It is a mistake to eliminate government science. Part II

In part one of this post I talked about the fundamental importance of federal research.  I’d like to talk a little bit more about the implications of contracting out research.  These implications depend largely on how that contracting is performed.  While we might expect contracting of federal science to universities to provide a sort of status quo situation, this is not entirely the case.  

The federal government has a number of checks in place to ensure that federal research is disseminated freely (or nearly freely).  One has only to look at major federal programs like GeoGratis, Environment Canada’s databases and the OpenData program (a major initiative recently announced by the federal government).  These are things of which we should be justifiably proud.  This is Canadian taxpayer’s money providing a service, free of charge to Canadian citizens to boost scientific and business productivity.

The Canadian Government's Open Government logo
Figure 1. The Open Government initiative was announced by the federal government in 2011.

Compare this freedom of access to what often times happens in academia.  While free access to data is generally the norm, there are many instances of datasets being unavailable (for example these missing sites for the North American Pollen Database) as a result of a failure to distribute data following publication.  I am by no means arguing that all research should be placed in federal hands, but research that provides significant value to the federal government and has the ability to provide significant value to a broad number of stakeholders should be a priority for government.

There are a number of examples where governments have turned over research services to the private sector.  The UK turned over almost all capacity for plant breeding to Unilever in the 1987.  This had the effect of limiting ‘risky’ research since institutional plant breeding research became dependent on industrial support, and ultimately reduced the overall breadth of scientific research into plant breeding.  Thirtle et al. (1997) make the argument that privatization of science in this case has failed since the industrial partners failed to take on the research burden supported by the government, while at the same time taking over the profits generated by the ‘market’ research.  Levitow and Carr (2002) argue that we expect our public sector reseach to perform two tasks, to generate research of value to the public, and to act as a watchdog, assessing potential risks to the public health or in a broader sense to public assets.  With the increasing privatization we potentially lose both of those functions, the loss of risky research prevents the development of new research avenues and the loss of independence prevents the proper functioning of the ‘watchdog’.

A figure to brighten up the page.
Figure 2.  Publications and retractions in the biomedical literature are both increasing, but the rate of retractions is increasing faster. The drop in retractions in 2007 is a result of a lag between publication and retraction.  From: Cokol et al. 2008. Click on the figure for link.

The loss of openness in research comes with significant costs, both to the public and to industrial partners (although if this is the cost of owning the technology then ultimately the public pays again).  There is considerable waste in industrial research that comes from having a closed research shop, and from the very nature of private industry.  Comparing industry-centric conferences to publicly-funded research conferences is an exercise in measuring waste, not that I’m jealous or anything.

Industry sponsored biomedical research has further problems.  Positive publication bias in the literature has a cost that is underwritten by both industry and the public, as do retractions.  Closed data can also serve to mask poor quality research, or PR masking itself as research.  Retractions are particularly insidious, even after the publication of a retraction notice the original flawed publication is often cited positively in subsequent biomedical research (Neale et al., 2010).  If this is the case, then what are we to make of the citing articles?  They are using flawed research to support their own findings, does this lower the value of the citing articles?  Lower the monetary value of the research?  What is the cost to the public, to the industry supporting the research?  Is there value to an industrial partner in publishing a flawed report and then retracting it if that report will continue to elicit citations?

A recreation of Tyndall's light pipe, precursor to modern fibre optics
Figure 3. Shining a light through flowing water had little economic use for 19th century industry, but it is the basis of billions of dollars of wealth in modern society. (link in french)

There are costs to privatizing research, and costs that are borne that are not immediately obvious to the public.  If the government is serious about contracting research services it must ensure that data openness is preserved in all federally funded research and that some baseline of fundamental, and ultimately ‘blue sky’ research is preserved.  It is this risky research that often leads to major innovations.  Innovations that are transformative, that promote new industry and drive productivity.  Often scientists cannot imagine the economic potential for their discoveries, that, ultimately, is the domain of industry.  However, without fundamental research, industry would never be able to develop x-ray technologies, nylon or fiber optics.  After all, if we hadn’t asked why the sky was blue we’d never have understood how to squish all that light into those tiny glass threads.

Listening to:  Marvin Gaye – Come get to this (this video is pretty cool!  Just Marvin and his piano)

About these ads

Published by


Assistant scientist in the Department of Geography at the University of Wisconsin, Madison. Studying paleoecology and the challenges of large data synthesis.

4 thoughts on “It is a mistake to eliminate government science. Part II”

Leave a Reply

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

You are commenting using your 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