Chapter 22 Exercises

These exercises move from data retrieval to analysis to application. Several connect to your Democracy Audit project on your home congressional district. All require you to engage with real data, not just summarize the chapter.

Exercise 1: Pull turnout data for your district from the Secretary of State

Go to your state's Secretary of State website and find official certified results for the most recent presidential election (2024) in your congressional district.

Find or calculate:

  1. Total ballots cast in your district in 2024.
  2. Total registered voters in your district as of the November registration deadline.
  3. Turnout as a percentage of registered voters.
  4. The estimated VEP for your district (you may need to use Census ACS data plus an estimate of non-citizen population; the University of Florida Election Lab has state-level VEP estimates).
  5. Turnout as a percentage of VEP.

Report all five numbers. Then write 250 words on what these numbers mean. Was your district's turnout above or below the 2024 national average (~64% VEP)? What might explain the difference?

Exercise 2: Compare your district's 2020 and 2024 demographic shifts

Using crosstab data from AP-NORC VoteCast (or your state's exit-poll data, or Pew validated-voter analyses if available for your state), compare how the following groups voted in your state in 2020 and 2024:

  • White college-educated voters
  • White non-college voters
  • Black voters
  • Hispanic voters (if your state has a measurable Hispanic vote)
  • Voters under 30
  • Voters over 65
  • Women
  • Men

For each group, report the Democratic vote share in 2020 and 2024, and the swing (change in points). Identify the two groups in your state that shifted most, in either direction.

Then write 400 words placing your state's shifts in the context of national 2020→2024 trends. Was your state's pattern typical, or were there state-specific factors? (If you live in a battleground state, this analysis is the heart of post-election commentary.)

Exercise 3: Calculate the registration gap

Using Census Bureau data (the Voting and Registration Tables, published every two years), find the registration rate for your congressional district or state — the percentage of voting-eligible adults who are registered.

Then break this down by:

  • Age (18-29, 30-44, 45-64, 65+)
  • Race/ethnicity (white, Black, Hispanic, Asian American)
  • Education (college, some college, high school only, less than high school)
  • Income quartile

Identify the demographic group in your district with the lowest registration rate. Estimate how many additional voters would be added to the rolls if their registration rate matched the highest-registering group.

Write 300 words on what this gap means in your district. Are there visible mobilization efforts targeting this group? What might explain why registration efforts haven't closed the gap?

Exercise 4: Steel-man both sides of voter ID

Using the chapter's empirical sources (Hajnal/Lajevardi/Nielson, Hopkins/Hill, Cantoni and Pons, Anoll et al.), write the strongest version of two arguments:

Argument A: A 600-word brief making the strongest empirically supported case that strict voter-ID laws should be opposed because they reduce turnout disparately among low-income and minority voters. Cite specific studies; address counter-evidence honestly.

Argument B: A 600-word brief making the strongest empirically supported case that strict voter-ID laws should be supported because they protect election integrity at minimal turnout cost. Cite specific studies; address counter-evidence honestly.

Submit both. The grade is on quality of argumentation and faithfulness to the cited evidence, not on which "side" you actually agree with. (For most students, writing the side they disagree with is harder and the more useful exercise. Do that side first.)

Exercise 5: Democracy Audit on turnout in your district vs. national average

This exercise extends your running Democracy Audit project to turnout data.

Compile a one-page audit memo on your home congressional district that answers:

  1. What is the district's turnout (presidential and midterm) in the last two cycles vs. the national average?
  2. Which demographic groups in the district vote at the highest and lowest rates?
  3. What fraction of the district's eligible adults are unregistered?
  4. What is the partisan composition of the registered electorate vs. the actual voting electorate? (Is there a turnout gap that favors one party?)
  5. How does the registered voter base of the district compare demographically to its eligible-but-unregistered population? (This tells you the potential for the electorate to change if registration efforts succeed.)

Memo length: 600-800 words plus a data appendix (charts, tables — at least three of each). Cite all sources.

This memo is the foundation for the participation-and-turnout section of your final Democracy Audit at the end of the course.

Exercise 6: Map the partisan shifts among Hispanic voters in your state

Using validated voter-file data (Catalist, TargetSmart, Pew) or VoteCast crosstabs, examine how Hispanic voters in your state have voted in:

  • 2012
  • 2016
  • 2020
  • 2024

If your state's Hispanic vote sample is too small for reliable estimates, use the most-similar nearby state (Texas, Florida, Arizona, Nevada, California, Illinois, New Jersey, New York are all good candidates).

Identify whether the shift in your state was:

  • Larger than the national shift (~12-15 points toward Republicans, 2012-2024)
  • About the national shift
  • Smaller than the national shift

Then write 500 words exploring why your state's Hispanic vote moved the way it did. Consider: country-of-origin composition (Cuban, Mexican, Puerto Rican, Central American, etc.), age structure, religious composition, urban/rural distribution, immigration salience as an issue, candidate-specific effects. What evidence would you cite?

Exercise 7: Test the resource model with a household interview

Identify two adults in your network — one a regular voter, one a non-voter or irregular voter. (Not yourself.) Interview each for 15-20 minutes, asking questions that probe the three resources from the Verba/Schlozman/Brady model:

  • Time: What's their work schedule? Childcare obligations? Commute? When do they have time to engage with politics?
  • Money: How much would taking time off to vote cost them in lost wages or childcare? Have they ever skipped voting because of cost?
  • Civic skills: How do they get political information? Do they feel confident reading a ballot? Do they know who their elected officials are?

Write a 700-word reflection comparing the two interviews. Did the resource model predict the difference between them? What factors not captured by the resource model also seemed to matter (habit, family influence, mobilization, motivation)?

Anonymize names. This is a qualitative exercise; the goal is to make the abstractions of the chapter concrete by attaching them to actual people you know.

Exercise 8: Apply the rational-choice paradox to a non-voter you know

Identify (carefully, anonymously) one adult in your family or social network who is consistently a non-voter. Without confronting them, reflect on which of the rational-choice paradox's proposed solutions would or would not apply to them.

Write a 500-word reflection covering:

  1. Civic-duty value. Does this person feel that voting is something they "ought to" do, but don't? Or do they not feel a duty to vote at all? How might this be different from someone in their demographic group who does vote?
  2. Expressive value. Are there other ways this person expresses their political identity (yard signs, conversations, social media, donations to causes, volunteering)? Or have they checked out of political expression entirely?
  3. Group-conformity value. Do their family and friends vote? Are they swimming against a turnout current in their immediate network, or with one?
  4. Resource constraints. What time, money, or civic-skill barriers do they face? Are these the binding constraints, or are there other barriers (motivation, distrust of the system) that resources don't capture?

Identify the one factor that you think most explains their non-voting. Then identify the one realistic intervention (a friend's mobilization, an automatic registration system, a more compelling candidate, a major personal life change) that might most plausibly change their behavior.

This exercise is not a substitute for actually talking to them — but it is a good test of whether the chapter's frameworks help you understand the people in your life or only the people in survey datasets.

Exercise 9: Project — Design a non-partisan voter registration campaign

Pick a demographic group in your district with a low registration rate (use Exercise 3's findings). Design a non-partisan voter-registration campaign aimed at that group, in 800-1,000 words.

Your design should include:

  1. Target group definition. Who, exactly? (Be specific: "18-24-year-olds at the community college" is better than "young voters.")
  2. Channel mix. How will you reach them? (In-person, social media, direct mail, partnerships with employers/churches/schools — and why those specifically?)
  3. Message strategy. Why should they register? (What's the message that lands with this group, given the chapter's discussion of mobilization research?)
  4. Cost estimate. What's your budget? Cost per registration?
  5. Measurement plan. How will you know if the campaign worked?
  6. Honest assessment of barriers. What might cause this campaign to fail? What are the limits of registration efforts in reaching habitual non-voters?

Submit as a memo. The most successful designs in past iterations of this exercise have been the ones that took the difficulty of habitual non-voters most seriously.

Exercise 10: Compare two states' voting administration

Pick two states with substantially different voting administration. Suggested pairs:

  • Oregon (all-mail voting, automatic registration) vs. Texas (no-excuse absentee restricted, no online registration in most cases)
  • Minnesota (same-day registration, no-excuse absentee) vs. New Hampshire (same-day registration, restricted absentee)
  • Colorado (all-mail voting, AVR) vs. Wisconsin (some early voting, photo ID required)

For each state, document:

  1. Registration architecture. Online registration? Same-day? Automatic? Pre-registration of 16-17-year-olds?
  2. Voter ID rules. What ID is required? Are alternatives accepted?
  3. Mail/absentee voting. Excuse required? Universal? Drop boxes? Ballot collection?
  4. Early voting. How many days? Where?
  5. Polling-place administration. How many polling places per registered voter? What are typical wait times?
  6. VEP turnout in 2024. What was it?

Compare the two states' turnout. Then write 600 words on whether the difference in turnout is plausibly explained by the difference in administration. (Be careful: turnout is also driven by demographics, partisan composition, and election-specific stakes; isolate the administration effect from these confounders.)

What does the comparison suggest about the size of the policy lever available through administrative reform? Is administrative architecture a small effect, a medium effect, or a large effect on turnout? Defend your answer with specific evidence.

Exercise 11: The retrospective-voting test

Pick a recent congressional or gubernatorial election in your state. Identify the incumbent and the challenger. Compile information on:

  1. Objective conditions during the incumbent's term. Economic indicators, major events, policy outcomes attributable to the incumbent.
  2. The vote outcome. Did the incumbent win or lose? By what margin?
  3. The retrospective evaluation in exit polls or VoteCast crosstabs. How did voters say they felt about the incumbent's job performance?

Apply the Achen-Bartels and Healy-Lenz frameworks. Was this election a case of:

  • Rational performance evaluation (voters punished or rewarded based on actual conditions)?
  • Arbitrary blame-assignment (voters punished or rewarded based on partisan filter or events outside the incumbent's control)?
  • Something in between?

Write 500 words defending your interpretation. Cite the exit-poll or VoteCast data. Identify what evidence would change your mind in either direction.

This exercise builds the analytical skill of distinguishing between what voters say they are doing and what they are actually doing — a critical skill for reading any post-election commentary.

A note on the exercises

These exercises move from individual-voter analysis to district-level analysis to system-level analysis. Together, they should give you both the empirical literacy and the practical familiarity to engage seriously with American voting behavior. None require math beyond percentages; all require careful sourcing.

If you complete five of these eleven exercises during the chapter unit, you will have a substantially better grasp of American electoral behavior than most graduates of an introductory political science course. If you complete all eleven, you will have written something close to a senior thesis chapter on your own electoral environment.

The Democracy Audit project that runs through this textbook benefits especially from Exercises 1, 3, 5, and 10. By the end of the course, you will have compiled a detailed audit of your own congressional district's electoral behavior, structural barriers, and reform opportunities. Begin building that file now.