Randomised control trials for government policy?

I found this chapter online by Andrew Leigh called ‘Evidence-based policy: summon the randomistas?’ It is short so you are probably better off reading it rather than reading my account of the chapter. The reason I am so keen to post it is that I think he really chose the wrong example to start with.

He starts with an account of him taking his morning vitamin with his coffee, until he reading a review of randomised control trials as to whether or not people who take vitamin supplements live longer. I agree the $30 he said he used to spend on vitamins was probably a waste of money as I am pretty sure it is not just naturopaths who say that coffee interferes with vitamin absorption. However, I disagree that a review of randomised control trials was the best basis on which to make such a decision. A better evidence would be to check the evidence to see if vitamin deficiencies could cause death and then for what vitamin levels are optimal of somebody his age and gender, understand what impact the multivitamin was (or was not) having on his individual vitamin levels and then act accordingly.

It is a pretty similar problem that I am left contemplating at the end of the chapter. I might be wrong, but it seems like a lot of research goes into understanding how different bodily processes, and the chemicals they are interested in bringing in to modify them, work before they are rolled out for a randomised control trial in humans. Even after medications have been approved for general use, usually there are also some diagnostics before the medications are prescribed. (OK, I know that there are lots of drugs, such as anti-psychotics, that we have very little idea as to how they work and for which people they will work. However, this is the source of much concern in science.)

If we do not know what context the policy is being asked to work on then we are not really sure what cases are the same or different. Perhaps more significantly, societies are not patients, and we cannot simply check all aspects of every individual in a group’s context in the way that a prescribing doctor can check a current list of medications.


Tags: ,

6 Responses to “Randomised control trials for government policy?”

  1. Tomboktu Says:

    I presume you know what Leigh’s new job is?

    • Tracey Says:

      Ah! Thanks for the comment because otherwise I would not have gone back to look him up. I just keep getting surprised by just how many of people writing on my topic are connected with the ALP (or suddenly stopped writing anything really critical when the ALP got in and they happened to end up with jobs in the public service).

  2. Tomboktu Says:

    Funny that. Over the last seven or eight years, between my day job and my postgraduate study, I’ve on more than one occasion commented at a seminar or conference on social justice, poverty, or community development, that the speakers were the Irish Labour Party at its day job. So many of its non-politicians activists hold posts in those kinds of organisations.

  3. Arguments for bridging social capital are making it into our newspapers « Another student blog Says:

    […] Leigh’s work was the topic of an earlier post on this blog. In that post I was not all that complimentary in regards to the chapter I had read and I can […]

  4. Another student blog Says:

    […] think isomorphic mimicry is a problem, and I am really not sure that we ever have enough information about the social world to know which contexts are the same as each other. Perhaps this is why I really should not be in an applied school? It turns out I […]

  5. A quick reference to Leigh and social capital « Another student blog Says:

    […] to spend a bit more time in the ‘real world’ to be able to appreciate his arguments for randomised control trials and what he chooses to measure. Still, I have been collecting various references to his work, so I […]

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s