Thursday, March 17, 2011



The Triumph of Innumeracy


by Mark Powell

Vermonters for Sustainable Population

At a December 21st press conference, Census Director Robert Groves released the results of the 2010 census. News coverage across the board has echoed the soundbite that population growth had slowed signficantly, with a typical headline calling it “ the slowest since the Great Depression.” In reality, the US population in the past twenty years has grown faster than ever before, but that was not the message we heard. That was not the message we heard from the Census Bureau, however. In its press appearances and written statements, Director Groves and other Census Bureau spokepersons emphasized percentage rates of growth almost exclusively, providing real numbers only in their entirety and not at all in any graphical context. As if aware of the potentially explosive response he was hoping to avoid, Director Groves showed a slight squeamishness as he stated, “The percentage growth this last decade . . . . is thus the second lowest of the past century.”[i]

The characterization is fair and accurate only if the increase in the U.S. population is measured as a percentage of the existing population, but is irresponsibly inaccurate when considering the actual numerical increase in comparison to past decades of growth. At a minimum, the Census has reported an increase of 27.5 million people in the U.S. from the year 2000 to the year 2010, a 9.7% increase. Looking at the percentage increases of the past century, this is indeed a modest growth that is surpassed by every one of the previous five decades. Since the U.S. population at the start of this decade was more than twice the size of the 1940 population, however, the use of percentages understates the significance of this decade’s growth. Looking at the actual numbers of decennial census results, on the other hand, shows that this decade is surpassed only by two others. One of these, the previous decade from 1990 to 2000, recorded an unprecedented 32.7 million increase to the size of the U.S. population. The other comparable decade, from 1950 to 1960, the high point of the legendary baby boom, saw an increase of 27.9 million people to the U.S. population, barely enough to supersede the current decade, which, as I mentioned, was portrayed in headlines as the slowest decade of growth since the Depression. I think this is an important distinction that is clearly de-emphasized in the language and analysis that prevails in the news coverage of this preliminary report.

It is fair to ask whether the percentage increase or the increase in actual numbers is a better measure of the impact of continued U.S. population growth, but I fear that the Census Bureau, in choosing to report out on the less-controversial interpretations based on percentage increase, is de-emphasizing the significance of current immigration policy and its impact on current and future population growth. If it were not for the recent economic downturn, experts all agree, immigration would have continued at its record pace through the end of this decade. If it had done so, it would probably have exceeded even the 1990’s in the numerical increase of the U.S. population.

In the chart below, I have graphed U.S. population growth from 1900 to 2010 in two different ways. The first shows the percentage increase, which appears fairly innocuous in that it shows a modest trend of declining population growth relative to the existing size of the U.S. population. The second shows the actual numerical increase by decade, indicating in a much more definitive way the fact that U.S. population growth has rebounded considerably in the past twenty years.

US Population: Percentage Growth by Decade as Emphasized in 2010 Census Coverage

Decade Ending:

U.S. Population: Numerical Growth by Decade as Ignored by 2010 Census Coverage

Decade Ending:

If they had been interested in showing a clearer picture of the historical trends, they could have used a rolling average, as statisticians often do in situations where short term variations are potentially unrepresentative. Using a twenty-year rolling average to portray U.S. population growth over the past century would produce the following graph:

Even if you were to specifically bracket the baby boom years, comparing the growth from 1945 to 1965 rather than on the basis of distinct decades, the two decades of the baby boom saw an increase of 54 million people in our population, while the twenty years just ending saw an increase of at least 59 million. The baby boom has just been eclipsed by the 20 year period just ending, and we are poised for even more population growth as the economy rebounds. When the baby boom reached the age of 20, it called it quits and went to bed. This boom, at 20 years old, is snoozing at the moment. However, with a likely resurgence in immigration as the economy improves, and taking into account the fertility of recent immigrants, this boom is likely to wake up and drive forward with continued unsustainable growth in our population, making this not only the largest population surge in our history but also the longest running one.

This is the most widely distributed graph portraying the 2010 Census Results.


As you can see, this graph does portray the overall population size, using green bars, and that is obviously at a higher level today than ever before. They have used a scale, however, that portrays the per-decade increases on a much diminished significance, making it very difficult to discern even minor variations between decades. Had they used the bar elements of the graph to portray decennial increases rather than cumulative totals, however, the differences in numerical growth between decades would be much more emphatic. So what this graph most readily conveys is not the decade-by-decade total size as shown by the green bars but the decade to decade shifts in the percentage growth, indicated by the jagged red line, so that the percentage growth of one decade becomes the visual standard by which the succeeding decade is measured. Since the 1990s witnessed the largest numerical growth in our history, anything that follows it has to meet a higher standard of growth just to avoid an inferiority complex. As mentioned earlier, when all is said and done the Aughts will likely be recognized as the second fastest growing decade, population wise, in our history. Sure doesn’t come across in this 2010 graphic, though, does it? Now have a look at a graphic that was used in 2001 to inform the American public of the demographic earthquake recorded in the 2000 census.

Clearly, then, U.S. population growth is not slowing down when considered from a long-term, statistically meaningful way. This begs the question; does the Census Bureau have an interest in portraying our current population growth as slower than it actually is? And if so, why?