Democracy and Autocratization 3

Final Project Essentials: Interaction Terms

Author
Affiliation

Jeremy Springman

University of Pennsylvania

Logistics

Assignments

  • Last Thursday
    • I forgot the quiz; sorry
  • Today
  • Thursday (3/14)
    • Distribute Data Assignment 1
    • Two readings on Accountability and Information

Agenda


  1. Garcia-Ponce and Wantchekon
  2. Hassan
  3. Lebas and Young
  4. Final Project Essentials: Interaction terms and Fixed Effects

Garcia-Ponce and Wantchekon

Why countries Democratize

  • Correlation between income and democracy
    • Modernization theory: wealth, education, and urbanization are associated with democratization
  • Why might this be true?

In Africa, democratization after the Cold War was shaped by the level of development (urbanization) at the start of countries’ anti-colonial movements

Context


  • Anti-colonial movements chose between two strategies:
    • Rural insurgency
    • Urban protest
  • Democratization was delayed by the Cold War
  1. What shaped opposition movements’ choice between strategies?
  2. What delayed democratization during the Cold War?

Theory

Why does opposition strategy affect democratization decades later?

  • Inclusion
  • Organization

Mechanisms:

  • Institutions
  • Behaviors

Findings

Threats to inference

  • Things that affect whether or not opposition uses insurgency vs protest might also affect current levels of democracy

Attaching package: 'ggdag'
The following object is masked from 'package:stats':

    filter

Identification

  • Covariates
  • Sensitivity analysis
  • Difference-in-Differences
    • Leverages changes within countries over time
    • Before the end of the cold war, groups look similar
    • After the end of the cold war, groups trend away from each other
  • Instrumental Variables
    • Leverages ‘plausibly exogeneous’ variation in terrain
    • Terrain predicts strategy choice

Instrumental Variable

Mechanisms: Institutions

We do not seen an impact of Post-WWII Opposition Strategy on Post-1990 Institutions

Mechanisms: Behavior

Policy Implications


  • Historical social movements help understand the link between structural conditions and democratic development
  • Non-violent collective action appears critical for the establishment of lasting democracy

Hassan et al. (2022)

Political Control


Strategies of Control

  • Violence
  • Indoctrination
  • Coercive distribution
  • Infiltration

Lebas and Young (2023)

Repression and Backlash


Moments of uncertainty about political opportunity

  • Citizens uncertain about relative strength of regime and opposition
    • Risk of repression (cost)
    • Size of protests (benefit)

Research Questions


  • How are are individual decisions about protest affected by dissent and repression of others
  • What are the mechanisms?
    • Information on costs and benefits
    • Affective polarization
    • Emotions

Context

Political moment

  • Ruling party leadership transition
  • Contentious election
  • Economic crisis

Treatment

  • Exposure to local dissent
  • Exposure to local repression

Measurement

Design

  • Difference-in-difference estimator can partials out time-invariant differences across participants
  • Assumes units are on similar trajectories prior to the treatment
  • Bias could be introduced if individual shocks co-occur with treatment

Difference-in-Differences

Show code 
Year = c(0,1,2,3)
Outcome = c(NA, 1.3, 1.5, 1.7,
            1, 1.2, 1.4,1.6,
            1.1, 1.3, 1.7, 1.9)
Treatment = c("Comparison","Comparison","Comparison","Comparison",
              "Control", "Control","Control","Control",
              "Treatment", "Treatment", "Treatment", "Treatment")

dat = data.frame(Year, Outcome, Treatment)
dat$Treatment = factor(dat$Treatment, levels = c("Treatment", "Comparison", "Control"))

ggplot(data = dat, aes(x = Year, y = Outcome,  color = Treatment)) +
  geom_line(aes(linetype=Treatment),size=2) +
  geom_point(size = 6) +
  ylim(0.8, 2.2) +
  scale_linetype_manual(values=c("solid", "dotted", "solid")) +
  scale_color_manual(values = c("blue", "black", "red"  ) ) +
  theme(legend.position = c(0.8, 0.2), text = element_text(size=20),
        legend.title=element_blank())

Findings

Findings

  • Exposure to dissent and repression increases willingness of non-supporters to protest
  • Different mechanisms drive diffusion and backlash
    • Observing dissent affects beliefs about dissent (but not perceptions of risk)
    • Observing repression affects emotions (but also risk)
  • Local events drive diffusion and backlash; media events do not

Final Project Essentials

Interaction Terms

What is an interaction term?

  • Simple linear models assume that the effect of predictors is independent of other factors
  • Interaction terms allow us to estimate the difference in the slope of a predictor across unit characteristics

What are they used for?

  • Heterogeneous effects
  • Difference-in-differences

Interaction Terms

How do they do work?

\[ Y_i = \alpha + \beta_1 X_{i1} + \beta_2 X_{i2} + \beta_3 (X_{i1}*X_{i2}) + \epsilon_i \]

Example: Continuous outcome with two binary predictors

  • \(\alpha\): Intercept when \(X_{i1}\) and \(X_{i2}\) are 0
  • \(\beta_1\) Slope when \(X_{i2} = 0\)
  • \(\beta_2\) Difference in \(\alpha\) between \(X_{i2}=0\) and \(X_{i2}=1\)
  • \(\beta_3\) Difference in \(\beta_1\) between \(X_{i2}=0\) and \(X_{i2}=1\)