Statistical Bait and Switch for Socialism



Capitalism is all about competition, so Americans love top ten lists, especially if those lists compare states. Not only are we often nationalists but also regionalists. In other words, we like to be proud of ourselves – where we live and how we stack up compared to others. Unfortunately, we rarely take stock of how these comparisons are formulated.

For instance, the Milken Institute recently ranked each state using its “State Technology and Science Index.” Massachusetts ranked first and Mississippi last, with South Dakota ranked forty-first, Wyoming forty-third, and North Dakota jumping from forty-fifth to thirty-first place. (See the rankings here: )

Most will simply accept the given rankings (notwithstanding that technology is no more important than agriculture) and few will go so far as to research how these rankings were actually compiled. Here’s the inside scoop. The study used five subcomponents consisting of seventy-seven total areas to determine its ratings. The following are just a few of the criteria used:

– Per capita state appropriations for higher education (2003)
– Per capita state spending on student aid (2001)
– Percent change in state appropriations for higher education (2002-2003)

Take, for example, “Per capita state appropriations for higher education” and research that data. By fact, in 2007, Wyoming ranked #1, North Dakota #7, and South Dakota #30. But that was 2007. So why was data from 2003 used in this just recently (June 22, 2008) released report? (See Milken’s own data here from 2001 to 2007:

Similarly, the data used for the two other categories (above) came from 2001 and 2000-2003 while the most recently available data is from 2005-2006 and 2006-2007. So what is going on and why is this data so selectively skewed?

Foregoing any further analysis concerning the accuracy of the base data and the weight given to each by the Milken Institute, there is a broader issue here which is consistently overlooked.

State rankings are increasingly based on measurements of state aid. This increasing belief that state aid is desirable and beneficial when it comes to technology seems to be cleverly reinforcing when more outdated data is used.

For instance, without the specific knowledge that Wyoming was ranked #1 last year in “state appropriations for higher education,” the natural political instinct is to increase that spending for technology students because Wyoming is ranked forty-third overall, etc.

In other words, the public mind is conditioned to accept general data and rankings that are not necessarily applicable to specific issues. And the hidden agenda is often one of increased state and federal funding, in this case for technology based industries, which raises a state’s rankings (else why include it as a positive instead of a negative measurement throughout the five subcomponents?).

A deeper look at these kinds of news releases suggests that such rankings and polls are nothing more than clever arrows in the ever expanding quivers of socialism. It behooves us – capitalists all? – to be more diligent these days in our consumption of daily news. Not reported from this study is that South Dakota ranked twenty-sixth in “Risk capital and entrepreneurial infrastructure,” which are the real teeth of capitalism and economic growth, even as it concerns technology.

UPDATE: The statistical information referenced in this article is now more current than what was available at the time the author wrote the piece.

Anton Kaiser was born in Aberdeen, South Dakota, and retired in Rapid City after serving twenty-seven years as a U.S. Army infantry officer. He is a graduate of the United States Military Academy, West Point, and holds Masters Degrees in Business and in Public Administration from Webster College, St. Louis, MO. He is also a veteran of Vietnam, Berlin, Operation Just Cause (Panama) and an honor graduate of the Command and General Staff College, Fort Leavenworth, KS.

3 Responses to “Statistical Bait and Switch for Socialism”

  1. The Milken Institute appreciates Mr. Kaiser taking the time to examine the data underlying our just-released 2008 Technology and Science Index. However, there appears to be a misunderstanding about that data.

    All of the data used in the 2008 report is the most recently available. For example, in the category “Per capita state appropriations for higher education,” the data used in the 2008 report was from 2007, not 2003 as Mr. Kaiser states. For “Per capital state spending on student aid,” the year used was 2005-2006, not 2001. And for “Percent change in state appropriations for higher education,” we used data from 2006-2007, not 2002-2003.

    The confusion may come from the fact that we have posted on our web site ( all of the data from not only this year’s report, but from our previous two reports in 2004 and 2002 to allow a comparison for each measurement.

    None of this data, its measurement or its reporting has been “skewed.” All of the numbers are the most recent official numbers available, and all of them are there for anyone to see on our web site.

    Skip Rimer
    Director of Programs and Communications
    Milken Institute

  2. I am attempting to contact the author right now. We will make any and all corrections or clarifications after I talk with Mr. Kaiser.

    Thanks for contacting us.

  3. I checked with Mr. Kaiser this afternoon and he told me that the information available on the website had changed from the time he checked the data, that it was now more current than it had been.

    Also, as he explained it to me and as I understand his intent, when he said “skewed,” he meant not that the data itself was inaccurate, but that it could give a skewed impression that government funding is automatically a plus for technological development.

    Here’s what he sent me:

    “As a result of my article, it appears the Milken Institute has updated their reference data for this portion of their report:

    Nevertheless, other subject categories on this site display data dates from as far back as 2003, which continues to skew their data. For instance, the “Percent households with computers” data is from the year 2003. Given South Dakota’s recent push for computers in classrooms and other internet initiatives in the last few years, that data has to be terribly inaccurate a half decade later. The point is that “most recent” data should not be confused with “most accurate” data when it comes to the politics of science.

    And, certainly, the point that listing so many state and federal aid programs as positives for technology remains a valid challenge to their overall design concept.”