In the first post of this series, I noted that much of the reduction in SSHRC research grant success rates during the transition from the old 3-year Standard Research Grants (SRG) to the new 5-year Insight Grants (IG) could be explained by the stock-flow dynamics of changing grant durations. But that's not the whole story: success rates also depend on the total funding envelope.
Here is the history of success rates (projects funded as a percentage of applications) and funding rates (total research funds awarded as a percentage of total requested) for the entire SRG program:
I don't have data before 1995, so I don't know if the relatively low success and funding rates of the mid-1990s reflected the historical rates that prevailed before then, or if they had been suppressed during the budget cuts of that era. What is clear is that SSHRC research budgets were increased in the latter part of the 1990s. For roughly a decade, success rates were just above 40%, and funding rates were almost 35%. These rates dropped sharply after the Conservatives took power.
At the time, I figured that the Conservatives had simply decided to cut funding to the "knuckle-headed Marxists who wouldn't vote Conservative in a million years." I now think that this is an unfair characterisation, and that it's more likely that the Conservatives were simply trying to bring spending under control.
If you have a research grant program that maintains roughly constant success and funding rates, then the total awards budget is essentially determined by how much is requested: if funding requests double, then so does the awards budget. And that's more or less what happened during that decade.
Here is a graph of total SRG awards. All expenditure terms in this series are in constant 2015 dollars, using the GDP deflator:
Total real SRG awards almost tripled in 8 years before they were leveled off in 2004. The cuts in 2007-08 still left the real awards budget roughly twice as big as it had been ten years earlier.
Much of this increase was driven by an increase in grant applications, from around 1600/year in 1995-1997 to 2500/year in 2005-2008. But I suspect that what concerned policy-makers at SSHRC in the mid-2000s was what was happening to grant award sizes. Here are real average SRG awards expressed at annual rates - that is, the average award divided by three. (Expressing awards in terms of annual averages will make it easier to make comparisons with the 5-year IG awards later on.)
I don't know to what extent the implications of SSHRC's policy of constant funding and success ratios were known to the research community - I certainly wasn't aware of it at the time - but it would appear that the policy of giving out more money when more money is asked had the expected effect on awards. When you look at this chart, it seems at least plausible that the goal of the cutbacks in the mid-2000s was to trim the size of real average awards back to where they had been ten years earlier.
So now we move to the IG regime - what happens the the budget?
This graph actually underestimates the budget increase. In addition to the IG program, SSHRC also introduced the 'Insight Development Program', which essentially sought to serve people who might have applied for SRGs as 'new scholars'. (I like to think I played a part in the creation of the IDG, but that's another story.) The IDG budget in 2015-16 was some $17m, so the total SSHRC research grant budget is now roughly comparable to its mid-2000s peak.
Again, I'm guessing here, but I think that the hope at SSHRC was that now that some order had been restored to grant sizes, that discipline would be maintained so that the effect of a budget increase would show up in success rates.
Not so much, as it turned out:
For whatever reason, one of the effects of moving from the SRG to the IG was a drastic increase in grant awards, far beyond what you'd expect from the mechanical increase of moving from three-year to five-year grants. Why?
This is the second of a four-part series:
I - Why are success rates so low?
III - Research in Buzzword Studies
IV - A collective action problem
It could be a combination of push and pull factors.
If researchers had internalized the previous low funding rate, they could have inflated their grant proposals. The idea would be "okay, I might only get 60% of what I ask for, but that 60% will be the part I really want."
The other, enabling half might be a change of research direction. As I understand things, the Harper government was often more interested in "high impact" grants, where fewer researchers would get higher-value grants. This kind of thinking was the genesis of the Canada Research Chair program.
Posted by: Majromax | August 10, 2016 at 10:00 AM