You are on page 1of 22

Comparative Analysis of Social and Public Policy

(CASPP)

Week 8

Single Case Analysis

1
Outline

I. The Case Study and Comparative Approaches

II. (Single) Case Selection Strategies

III. Other Arguments for Case Studies

IV. Conclusions

2
(I) The case study
• the investigation of a single country (or region, health care
system, policy process, etc) may or may not be a case study

……: calling a single country study a case study is a comparative act!

>> cases do not exist – we ‘construct’ them (x is a case of….)

• linking an object of study (e.g. a particular middle income country, a


liberal democracy, a welfare state, national health care system etc) to
a broader universe of similar objects (all middle income countries,
liberal democracies, welfare states, public health systems)

• suggests use of concepts that can ‘travel’ (earlier session)

• Why one case?: Opens possibility for contributing to cumulative


knowledge of multiple cases……but often seen as having weak
potential to directly contribute to theory building, because:
3
Single case study: N=1

Critical claim = a single case only provides one observation of a


phenomenon, providing a weak basis to identify wider causal
relationships or generalise findings to other cases. A single case
therefore cannot provide valuable comparative insights.

but is this true?

4
1. comparison within one case
• ‘(de)construct’ a case (e.g. country) of particular
interest into multiple cases for comparison

• exploit relevant macro-social/political variations


– socio-economic structures
– policy delivery (e.g. health care, education)
– governments
– etc.

• studies about sub-national entities can contribute to


theories about nations

5
2. comparison within case over time

• exploit relevant variations (in governments,


economic conditions, policies etc.) over time
• “chunk” a real-existing historical narrative into
temporally distinct phases
• problem: cases rarely independent from each
other (Galton’s problem) – which is a potential
problem for any (also statistical) cross-national
comparison

Useful reference: Bengtsson, B., & Ruonavaara, H. (2017). Comparative


Process Tracing: Making Historical Comparison Structured and Focused.
Philosophy of the Social Sciences, 47(1), 44–66 6
Case study: N=1?

1. within country comparison (e.g. sub-national


units)
2. comparison over time (2 or more time
periods)
3. within country comparison over time
(combining 1 and 2)
>> effectively creating multiple cases out of a
single unit
>> but again: always with reference to general
phenomenon (otherwise not a case study)
7
Can single case studies provide theoretical
contributions?
• It is often argued that single cases can do little to test or
develop existing bodies of theory.

• However a case study may assist comparative research


interested in causation:

• hypothesis-generating (Lijphart) or ‘heuristic’ (Eckstein)


case studies.
 case studies can serve to generate propositions (‘pre-
theories’) that can then be tested through comparative
(cross-case) research.
• hypothesis-testing – testing if a chosen theory applies to
a particular case.

 ….much depends on ‘case selection’ 8


Studies Focused on Outputs and Outcomes

Cases studies may be used to explore the relationship


between policy design decisions vs outputs/outcomes:

benefit levels length of unemployment


tuition fees enrolment (socio-economic background)
free health total national health spending
child care female employment
job security extent of temporary employment

9
(II) Case selection strategies
• if cases are selected carefully and strategically, case
studies may be theoretically relevant

• there are various options for deliberate or strategic


case selection
- a warning: many overlapping (and confusing) labels in the literature!

• all presuppose that cases are selected with reference


to
– a) particular empirical features of the case (e.g. a federal
country)
and
– b) existing theory/generalisations (e.g. about institutional
influences on policy reform in federal countries)
10
Strategic case selection for case studies

• various types of deliberate case selection are


possible
• for example:
1. typical case
2. deviant case
3. diverse cases
4. crucial case (type a; type b)

11
1. The typical case

• representative case; well explained by theory

• aim: elucidating the causal mechanisms that


underpin associations between variables:

– resolving problems of causal direction


– exploring ‘micro-foundations’ on which theories
developed at macro-level are based
– testing theories explicitly about policy processes

12
1. The typical case

• Example: trade union density (membership) vs


social spending (Sweden as typical case)

• Contributions
– can explore causes rather than assuming them
– might identify different pathways to similar
outcomes

• Problems
– may easily degenerate into simple historical
narrative (descriptive story telling)
13
2. The deviant case
• opposite of typical case selection (not explained by
existing theory)
• a case which is anomalous with reference to
standard understanding - surprising values
 e.g. France: low trade union membership but high social
spending
• aim: probe for new explanations
• finding the unspecified factor which explains an
association (after which it is not longer deviant!)
• contribution: specification (refining) of theory
– improving existing models; probe new explanations
– confirming general rule (“exception that proves the rule”)
14
3. Diverse cases

• at least two cases to cover full range of


variation (are at opposite ends of scale)
• aim: elucidating the causal mechanisms in
each case (e.g. different paths may lead to
same outcome)
• exploratory (hypothesis generating) or
confirmatory (hypothesis testing)

 an example….
15
Example
Theory:
Employment protection (EPL) for regular
(permanent) workers has an effect on national
employment patterns
Hypothesis:
the strength of national employment protection
affects the level of temporary work in a country.
……so the stricter the EPL (job security) for
regular jobs the more prevalent temporary work
16
selecting case studies

17
Diverse cases (e.g. EPL and extent of
temporary work)
• at least two cases to cover full range of
variation
e.g. UK (or DK) and Portugal (or Slovenia, SI) –
both ends of the ‘spectrum’

• elucidating the causal mechanisms in each


case (e.g. role of employers, other policies,
social composition of temporary workers etc)

• we can identify causal paths, and perhaps


confirm or question our initial hypothesis
18
4. Theory guided selection:
the crucial case (a)
• type 1: ‘least likely’/confirmatory
• aim: show that a general theoretical argument is likely to be
valid by (successfully) exposing it to a demanding test
– “Sinatra: “If I can make it there, I'll make it anywhere”.
(New York, New York)

• contribution: prove a theoretical argument


 example….
 theory: child care spending has positive impact on gender
equality (pay; jobs)
 select a country with male dominated labour market &
strong family care culture (demanding conditions, does
theory hold?) 19
Theory guided selection:
the crucial case (b)
• type 2: ‘most likely’/disconfirmatory

• aim: show that a general theoretical argument is likely to be


invalid by (unsuccessfully) exposing it to a particularly
undemanding test
• contribution: disprove an existing theoretical argument
 example….
 theory: child care spending positively affects gender equality (pay;
jobs)
 select a country with little gender biased labour markets & weak
family care culture (theory may not hold even under ‘undemanding’
conditions)

Real example: Paster, T. (2011) German Employers and the Origins of Unemployment
Insurance: Skills Interest or Strategic Accommodation? MPIfG Discussion Paper 11/5; 20
(III) Other arguments for single
country/case studies

• cost, practicalities etc.

• greater likelihood of researcher having (or


acquiring) contextual cultural knowledge of
the case studied

• perhaps preferable to multi-case studies


where case selection may driven by data
availability, data reliability, access issues etc.
21
(IV) Conclusions
• you can be a comparative policy researcher even if you focus
on one country (or other case) only!

• exploit relevant variation within countries (between sub-


units and/or over time) to construct comparisons
or
• select a single country case strategically so that it can ‘speak’
to existing theories

but always be clear and explicit about:


• the rationale for your case selection
• the comparative dimension of your country study
22

You might also like