In my day job, I'm asked to spend time thinking about measuring things
that are vaguely defined. Why?

I manage Unlock, one of the newest and smallest of around 50
information services funded by JISC. Around half of JISC's budget goes
towards JANET, the high-bandwidth network - JANET is to JISC what the
Royal Opera House is to the Arts Council.

The rest of JISC's cash goes to support research information
projects and services - licensing data, collecting it, making it
available on the web, building tools to analyse it, providing users
with technical and legal advice, making grants for IT innovation.

Along with all my peers, I'm asked to help JISC measure whether my
service delivers "Impact" and "Value". There's a big hidden cost in
this - time in meetings, data gathering, developer time for log
analysis, end-user time giving feedback, survey fatigue, and loss of
morale. JISC faces a budget cut of between 5% and 10%, year-on-year
for the next several years. JISC itself is funded by HEFCE, whose
half-a-million-squid budget cut was lately in the news.

So we all must measure the Impact And Value of our activities; measure up, and fear the axe.  But my idea of what Impact means is probably very different from yours, and yours from JISC's, and JISC's from Peter Mandelson's: we are all second-guessing each other.

My favourite thing about the Social Policy Bond idea is the focus on outcomes, on defining achievable goals. Outcomes must be meaningfully measurable. For any grant-making organisation, whether a charity, a quango or a private foundation, a focus on outcomes will help to maximise value for money.

"Impact" may be measurable, but it is never clearly defined - for some
it means potential economic value of commercialisation - for others it
means citation or reuse of work by other researchers. Impact is
different for every research field. An archaeologist told me that
impact in his field may not even be felt for another 50 years.

So how could JISC make good use of social policy bonds to define and achieve its goals?

For example - JISC and other academic funding bodies want to encourage
open access to research papers and underlying data. Funders tie strings
to grants that require open access results, or they make grants to
universities to build OA repository systems. Instead, JISC could issue an
Open Access Bond. The bond could be set to mature when, say, 98% of
HEFCE-funded research was available on an open access basis. Currently
the volume of research that is open access is more like 15%, so
initially the bond would likely be heavily discounted.

All kinds of institutions could invest in the bond - universities,
corporations, foundations. Some would be doing it for "ethical"
reasons, others for purely economic ones. Investors would then have an
incentive to see more research made open access - by providing free
hosting, or teaching researchers about the benefits, or applying peer
pressure, or generating media coverage about the benefits.

Another example - JISC could issue a Zettabyte Environment Bond that
matures when 1 Zettabyte (to pick an arbitrary very high number) of
environmental measurement data is being published annually by UK
research institutions. There would have to be conditions, to avoid
junk data, excessive duplication. Institutions who are already helping
to achieve this goal would buy the bond cheap; when it matured, they
could plough the profit back into ongoing research. Corporations whose
work supports this goal would buy the bond to provide evidence that
they want to be stakeholders, in the effort for the long term.

Another example - JISC could issue an Archive Digitisation Bond that
matures when 75% of university library archives are available in a
digital form. Institutions that are supporting their own digitisation
projects would buy the bond cheaply now. Their commercial partners
would have an extra incentive to pull the stops out, as when the bond
matures, the libraries will have profits to invest on completing the
efforts, and want to hire people who've already proven goodwill.

I could go on, but you hopefully get the idea, and can think of ways
to apply this to the goals of any grant-making funding organisation.

For example the Joseph Rowntree Reform Trust funds campaigning
organisations, who use the money to buy advertising, pay for poll
questions, hire venues, pay for consultants' travel - the grant money
quickly finds its way into hands that aren't working towards the
grantees' social goals. So the Trust could think instead about what
those goals are, about how to know whether the goals have been
achieved, and how to design a social policy bond that supports their
achievement.

And that helps me decide what I'd do with the tory party's imaginary 1
million quid - use it to develop a site that helps organisations administer
and broker social policy bonds.

Thanks, Hugh.