The public policy blog of the American Enterprise Institute

Subscribe to the blog

Discussion: (6 comments)

  1. Here’s what you got – a non-clickable link to a BLS
    website that is said to be what the author used in his calculations.

    this is an old AEIkl tried and true technique of synthesizing data and being less than clear as to their source or their calculations..

    where does the BLS link take you for Chart 1?

    well here:

    try it yourself and tell me the data that the author used,

    this is what AEI is all about now days … dishonest propaganda… misinformation, disinformation.

    If you want to make a claim, you have to provide the data and calculations you used or else what you’ve down is simply not credible nor legitimate.

  2. .

  3. Andrew Biggs

    Larry, It seems what you’re saying is that research is illegitimate unless a government survey both collects the data and does the tabulations for you, such that all the researcher is doing is link to calculations someone else has done. That’s not how research works (although we do cite a BLS paper on teacher work hours that matches the figures we produce). We point to the data we use and give sufficient detail that another researcher could reconstruct what we do. And when people have contacted us looking for more details on how we do our calculations, what we assume, etc. we prove it. That’s the standard research process.

    1. @Andrew – no. the standard is how Wiki does it.

      you do two things:

      you reference the original material by the specific page and if the data is not directly observable and is calculated, you provide the methodology so that the person can re-create it / verify it.

      If you are not willing to do that – then you’re not really providing a straight-forward ability for the reading to understand how you did the data.

      I followed the link by manually recreating it since it was not clickable… but the link took me to a page with several different sets of data – none of which matched the August data that was cited.

      Why not give a direct link to the exact page that you used and/or how the data was calculated/synthesized.

      I think the AFSME comment was on target… as to the verifiability… I was not able to re-create it either.

      In my mind, that’s not how you back up your assertions if you really want people to do it.

      If I somehow got this wrong, then show me.

      1. Andrew Biggs

        Larry, If you go through Jason Richwine’s paper it explains in some detail what we did: we accessed the data, which for each respondent details what they did through a given sample day; we then calculated for each person how many hours were spent working, with several different definitions of what constituted ‘work’ (e.g., including or excluding work travel, etc.). Then, we averaged work hours across different occupations and/or sectors or employment. You probably can’t recreate it unless you’ve got some familiarity with survey data and programs like Stata that are used to analyze it, but a decent economist or policy analyst who already works in that area would be able to do so. In any case, our basic results regarding teachers are consistent with an existing BLS paper, so we’re confident we’re not calculating things differently than BLS did.

        1. so we’re confident we’re not calculating things differently than BLS did.

          can you give the page numbers of the BLS data that you used (along with the link to the doc) and can you give the specific link to your calculations.

          I still think the AFSCME guy was correct. it’s not entirely clear what you have done it and what you used to do it as you have quite a number of different links – including one on Social “Insecurity”… etc.

          if you want to produce a convincing and compelling argument at least in my book you have to present a compelling and clearly articulated argument backed up by easily verifiable data and references.

Comments are closed.

Sort By:

Refine Content:


Additional Keywords:

Refine Results

or to save searches.

Refine Content