Bailey Voter Turnout Column Correction

|

Mea culpa. In my most recent column, I misinterpreted data about voter turnout for Presidential elections since 1828, which I claimed showed that US voter turnout had increased dramatically as the Federal government grew is size and scope. I had interpreted the data as being the percentage of eligible voters when it really is data showing turnout as a percentage of voting age population.

My erroneous assumption was that a turnout count would necessarily have to consider only voters who would have been eligible at the time, e.g., white, property-owning males over age 21. Instead the dataset I cited somehow makes a retrospective estimate of the percentage of possible voters based solely on their age, ignoring eligibility requirements of the time.

It turns out, according to Professor Michael Alvarez at Caltech that, ?the data are quite different when you attempt to factor in eligibility into the historical data. The qualitative picture that emerges is one of very low turnout through the late 1980's, high turnout in the 1830's through the mid 1880's, and then moderate turnout thereafter.? Of course, the big jump in percentage turnout related to population occurred after women got the vote in 1920.

However, I do note that there is still some evidence to support my assertion that in recent years voter turnout declines as the proportion of GDP the government takes declines. In 1960, voter turnout was just over 60% and fell to just under 50% by 1996. Using data from the Bureau of Economic Affairs, one finds that government expenditures as percentage of GDP from 1960 to 2000 are as follows:

1960—25%
1964—27%
1968—27%
1972—23%
1976—22%
1980—21%
1984—20%
1988—20%
1992—20%
1996—18%
2000—17%

Finally, my misinterpretation of voter turnout data does nothing to undermine the Nebraska study cited in the column showing that Americans are not eager to participate in politics. My apologies for the error and my thanks to the readers who set me straight.

NEXT: Jury Rigging

Editor's Note: We invite comments and request that they be civil and on-topic. We do not moderate or assume any responsibility for comments, which are owned by the readers who post them. Comments do not represent the views of Reason.com or Reason Foundation. We reserve the right to delete any comment for any reason at any time. Report abuses.

  1. You know, it also looks like voter turnout declined along the same curve as marijuana usage declined over the years.

    Toke up and vote!

  2. It’s interesting to note that the figures show a smaller government in relation to GDP.

  3. How did you produce those numbers? Downloading the GDP CSV and dividing government consumption / GDP gives:

    1960: 21.5%
    1964: 21.8%
    1968: 23.3%
    1972: 21.8%
    1976: 21%
    1980: 20.4%
    1984: 20.4%
    1988: 20.3%
    1992: 20.1%
    1996: 18.2%
    2000: 17.8%

    Not as big a shift as your numbers. Oddly, OMB data gives entirely different numbers: check out table 15.3.

    1960: 24%
    1964: 25.3%
    1968: 27.9%
    1972: 28.2%
    1976: 30.6%
    1980: 30%
    1984: 30.8%
    1988: 30.4%
    1992: 32.2%
    1996: 29.9%
    2000: 27.9%

    No idea where the conflict is coming from, or which is correct.

  4. Oh, I forgot to mention: I suspect there’s something we’re missing in the BEA data, because the government was in no way only 18% of GDP in 2001.

  5. Oh, it’s because the BEA doesn’t include transfer payments (Social Security, etc.) in the government expenditures line. Table 3.1 shows this; the amounts they give there, which include transfer payments, match the OMB data.

Please to post comments

Comments are closed.