2 Classification and Comparison Normative Approach of Research

2 Classification and Comparison

Normative Approach of Research

• Emphasize what ought to happen, what should be

• Often include value judgments about optimal standards (norms)

• Closely connected with political ideology/philosophy

• Are the basis or starting point for most questions in empirical political science

Empirical Approach of Research

• Seek to discover, describe, and explain facts and factual relationships

• Tries to remain value-free, eliminate biases

• Use concepts, models, and “the scientific method” to uncover such relationships.

An Example

• An example: Democracy and comparative politics

• The normative perspective

• Is democracy desirable?

• What types of democracy are desirable?

• The empirical perspective

• How do we define democracy? (working definition)

• How do we measure the quality of democracy? (operationalization of the concept)

• What are the causes and consequences of democracy (the correlates of inequality)?

• Are democracies more equal than other forms of government

• Does democracy lead to development?

Empirical Findings

• Are democracies more “equal” in terms of income distribution than other forms of government?

• ANSWER: No

• EVIDENCE: Democratization in Eastern Europe

• Does democracy lead to development?

• ANSWER: Don’t know

• EVIDENCE: Przeworski (2000), Sirowy and Inkeles (1991)

The Goal of Comparative Politics

• Empirical approach

• To identify similarities and differences

• To identify and explain patterns of political phenomena and political behavior

• To construct science-like generalizations as regard to political phenomena

• To predict the outcome of other cases not yet observed

Is Prediction Really Possible?

Multiple Forecasts of Oil Drilling: 1980–90

Outline

• Classification:

• How to identify similarities and differences

• How to identify patterns of political phenomena and political behavior

• The logic of comparative analysis

• How to explain patterns of political phenomena and political behavior

• How to construct science-like generalizations as regard to political phenomena

• Comparative methods

Classification

Classification of What?

• Typology of political regime

• Typology of government

• Typology of the state

• Typology of socio-economic systems

• Typology of political culture

Plato’s Typology of Political Regimes

Aristotle’s Typology of Political Regimes

Machiavelli and Montesquieu

Machiavelli

• Monarchies

• Despotic

• Non-despotic

• Republics

Montesquieu

• Despotism, or the lawless exercise of power by the single ruler

• Constitutional forms of government of the monarchy and the republic

Modern Classification of Political Regimes

After 1917:

• Capitalist vs. Socialist systems

The 1930s and 1940s:

• Fascism vs. Anti-Fascism (Both capitalist and socialist systems)

The 1950s and early 1960s:

• Democracy vs. Totalitarianism

The Mid-1960s:

• Challenges to the dichotomy of democracy vs. totalitarianism

After the late-1960s 

• Democracy

• Authoritarianism

• Totalitarianism

The mid-1990s: Linz’s new typology

• Democratic regime

• Authoritarian regime

• Totalitarian regime

• Post-totalitarian regime

• Sultanistic regime

Linz’s Typology

• Similarities and differences along four dimensions

• Linz and Stepan (1996), Table 3.1

Democratic Regime

• A form of government in which the right to make political decisions is exercised directly by the whole body of citizens, acting under procedures of majority rule, usually known as direct democracy.

• A form of government in which the citizens exercise the same right not in person but through representatives chosen by and responsible to them, known as representative democracy.

• Liberal constitutional, democracy refers to political systems in which there are attempts to defend and increase civil liberties against the encroachment of governments, institutions and powerful forces in society

Totalitarian Regime

• A form of government that theoretically permits no individual freedom and that seeks to subordinate all aspects of the individual's life to the authority of the government.

• Dominant leader using a mass party and ideology to mobilize people to achieve state objectives

• Control citizen and group behavior totally

• State attempts to control the whole of society

• Concentrates power in the hands of an individual or a group

• Use of secret police, concentration or labor camps, ideological control or indoctrination

• Examples: the former Soviet Union under Stalin, Nazi Germany, Fascist Italy and Maoist China

Authoritarian Regime

• Authoritarian system refers to any political system that concentrates power in the hands of a leader or a small elite that is not constitutionally responsible to the body of the people.

• It also differs from totalitarianism, since authoritarian governments usually have no highly developed guiding ideology, tolerate some pluralism in social organization, lack the power to mobilize the entire population in pursuit of national goals, and exercise that power within relatively predictable limits.

Post-totalitarian Regime

• Post-totalitarianism refers to a type of non-democratic regime before the transition to democracy.

• How is post-totalitarianism different from totalitarianism and authoritarianism?

• Pluralism

• Ideology

• Mobilization

• Leadership

Sultanist Regimes

• This is a system based on personal rulership, but loyalty to the ruler is motivated not by his embodying or articulating an ideology, nor by a unique personal mission, nor by any charismatic qualities, but by a mixture of fear and rewards to his collaborators.

• Characteristics

• The blurring of the line between regime and state

• Personalism

• Constitutional hypocrisy

• Narrow social bases of Sultanism

• Distorted capitalism

• Corruption at all levels of society

Examples of Sultanism

• Rafael Leonidas Trujillo Molina in the Dominican Republic

• Jean-Claude Duvalier’s rule in Haiti

• Fulgencio Batista’s dictatorship in Cuba

• The rule of the Somoza family in Nicaragua

• The late stages of Pahlavi shahs’s reign in Tran,

• Ferinand Marcos’s presidency after his declaration of martial law in 1972

• Manuel Noriega’s dictatorship in Panama

• Many of personalistic dictatorship in sub-Saharan Africa

Typology of the State

• Developmental state vs. predatory state

• Corporatist state vs. liberal state

• Regulatory state vs. positive state

• Welfare state vs. liberal state

• Strong state vs. fragile (weak, dysfunctional, failed, failing) states

State Capacity & Its Indicators

States in Asia

Fragile States in Asia

Typology of Government

Classification by the way of representation

• No Representation

• Functional representation: Corporatist system

• Territorial representation

• Majorotarian democracy

• Consensus democracy

Classification by the nature of central-local relations

• Unitary system

• Federal system

Classification by the nature of executive-legislature relations

• Parliamentary system

• Presidential systems

• Mixed systems

Typology of Socioeconomic Systems

• Classification by the relationship between the state and economy  

• Capitalism

• Liberal capitalism

• Corporatist capitalism

• Welfare capitalism

• Socialism

• State socialism

• Market socialism

• Classification by the level of economic development

• First world

• Second world

• Third world

The Logic of Comparative Analysis

What Constitutes a Scientific Study?

• Science is based on observable facts

• Science is a generalizing activity.

• It entails a search for regularities

• Science must identify and explain the regularities

• Scientific explanation must be logic

• Scientific findings can be cumulative

• Ideally, science can provide prediction

Basic Scientific Procedures

• Observation: Examine the data (variables)

• Identification of a problem (dependent variables)

• Gauging the causes (independent variables)

• Formulation of hypothesis

• Test the hypothesis

• Acceptance, revision, or rejection of the hypothesis

• Form theories

Variables

• An empirically observable characteristic of some phenomenon that can take on more than one value

• A concept that contains a notion of degree or differentiation

• Types of variables?

• Geographic

• Demographic

• Level of economic development

• Institutions

• Public Policies

• Ideas (political ideology)

• Culture (political culture)

• Legitimacy

Dependent Variable

• What we are trying to explain—they depends on others for its outcome)

• The item which we expect to see changed by other variables

• Dependent variable: the effect or outcome influenced or caused by another variable(s)

• Examples?

• Fragile states

Independent Variables

• Factors that explain the outcome we observe—why does it happen?

• Those factors which you hypothesize cause the change in the dependent variable

• Examples?

• Geographic factors (islands, artificial boundaries)

• Level of economic development

• Ethnic, and linguistic fractionalization

Hypothesis

• A statement of what we believe to be factual

• A statement proposing a relationship between two or more variables

• A proposition intended to serve as a test of the theory

• Correlation or causality?

• Example

• Ethnic fractionalization is a cause of state fragility

How to Test a Hypothesis?

• Make observation about diverse cases

• Isolate the most probable cause

• Provide a theory to explain the relationship between dependent and independent variables

Develop a Simple Causal Model

Theory

• A theory is something we use to explain and predict something that occurs in a given environment

• A theory is a set of logically related propositions that identifies patterns of what we think happens in the world

• Theories are frameworks that academics use to understand the world or to understand relations between the dependent variable and independent variables

• Theories do not have to be complicated!

Which States are Likely to Become Fragile?

Observations

• Large countries are not more likely to become fragile

• Poor countries tend to be fragile

• Multiethnic and multi-linguistic societies are likely to become fragile

• Multi-religious societies are not more likely to become fragile

• Electoral democracies could be fragile (Indonesia and the Philippine)

Hypothesis and Theory

Methodology of Comparative Politics

Basic Methods of Establishing General Empirical Propositions

• The experimental method

• The statistical method (many or large-n)

• The comparative method (few or small-n)

• The case study method (one or single case)

Experimental Method

• The archetype of scientific research is the experimental method, which involves manipulation of variables

• Control all inputs except one in two "test groups" and see if there is a different outcome.

• If there is difference in outcome, assumed to result from this one input or variable.

• Difficult to do in political research

• We can experiment presidential system for 12 years and then try parliamentary system for another 12 years

• Given that we cannot put countries in the lab and see how they behave, we rely on indirect ways of collecting observations and controlling for external interferences.

• Case study

• Comparative study

• Quantitative study

Case Study

Case Study

• Investigator selects a single case and studies it intensively

• Case study method allows the development of expertise in what one studies

• The study of a single case is considered comparative if it uses or develops concepts applicable to other cases, and/or seeks to make larger inferences.

• Ideal to examine “deviant cases” (black swan) to generate hypotheses, to develop new classifications.

Types of Case Studies

• Atheoretical

• Interpretive

• Theory-confirming

• Key deviant case studies: to falsify a hypothesis

• Detailed study to generate hypothesis

Advantages & Disadvantages

• Advantages

• Depth especially for understanding processes

• Grounded in historical developments and context

• Generates further hypotheses for research

• Analyzes deviant cases

• Disadvantages

• Demands extensive fieldwork/language skills/immersion

• Possibility of selection bias

• Inferences based upon one case are less secure.

• Limited generalizability and theory building

Comparative Methods

A Few Cases

• Two or more case studies put together

• Focus on a particular structure or behavior and put it in comparative context

• Look for similarities and differences in different settings

• Compare in one setting but across time

• Area studies

• Compare regions within a single country

• Good for theory building

The Comparative Method

• John Stuart Mill

• The method of agreement: The method of agreement consists in making paired comparisons in order to ascertain what causes changes we observe in the world.

• The method of difference consists in a double application of the method of agreement. This double application springs from the need to make counterfactual comparisons when the explanatory variable is absent in the cases we chose to compare.

Causes of Revolution

• "Revolution is caused by the combination of three factors: 1. High income inequality, 2. conflict within the governing group, 3. defeat in war.“

• Whenever and wherever "1", "2", and "3" are present revolution will occur-- a comparative (general) statement.

Method of Agreement

Method of Difference

Two Approaches of Comparison

• Most similar systems design (MSSD):

• Focuses on differences within a group of cases (region/area)

• Compares like-with-like to ‘control’ for shared factors such as culture, history, social or economic structure

• Seeks to identify key features that are different among similar countries, which account for the observed political outcome.

• Most different systems design (MDSD):

• Focuses on similarities in outcome among cases that differ from each other in most aspects (culture, history, social or economic structure)

• Seek to identify key features that are similar among different countries, which account for the observed political outcome.

An Example

• In Problems of Democratic Transition and Consolidation, Linz & Stepan use MSSD to examine democratic consolidation within regions (South America, Southern Europe, and Eastern Europe), and then use MDSD to compare democratic consolidation across regions.

Advantages & Disadvantages of Comparative Method

• Advantages

• Combines both depth and breadth

• Identifies variations within culturally-similar regional area or similarities across culturally different regional area

• Builds middle-level theories

• Disadvantages

• Higher demands for contextual fieldwork and language skills

• Too many independent variables and too few nations

• Limited generalizations outside of area/region

Quantitative Method

Why Quantitative Method?

• The comparative method is ill prepared to deal with a large number of cases and to tackle the issue of plural or conjunctural causation.

• Moreover, there are many instances where the information we collected is quantitative, not qualitative.

• The majority of studies that compare many countries use quantitative methods.

• Because of some of these and other good reasons, the statistical method is also used in comparative politics.

Feature of Quantitative Method

• “Variable-oriented”: examine the relationship between variables at a global level of analysis.

• The more the cases, the stronger the inferences (the more “experimental”)

• Helps to identify “deviant” cases.

• The qualitative study of many cases is difficult

• Data requirement

• Statistical method

An Example

• The higher the level of economic development, the greater the likelihood of democratic development

• How to measure democracy?

• What are the indicators of economic development?

• Need to make concepts into measurable “variables”

• Collect information on a number of variables related to those factors for 100 or more cases

• Use multivariate regression to see which factors explain most of the variation in your dependent variable, the level of democracy.

• Look for statistical relationships among the variables that support or challenge your assumption about the relationship among them.

A Research Design I

• Question:

• What are the electoral institutions that have contributed to the advancement of women in politics the most?

• Task: Compare numbers: the relative number of women to men in political assemblies around the world.

• Data collection:

• Collection of information about the proportion of women in parliaments and congresses around the world.

• Tested theory:

• A prominent theory in the field was that the electoral institution that have a large effect on the proportion of women in parliaments was the electoral formula: how votes are translated into seats.

A Research Design II

• Procedure:

• Sort out countries depending on their electoral formula:

• Either plurality or proportional representation

• Then compared the mean value of the proportion of women in each group.

• Finding:

• The mean of proportions in countries with plurality is 11.92% and in PR or semi-PR is 18.93%.

• Both values are statistically significant.

A Research Design III

• Complication

• The value of the proportion of women in all countries that have quotas was 2 standards deviations above the mean!

• Exclude the countries with quotas from the sample and compared the mean of their proportions again:

• New finding:

• The difference of their values were not statistically significant anymore!

Source: Gomez Albarello, Juan Gabriel. 2002. “What’s Wrong with This Picture”, available at http://artsci.wustl.edu/%7Ejggomeza/papers/picture.pdf

Advantages & Disadvantages of Quantitative Method

• Advantages

• Comprehensive generalizations

• Identifies outliers and deviant cases

• Build and test general theories

• Disadvantages?

• Limited availability of data

• Valid cross-cultural measures

• Lack of insights into ‘black-box’ policy processes

• Skills needed to analyze data

• Too abstract and too far removed from context and processes to facilitate policy interventions

What would You Try?

• Qualitative or quantitative? Why?

• In what circumstances would you choose many cases, a few cases, or single case studies?

• Why?

Common Problems
What would you do to overcome these difficulties?

• Too many variables, too few countries

• Increase units of observation? Eg parties, states, time periods

• Equivalence across cultures/societies

• Universal or relative concepts

• Democracy? Gender equality?

• Functionally-equivalent but different indicators

• Income quintiles

• Selection bias

• Few countries/cases may confirm propositions but limited inferences elsewhere

Common Problems

• Value bias

• Can “IS” and “OUGHT” questions be separated?

• Empirical and evaluative statements

• Explicit not implicit assumptions

• Spuriousness

• Omitted variables

• Correlation of X>Y, but Z leads to both X and Y

• Ecological fallacy

• Multilevel analysis from aggregate to individual (statistical)

• Average income in HK is higher than that in the mainland

• Individualist fallacy

• From individual to aggregate (one or few cases)

What is Scientific about Political Science?

• Subject matter's behavior is completely determined – not the case

• Subject matter's behavior is clearly and completely understood – not the case

• Subject can be summarized in terms of general laws of behavior – not the case

• Political Science is scientific in the sense that the study of the subject is pursued in a scientific manner, with reliance on evidence, testing of hypotheses

Is it Possible to Study Politics Scientifically?

• Complexity of individual human beings

• Cultural diversity: Too different to be generalized

• Ability of humans to learn and change behavior

• Limited capacity for controlled experiment

• Difficulties in quantifying and measuring social phenomena

• Impact of personal and social values on research

• So it is impossible to develop a real "political science,” but we should try to study politics in a scientific way