Social Impacts Measurement
in Government and Academia
Daniel Fujiwara
d.f.fujiwara@lse.ac.uk
Cabinet Office & London School of Economics
I. Social impacts in public policy
• Economists in government have a long tradition of
measuring social impact: Cost-Benefit Analysis (CBA).
The Green Book (1)
• CBA enters at the Appraisal stage
• Appraise policy interventions in terms of their social
costs and benefits.
• Further supported through the Social Value Act (2012)
ROAMEF cycle
The Green Book (2)
• Supplementary Green
Book guidance provides
guidelines on how to
value social impacts
The Magenta Book
• Assessing whether the policy had a causal
effect/impact
II. Social impacts research (1)
• Valuation and Evaluation (causality) are key to
understanding social impacts.
Valuation
• Theory: to find the monetary equivalent of the change in
welfare or wellbeing associated with experiencing or
consuming the ‘good’.
• We could look at people’s preferences or their
wellbeing
• Wellbeing valuation: recent research is looking at
assessing value in terms of changes in subjective
wellbeing (Fujiwara & Campbell, 2011) with lots of
potential for housing issues.
Social impacts research (2)
Evaluation (causality)
Evaluation scale (Dolan & Fujiwara (2012) BIS technical report)
•
•
Theory: Identifying
and measuring the
counterfactual
Many public sector
organisations moving
to the Maryland
Evaluation scale.
This scale ranks how
well counterfactuals have
been measured in the
analysis.
Level
Design
Statistical method
5
Randomised trials
Evaluations with well implemented
random assignment of treatment to
subjects in treatment and control
groups.
4
QuasiExperiments
Evaluations that use a naturally
occurring event (that makes the
treatment assignment as good as
random)
3
Matching
techniques;
Regression
analysis
Non-experimental evaluations where
treatment and comparison groups are
matched on observable characteristics
2
Simple
comparisons
Studies with a treated and comparison
group, but with no attempt made to
control for differences among the
groups.
1
Pre- and post
analysis
Studies where no comparison group is
used. Outcomes are measured pre
and post-treatment.
III. Housing and social impact
Example: Attaching values to aspects of housing through
the Wellbeing Valuation approach
Preliminary findings (not for citation)