments are important in following markets and making long term or even short term investment decisions. Do you understand the “200 day moving average” in a stock price, for example? How should you use that knowledge to a decision to add to your investment or sell what you have and walk away.
Many decisions we might make, for example to redevelop and save the Social Security Program require that we first understand the high probability that without adjustment the program will grow increasingly insolvent over the next two decades. It is not in trouble now nor will it be in the next several years. The problems come from the fact that we are living longer and that the largest generation in American history is just now beginning to retire at the rate of about 10,000 per day. So the real problem comes 10-20 years out when there are more recipients than there are younger citizens contributing to the program. Clearly this distribution has to change. Ideas to this point have included: gradually increasing the starting age for payments to begin from 65 upward to 70, reducing payments to some based on size of their other retirement funds, savings or income, and increasing payroll contributions by existing and future workers.
Americans need to make additional complex decisions about current federal government budgets which impinge on long term national debt. In 2011 for example we spent 1.1 trillion dollars more than we gained in revenue. Two simplistic and unworkable alternatives have, of course, been suggested: add 1.1 trillion in revenue, or cut 1.1 trillion in expenses. Both ideas are dead on arrival for almost obvious reasons. Whatever decisions are taken some kind year to year and then future multiple year analysis, both statistical and non-statistical, has to be made within each budget sub-category on both the revenue and spending side.
On the revenue side 64% comes from income, corporate, excise and other taxes or a total of 1.47 trillion of the 2.3 trillion in revenue. The remainder on the revenue side comes from Social Security payroll taxes, which still are in excess of the pay out. They could be increased but further excess would likely only go to pay down current or expanding debt. Thus, if the full 1.1 trillion were required to be added to non-Social Security taxes that would require an increase of about 40percent of current taxes. Such an increase would be exceedingly difficult to sell politically. However, if 300 billion dollars were the target, that would mean an approximate 20 percent increase in tax collections. If this were combined with a serious revision of the tax code, which could include reductions of subsidies to agriculture, oil and other businesses who are paid by the government to change behavior (not always in smart ways), then the 20 percent noted above could be partly collected from tax code changes and partly by subsidy changes, which occur on the expenditure side of the budget. If something like that occurred, we still need another 800 billion to come from expenditures to completely remove yearly budget debt. It seems obvious that some of the 800 billion would come from reduction of the defense budget, some from savings from reorganization of Medicare/Medicade, some from reduced discretionary spending and some from reorganizing the amount of interest that needs to be paid on debt by selling 10 year and longer federal U.S. bonds that are currently available at all time low interest rates. Collectively, from both sides of the budget, it seems hard to imagine that 1.1 trillion in savings can be achieved within one year, but some additional cost lowering can be built in and achieved in the years immediately following. Nevertheless, when attempting to balance budgets for a longer term, a sequence of year-by-year improvements are far better than just letting the budget drift toward higher and higher debt levels.
As citizens we need to be able to understand how to get back to a safe, stable, predictable center of a statistical distribution when we are caught in trends that are far away from a desired mean. Under such conditions otherwise rare events may become more probable and occur quite unpredictably. In his landmark book, “The Black Swan,” Nassim Nicholas Taleb describes the highly improbable, but unpredictable event with massive impact. They are rare, but they do occur.
We know very little about how to negotiate the possibility of an unusual outcome, and it is often impossible to even think about them in advance. Such outcomes have become rather common lately as we find ourselves in the midst of an economic downturn which is so complicated that the real risk of undesired outcomes have turned out to be far greater than they appeared going in. Clearly, in these matters we are at the edge of ordinary garden-variety statistical thinking, and a strange land where our vision of standard statistical outcomes doesn’t work so well, a place where we cannot easily and accurately assess risk. This has major implications. Even bankers cannot or should not assess risks they do not fully understand. They do risk losing significant amounts of depositor’s funds, beyond a level at which they may be insured or are insurable. Such risks minimally cannot involve all the depositor’s funds available to the bank, and should be no more than a few percent of the funds entrusted to banks by depositors.
We all need to understand what the banks are doing, and minimally limit the behavior in such a way as it does not “break the bank,” and therefore the depositors. We, but most of all the banks, need to understand the types of investments in which a real understanding of risk appears to be still out of our grasp. We may be in a place where we don’t even know what we don’t know.