Comparative politics: Theories and methods

1
Comparative politicsTheories and methods

• Ana Rico, Associate Professor
• Department of Health Management and Health
  Economics
• ana.rico_at_medisin.uio.no

2
OUTLINE

• I. Health politics Content, motivation and goals
  of the course
• II. Political science and comparative politics
  problems methods
• Problems of political science research
• Small-N and Large-N approaches
• Comparative politics the fuzzy centre
• III. Types of theories
• From early theories monocausal (socioeconomic,
  cultural or institutional) determinism
• To multi-causal theories (comparative
  politics, political economy)
• IV. Causes of policy change (IVs)
• V. Applications (DVs)
• Causes of the expansion of the WS/HC systems in
  OECD countries
• Causes of retrenchment/restructuring/resilience
  of WS/HC systems
• VI. Course assignments

3
Health politics Contents, motivation and goals
of the course
4
HEALTH POLITICS Content and motivation

• 1. Content and motivation of the course
• A) Content
• Application of political science theories to the
  health care sector
• Based on the comparative-historical method ?
  tries to build the gap between qualitative
  case-studies and quantitative statistical studies
• Focus on how to build and criticise concepts and
  theoretical arguments
• B) Motivation
• Compare health care (HC) with other sectors of
  the welfare state (WS)
• Understand how the WS was built, and which were
  the causes of its emergence
• The main distinctive feature of Europe?
• Outside Europe, a lot of interest on how to
  replicate it
• It can also help to understand how to mantain it
  or expand it further
• Some paradoxes...
• Health politics has been developed mainly by US
  scholars
• In Europe, recent and often focused on how to cut
  back public health care
• We know a lot about health policies (e.g. Managed
  competition), but little on how to introduce them
  (health politics)

5
HEALTH POLITICS Goals

• 2. Goals of the course
• A) Conceptual
• Review and classify existing research traditions
  in political science
• Learn how to build and criticise concepts and
  theoretical arguments in a convincing way
• Learn how to build simple causal models out of
  complex theories and facts
• Analyze the political determinants of health
  policy
• B) Empirical
• Assess the analytical goodness of fit between
  theory and evidence (validity)
• Study the causes of health care (and welfare
  state) reform ? policy change
• Focus on how to build universal WS/HC systems,
  and how to expand them
• C) Practical
• Develop analytical skills as well as
  (experiential) clinical eye from reform
  processes
• Learn a new language concepts of political
  science and health system research

6
Political science Methods and problems
7
POLITICAL SCIENCE COMPARATIVE POLITICS

• 1. Problems of political science (PS) research
• 1.1. Lack of agreement across schools of thought
• Divided across theoretical (which particular
  cause they emphasize) and ideological lines
  marxism, liberalism, structuralism, culturalism,
  pluralism, institutionalism, rational choice
• Little communication, often tough competition and
  rivalry between them
• ? Lack of agreement on how to define
  operationalize basic terms
• E.g. Institutions
• 1.2. Many variables, few cases
• Object of study involves complex macro-social
  phenomena (e.g. WS) characterized by
• Multidimensional concepts, meaning varies across
  time place, difficult to operationalize
• A great number of potential causes, not
  independent among them
• Long causal chains and complex causal mechanisms
• Feedback effects, endogenous causation and
  selection bias (Przechevorski in Kohli)
• Object of study (e.g. WS) often occurs only in a
  few cases (e.g. OECD countries)
• Difficult to build simple concepts and models
  that can be tested across cases (reliability),
  but still resemble real world complexity
  (validity)

8
POLITICAL SCIENCE COMPARATIVE POLITICS

• ? Three methodological approaches to adress such
  problems
• Narrative case-study (N1) approach
• Qualitative, in-depth study of a single case
• Internal validity high but ad-hoc explanation,
  low reliability
• No generalisation or theory-based explanation
  possible
• Statistical (large N) analysis
• Too many cases to know in-depth what happened in
  each of them.
• Too little variables taken into account, simple
  or no causal mechanisms
• External validity goes at the cost of internal
  validity
• The comparative method medium-N (N2 or more)
• Seeking a balance between number of cases and
  number of variables
• Seeking a balance between internal validity and
  external validity
• The goal is to build general theories, but based
  on cases researched in depth
• The research design is quasi-experimental
• Research techniques can be qualitative or/and
  quantitative

9
POLITICAL SCIENCE COMPARATIVE POLITICS

• 1. Case(country)-based Small-N (1) studies,
  e.g. US Welfare State
• Old research tradition historical, in-depth
  analysis of one case-study
• It simplifies by reducing N so that all or most
  potential causes can be analysed
• It allows a detailed analysis of
• Specific characteristics of countries (e.g
  unique, model causes)
• complex causal mechanisms, and
• time sequence of events
• GOOD INTERNAL VALIDITY (causes represent well
  real world complexity),
• BUT
• Little external validity (causes cannot be
  tested/applied to other cases)
• It does not allow to discriminate between general
  and country-specific factors

10
POLITICAL SCIENCE COMPARATIVE POLITICS

• 2. Variables-based Large-N studies, e.g. Causes
  of democracy
• New research trends (modelling) study a few
  variables across a large number of cases
• It simplifies by (1) selecting a few variables
  considered as key causes (2) reducing
  complexity or variability by making restrictive
  assumptions
• (eg all the causes are independent of each other
  the causes are independent of the effects
• causal sequence does not matter
• Cross-national statistical studies are the main
  example (over 100 countries)
• GOOD EXTERNAL VALIDITY (under tight, often
  unrealistic assumptions), but
• Little internal validity too many omitted
  (extraneous) variables, disregards variables
  which are difficult to measure (often the most
  relevant), model is not a fair image of reality,
• It is used by many schools that believe we can
  find general regularities across cases
• The most radical is rational choice analysis
  game-theory, which works best when the
  restrictive behavioural assumptions of classic
  economics apply

11
Causal mechanisms in small-N case-studies
POLITICAL SCIENCE COMPARATIVE POLITICS
Causal mechanisms in large-N case-studies
Self-interest
Assumptions eg Behaviour
Social interaction
Norms values
12
POLITICAL SCIENCE COMPARATIVE POLITICS

• 3. The comparative-historical method
• Thinking without comparison is unthinkable
  (Swanson, 1971 quoted by JK Helderman)
• Seeks a balance between N and number of
  variables
• N1 ( reference case/s). Case-study treated
  comparatively
• Ideal type ( rest of cases) ? Often example of
  best practice
• Deviant/critical case ( average case) ? Causal
  mechanisms contradict establised, general theory
• N1 Sector 1/Sector 2 Time period 1/Time period
  2
• N2, 3, 4. Matched comparisons
• N gt 4. Requires
• Combination of quantitative/qualitative. QCA, OLS
• Secondary qualitative/historical evidence on Ngt4
  cases

13
Most-similar cases (method of difference)

• Case 1
• A
• B
• C
• X
• Outcome Y

• Case 2
• A
• B
• C
• Not X
• Outcome not Y

Main cause (X) is present in one case and missing
in the other. Both cases are similar (matched) in
all other respects. The main effect Y is present
when X present, and absent when X is absent
(points to a necessary and sufficient cause).
14
Case-study, before-after design

• Case 1, T0
• A
• B
• C
• Not X
• Outcome not Y

• Case 1, T0
• A
• B
• C
• X
• Outcome Y

Main cause (X) is present in one case and missing
in the other. Both cases are similar (matched) in
all other respects. The main effect Y is present
when X present, and absent when X is absent
(points to a necessary and sufficient cause).
15
Most-similar, graduation in cause-effect

• Case 1, Sector 1
• A
• B
• C
• X0
• Not Y

Case 1, Sector 2 A B C X1 Y1
Case 1, Sector 3 A B C X2 Y2
Main cause (X) is present to different degrees in
two cases and missing in the other. The cases are
similar (matched) in all other respects. The
effect Y is present when X present, and its
graduation corresponds to Xs one (additional
evidence of necessary and sufficient cause).
16
Close to most-similar, N4

• Case 1
• A
• C
• E
• X
• Y

Case 2 A C D Not X Not Y
Case 3 A B D X Y
Case 4 A B E Not X Not Y
Main cause (X) is present in two cases and absent
in the other two. The cases are only imperfectly
matched, but rest of potential causes do not
correspond to the effect (can be necessary or
suficcient, but not both). The effect Y is
present when X present., absent when is absent
(necessary and sufficient cause).
17
Most-different cases (method of agreement)

• Case 1
• A
• C
• E
• X
• Outcome Y

• Case 2
• B
• D
• F
• X
• Outcome Y

Main cause (X) is present in both cases. They are
very different in all other relevant respects.
This suggests that the effect Y always occurs
when X is present (necessary cause) the rest of
factors can be irrelevent or sufficient causes.
18
POLITICAL SCIENCE COMPARATIVE POLITICS

• 3. The comparative-historical method (cont.)
• It can be used for explanatory, descriptive and
  prescriptive analysis
• A. Under a quasi-experimental research design, it
  can be used for inference (explanatory studies) ?
  Matched comparisons, QCA, OLS
• QCA offers some advantages over OLS (Mahoney),
  eg
• It allows for the different categories of a
  tipology - as the DV (eg Esping-Andersen and the
  three worlds of welfare capitalism) being
  explained by different combinations of causes
• B. Less strict, more qualitative comparisons can
  be used for descriptive and prescriptive
  purposes, e.g.
• Concept formation and categorization e.g.
  Definition and types of WS, the concept of
  representation (Pitkin)
• Operazionalization of complex concepts e.g.
  Democratic institutions (Executive/Parliament
  Dominance, Federal/Unitary, Majoritarian/Consotiat
  ionalProportional, Corporatism/Pluralism) ?
  Liphart

19
POLITICAL SCIENCE COMPARATIVE POLITICS

• 3. The comparative-historical method (cont.)
• B. Less strict, more qualitative comparisons used
  in descriptive and prescriptive purposes, e.g.
  (CONT.)
• Building hypothesis and evidence on
• Complex case-specific interactions between IVs
• Effect of historical accidents as sufficient
  causes
• Temporal sequences of causes
• Complex causal mechanisms
• Mapping and comparing policy alternatives for
  policy-makers
• Studying the key causal mechanisms of a case of
  best practice in order to imitate it
• Uncovering specific temporal sequences of events
  in the history of a deviant case to facilitate
  removal of obstacles to change

20
CAUSAL MECHANISMS IN COMPARATIVE POLITICS
Sociopol. actors
Politicalactors
Process, interact.
Context
Policy
Institutions Resources
21
Types of theories
22
TYPE OF THEORIES CAUSES OF POLICY

• III. Types of theories
• From early theories monocausal (socioeconomic,
  cultural or institutional) determinism
• Good for advising/influencing policy-makers
• Often professional interests/ideology of
  researchers
• Limitations of quasi-experimental, qualitative
  comparisons
• Little information available
• To multi-causal theories (comparative
  politics, political economy)
• IV. Causes of policy change (IVs)
• 1. Structural, cultural and convergence theories
  SOCIAL CONTEXT (audience)
• 2. Actor-centred theories POLITICAL ACTORS
  (players/teams/clubs)
• 3. Institutionalist theories INSTITUTIONS (
  rules of the game)
• 4. Action-centred theories INTERACTIONS
  (game/league)
• 5. Policy-centred theories PATH DEPEPENDENCE,
  FEEDBACK, LEARNING

CONCEPTUAL COMPLEXITY INCREASES
23
POLITICAL SCIENCE COMPARATIVE POLITICS
Exercise Amenta et al. 2004
Types of theories

• EUROPE-BASED EXPLANATIONS
• Modernization ec.development
• Partisanship theory (SD, CD)
• Coalition theories
• Institutional theory centralization/
  /fragmentation of the polity
• State-centred theory state capacity
  and autonomy state bureaucracy
• Path-dependency, policy feedbacks
• US-BASED EXPLANATIONs_
• Race
• Social movements (citizens. mob.)
• Interest group theory capitalists
• Public opinion
• Patronage, non-ideological pol. parties
• Democratic polity openess access

• 1. Context theories
• Structural,
• Cultural
• Convergence theories
• 2. Actor-centred theories
• Interest groups
• Political parties
• State-centred
• State-society civil society
• 3. Institutionalist theories
• 4. Process-centred theories
• 5. Policy-centred theories
• Path dependence,
• Policy feedbacks policy learning

24
Causes of policy change
25
CAUSES OF POLICY CHANGE Operationalization in
WS/HC research

• Access participation
• Policy strategies
• Coalition-building
• Competition and cooperat.
• Changing resources
• Learning

• Conjunctural factors ec crisis, wars

• Socioeconomic structure
• Ownership, income
• Education, knowledge
• Social capital (status, connections)

CONTEXT

• Sociopolitical structure
• Cleavages and political identities

• Values Culture and subcultures

• Interest groups
• Profesional assocs.
• Poilitical parties
• State authorities
• Citizens PO/SM
• Mass media

• Distrib. of formal pol. power electoral law,
  constitution, federalism, corporatism
• Contracts and org. structures
• Norms of behaviour
• Sanctions/incentives

POLITICS InteractionsProcess
Preferences Resources
INSTITUTIONS
POLITICAL ACTORS
Individual and collective
-
Formal and informal

• Entitlements rights
• Regulation by law (of power, ownership,
  financing, behaviour, contracts)
• Redistribution Financing RA
• Production of goods services

POLICY
Adapted from Walt and Wilson 1994
26
Assignments
27
ASSIGNMENTS

• Presentation and participation in class (10
  10) ?
• 1 or 2 students per article
• 1 or 2 presentations
• Summary, partly based on graphic tools
• Criticisms to unclear or overstretched
  concepts, unconvincing arguments,
  counterarguments, lack of correspondence between
  concepts and evidence, insufficient evidence,
  important omitted variables, others
• Course paper (30) ? Groups 2-4
• History of Norwegian WS/HC compared with other
  case
• Recent reforms expanding the HC sector compared
  with case
• Exam (50) ? 5 December
• Concepts and theories
• Text to discuss

28
ASSIGNMENTS
Wed. 21 Sep., 1315-1600 Ana Rico The social context of health politics  Mechanic Rochefort 1996 Bouguet 2003 Svallsfors 1997
Wed. 28 sep., 1315-1600 Ana Rico Interest groups and political parties  Olsen 1982 Quadagno 2004 Hunold 2001
Wed. 5 Oct. 1015-1200 Student-led session Case 1 Interest groups in the US WS C2 Corporatism and professional self- regulation in EU HC. Navarro 1989 Quadagno 2004 Greß et al. 2004 Blom-Hansen 2000
Wed. 5 Oct., 1315-1600 Ana Rico The role of the state government, parliam-ent, and bureaucracy  Skocpol 1980 White 2003 Howlet Ramesh, ch.2.
Wed. 12 Oct., 1315-1600 Ana Rico Civil society policy experts, public opinion and mass-media Hall 1993 Manin 1989, ch6 Hoffman 2003
Wed. 19 Oct., 1015-1200 Student-led session Case 3 The political economy of the WS in US, UK Sweden Case 4 Mass media public opinion in Clintons HC reform Hall 1993 Weir Skocpol 1983 Jacobs 2001 Goldsteen et al 2001
29
ASSIGNMENTS
Wed. 19 Oct., 1315-1600 Ana Rico Institutions Division of powers, veto points and regulation Immergut 1992 Scharpf 2000
Wed. 26 Oct., 1315-1600 Ana Rico Action theories and the political process  Korpi 1989 Garrett 1993 Rico Costa 2005
Wed. 2 Nov., 1015-1200 Student-led session Case study 5 HC expansion in the UK, the US, and Canada Case study 6 The origins of the US WS Jacobs 1992 Maioni 1997 Briggs 2000 Jenkings and Brents 1989
Wed. 2 Nov., 1315-1600 Ana Rico The new welfare/health politics and the debate on retrenchment  Pierson 1996 Clayton Pontusson 1998 Tuhoy 1999
Wed. 9 Nov., 1015-1200 Student-led session Case 7 Evidence on retrenchment in WS Case 8 Politics of retrenchment in HC Korpi 2003 Allan Scruggs 04 Hacker 2004 Oliver 2004