## Fellowships in Stat only: good news?!

**A**s reported in Nature newsblog, the UK funding body, EPSRC (Engineering and Physical Sciences Research Council), “has scrapped fellowships in all but two areas of mathematical sciences, namely statistics and applied probability”. This decision may sound like a bonanza for statisticians and applied probabilists, however, when thinking about it a bit more widely, it is close to a disaster. Choosing to fund fellowships only in a narrow subset of the field is indeed unfair, unwise, and inefficient. ** Unfair** because the topics were chosen w/o consultation with mathematicians. It could have been numerical analysis or cryptography instead. In which case it would have impacted statisticians and applied probabilists as well. Thus, top UK statisticians like Peter Donnelly and Peter Green rightly signed a protest letter along colleagues from other mathematical fields. (Maybe the RSS has likewise reacted. I have not seen it.)

**, because, as noted in the letter sent a week ago by twenty-five top UK mathematicians to their Prime Minister, cutting funds in most of mathematics will mean that most UK PhD students will leave the UK to get fellowships abroad. With a fair chance of never returning. (Maybe a bonanza for France? Not really, either, as the funding has not increased here and the current French PhDs need to be funded as well. Even though they most often get hired within a few months of their defense. Or leave for a postdoc abroad…)**

*Unwise***, because the decision is taken without prior notice and cannot expect to impact the area of research of future PhD’s. Nor does it bring a solution for the future of current PhD’s in**

*Inefficient***…**

*Not!{statistics and applied probability}*
October 3, 2011 at 2:00 pm

[…] Christian Robert gives his views on the EPSRC’s decision to accept fellowship applications only in statistics and applied probability until further notice. Like this:LikeBe the first to like this post. […]

October 3, 2011 at 1:21 pm

That is an amazingly short-sighted decision by the EPSRC.

October 3, 2011 at 11:44 am

I’m not saying this is a good decision; it isn’t. But I also think it’s important not to exaggerate its potential impact – actually there are things EPSRC are doing that are probably worse for the mathematical sciences.

First, you say “most UK PhD students will leave the UK to get fellowships abroad”. This doesn’t follow at all. Most postdocs in mathematical subjects, funded by EPSRC, did not get fellowships in the past, they got post-doc research assistant posts (PDRAs), which is where they work on research project is supervised by someone else. According to EPSRC in their evidence to the international review of mathematical sciences they ran last year (see http://www.epsrc.ac.uk/SiteCollectionDocuments/other/MathsIR2010EvidenceDocumentsParts1-3.pdf , section 7.2 starting on page 42), there have been only about 10 PD research fellowships in maths per year (incidentally, none in statistics between 2005 and 2009), compared to about 50 PDRAs per year in the mathematical sciences. So “most” UK PhD students, if they got EPSRC funding, didn’t get fellowships, they got PDRA posts, so taking away fellowships won’t affect most of them. Cuts to research grant funding (which have also happened) are likely to have a bigger impact on maths postdocs, I think, because they affect the number of PDRA posts, and there seems to be much less complaint about that (at least, not that I have heard).

Second, postdoc research fellowships are not a common way of funding across the area of responsibility of EPSRC. E.g. in 2010 they existed only in theoretical physics, mathematical sciences, and what they call “cross-disciplinary interfaces”, so nothing in other sciences or in most of engineering or in computing (and that’s the case this year too). I don’t think that has brought those research areas to their knees.

Third, it’s a bit misleading, even though it’s true, to say there was no prior consultation of mathematicians. There was a big international review a bit less than a year ago, which, after considerable consultation of mathematicians and others, concluded (among other things) that statistics research is in a hazardous state in the UK for various structural reasons. On the whole these conclusions were welcomed. Now the EPSRC total budget is fixed by the Government, so if they are to spend more on statistics, it is inevitable that they must spend less on something else.

October 3, 2011 at 7:58 am

According to an addendum to the Nature blog you link to, they’re not really cutting fellowships to other areas of mathematics:

“Atti Emecz, director of communications at EPSRC, says that it has always been EPSRC’s intention later in the year to broaden the areas supported by mathematics fellowships.”

December 20, 2011 at 10:45 pm

@David (Reid?): It’s now the end of the year. A “further” set of fellowship priorities were announced in November. The decision is to keep postdoctoral fellowships closed to all disciplines except statistics and applied probability. Later stage fellowships have been broadened to two EPSRC-created categories “New Connections between Mathematical Sciences and Information Communication Technologies” (early career only) and to “Intradisciplinary Research” (early career and established career): http://www.epsrc.ac.uk/funding/fellows/Pages/mathematicalsciences.aspx