Pre-class Brief: Population Reading

Read, Extract, Arrive Ready (≈15 minutes prep)

0) What to skim first (5 minutes total)

  1. Abstract — highlight the causal claim or main descriptive finding (one sentence).
  2. Key visuals — glance at Figure 1 and Figure 2 (e.g., fertility / mortality / migration / age structure).
  3. Identification or Data section openers — first paragraph under Methods/Data.
  4. Conclusion first paragraph — how authors frame policy relevance.

1) Bring this to class (10–15 lines total)

  • 1-line takeaway (your words):
    > “In [country/region], [outcome] changed by [direction/size], mainly due to [mechanism].”
  • 2 lines of evidence: cite visuals precisely
    • Fig. [X] shows ______ (trend/level/comparison).
    • Table [Y] shows ______ (effect size / confidence).
  • 2 lines on method/data quality:
    • Identification hinges on ______ (DiD / IV / panel FE / decomposition / cohort analysis).
    • Biggest vulnerability is ______ (parallel trends / migration selectivity / measurement error).
  • 3 questions for discussion (concrete):
    1. If [assumption] is violated (e.g., selective out-migration), how would the estimate move (bias ↑/↓)?
    2. Would results hold for [other region/age/period]? Why?
    3. What alternative indicator (e.g., TFR vs. cohort fertility, adult mortality vs. life expectancy) might flip the story?
  • 1 replication/visual idea (optional bonus):
    • Re-plot Figure [X] with [log scale / per-capita / age-standardized]; or add a dependency-ratio panel.

2) Lanes we’ll use for the 30-minute discussion

Pick two to focus on.

A. Causal Design / Comparability
- What makes groups/periods comparable?
- Where might parallel trends or instrument exogeneity fail?
- Which figure/table best supports the design?

B. External Validity (Generalization)
- Conditions where the finding would reverse (institution, business cycle, urban/rural, cohort).
- Which subgroup (e.g., 65+, prime-age, in-migrants) is likely to drive the aggregate?

C. Measurement & Data
- Definitions used (TFR, NMR, life expectancy, net migration, dependency ratio).
- Likely measurement error and its bias direction (attenuation/upward).
- Cross-check with administrative vs. survey vs. vital statistics.

D. Mechanisms / Narrative
- Claimed channels (e.g., labor demand shocks → migration → age structure → wages/care load).
- One testable implication you’d add (leading indicators, cohort splits, placebo period).

E. Policy Implications
- One implementable action (who/when/how much).
- Trade-offs (short-run vs. long-run, regional spillovers).

3) Roles in class (rotate every 5 minutes)

  • Timekeeper — keeps us on the lane timeline.
  • Scribe — writes the 1-line conclusion per lane.
  • Devil’s Advocate — must produce one plausible counter-story.

4) Quick citation note (if you cite)

Use in-text keys like [@key] and include references at the end:

## References
::: {#refs}
:::

(If you know the exact visuals already, pre-fill “Figure [X] / Table [Y]” above to reduce search time in class.)