Recent budget disputes here in Minnesota furnish illustrations of how even a simple statistic can be misused in ways that mislead rather than enlighten. So does a recent op-ed about youth unemployment.
Start with our state’s fiscal problems. Many writers have juxtaposed the approximate 4 percent spending increase embodied in the recent budget deal with average increases of some 18 percent every two years over the half century from 1960 to 2010. This is a variation of the assertion that biennial budget increases averaged 21 percent from 1960 until Tim Pawlenty became governor in 2003 and only 3.5 percent after that.
All of this is true. However it also is misleading in that many understand it to mean growth rates were uniformly high over the period cited. This was not the case.
The high rates of growth are skewed by very rapid growth in outlays in the 1960s and 1970s, as state government took over education funding and other functions long borne by counties, cities and school districts.
From 1961 through 1977, state outlays increased by 32 percent every two years. But from 1977 through 2011, the biennial increase was only 11.5 percent.
That still sounds like a lot, but it ignores three important factors. First, prices rose substantially over that time, especially when the Federal Reserve let inflation get out of control in the 1970s. Second, the state’s population has increased by some 2 million people over the past 50 years. Third, the overall economy has grown, both in terms of value of goods and services produced and in terms of income.
Adjusting for inflation over the past half century cuts the rate of increase in the general fund budget by more than half, from 17.7 percent every two years to 8.6 percent. Since 1977, the increase adjusted for inflation would come to 3.2 percent every biennium.
Now, adjust for the fact that the population grew to 5.3 million Minnesotans from 3.4 million in 1960. On a per capita basis, the increase per biennium is even lower, 6.7 percent per biennium over 50 years and 1.4 percent every two years since 1977. (In fairness to Pawlenty, spending was higher in the quarter-century before he took office. Between 1977 and 2003, when he became governor, the general fund budget increased 2.6 percent on a per capita inflation-adjusted basis every two years.)
Making these calculations gives a sharply different perspective than the “18 percent per biennium for a half-century” take on things. The 4 percent increase embodied in the recent agreement is less than in preceding years. But the increases over the preceding 30 years or more are not all that high once one adjusts for the two factors of inflation and population growth.
Some people who don’t stop to understand the numbers fret about how we will ever be able to pay for state government if it keeps growing the way it is. The answer, of course, is that as long as the general economy grows faster, it is not hard to pay for a larger government. That is why the total cost of local and state government combined is smaller relative to the state economy than it was 15 years ago, not larger. And yes, the process can go on indefinitely as long as outlays don’t grow faster than output or income.
A second example of the misuse of percentages appeared in an op-ed by pundit Amity Shlaes that appeared in this newspaper some weeks ago. Arguing that the minimum wage is responsible for high unemployment among young people, she criticizes congressional Democrats for raising the federal minimum wage 41 percent from 2006 to 2009.
Her calculations are correct. The minimum actually had been $5.15 until May 2007. And Democrats did make up the majority of those voting to increase it in three steps to $7.25 by July 2009. So there was a 41 percent increase in just over two years.
However, it would be just as correct to say the wage went up 41 percent over 12 years. That is because the $5.15 level was set in 1997. Spread over 12 years, that 41 percent increase works out to a raise of 2.9 percent per year.
Adjusted for inflation, the increase was just 5.1 percent over 12 years rather than the 41 percent unadjusted figure Shlaes cites. And even though inflation has been very modest since the 2009 bump to $7.25, that amount now has no more buying power than $5.15 had in 1997. Adjusted for inflation, the minimum wage’s buying power is lower than at any point from March 1956 to August 1984.
I heard a third example of abuse of percentages when an environmentalist asserted that farmlands in southern Minnesota had lost some percent of topsoil to erosion since cultivation started in the 1870s. Her facts were correct, but largely irrelevant for current policy. An amount over 140 years says next to nothing about what is going on now. There were very high rates of erosion from the first plowing of the prairies through the 1930s. But changes in farmer attitudes and farming techniques cut erosion rates dramatically, even though off-farm environmental damage from erosion remains an important problem.
And that illustrates an important point. Misuse of statistics can undermine a legitimate argument. The environmentalist was correct about the basic issue of agricultural erosion harming ecosystems. Shlaes was correct that the minimum wage can increase unemployment, especially for low-skilled youth, although few labor economists would make the sweeping argument she did. But supporting such basically sound positions with misleading use of some number is more likely to make knowledgeable readers shy away from one’s argument than agree with it. And it certainly does not raise the level of public understanding.
© 2011 Edward Lotterman
Chanarambie Consulting, Inc.