Elections, voting and public opinion
Compare election systems, explain how people vote and assess how campaigns, demographics and polling shape political decisions.
Exam scope and assessment notes
This is a standalone AS Level qualification for centres in the United States of America. The syllabus is for June examinations in 2028–2030, uses digital exams and is not available in every administrative zone. Confirm local availability and entry with Cambridge and the exam centre. This original StudyForA course maps teaching and staged practice to all 15 syllabus sections, including separate race, gender and disability-rights lessons under section 4.4. Full digital examination practice, complete analysis of all mandatory documents and landmark cases against official resources, and subject-specialist review remain outstanding. It is not Cambridge-authored material and does not provide official marks.
Before you begin
- Know that the United States has national and state governments
- Distinguish a constitutional rule from a political claim about its effects
Work on paper. Try each question before revealing help, and explain why your method works.
Quick readiness check
Which pair gives the strongest basis for evaluating a political claim?
What you will learn
- Explain Electoral College allocation, winner-take-all and proportional approaches, House and Senate elections, redistricting and gerrymandering, runoffs, direct democracy and campaign-finance debates
- Apply rational-choice, retrospective and prospective voting models; assess demographic patterns, critical elections, realignment and dealignment, campaign effects and turnout
- Explain how public opinion and political socialisation relate to government and campaigns, and evaluate poll types, samples, wording, nonresponse and other sources of error
Election rules turn votes into offices in different ways
The Electoral College selects the President through state-allocated electors. Most states award all their electors to the statewide popular-vote winner; Maine and Nebraska allocate some electors by congressional district. A proportional approach is a distinct way a state could allocate electors. Do not assume the national popular-vote total directly determines the Electoral College result.
House members face election every two years from districts; redistricting redraws district boundaries after population data, while gerrymandering uses boundaries to advantage or disadvantage groups or parties. Baker v. Carr (1962) made federal judicial review available for certain legislative-apportionment claims. Runoffs require another election when the first round does not meet the applicable rule. Senators serve six-year terms, with roughly one-third of seats up every two years, and are elected statewide.
Referendums let voters decide a policy question put to them; initiatives allow voters to propose a measure under a state's rules. Campaign-finance regulation addresses how money is raised or spent in elections. The Bipartisan Campaign Reform Act (McCain–Feingold, 2002) and Citizens United v. FEC (2010) are important parts of the debate; distinguish rules for contributions from rules about independent political spending rather than saying all campaign finance limits disappeared.
Voting models and campaigns explain possible influences, not certain motives
Rational-choice voting treats voters as weighing expected benefits and costs. Retrospective voting evaluates an incumbent's past performance; prospective voting focuses on expected future policies or performance. These are analytical models, and a real voter may use more than one.
Voting patterns can vary with race, ethnicity, age, generation, gender, urban or rural residence, religion, education and class. A demographic association is not an explanation of every individual's vote. A critical election produces a lasting shift in party coalitions or voting patterns; realignment describes that shift and dealignment a weakening of durable party attachment.
Campaigns can influence turnout through rallies, candidate personality, debates and contact with voters, but they operate alongside party identification, policy preferences, resources and the wider context. A careful answer uses evidence to connect a campaign feature to a possible effect instead of claiming one event caused an election result by itself.
Public opinion is measured imperfectly and can affect government
Public opinion can shape branch priorities, legislation, campaigns and party platforms. It may be measured through polls, surveys and election results, but each measure answers a different question: election results show choices among candidates or options on the ballot, while a poll asks a sample about stated views at a particular time. Benchmark polls establish a starting estimate, tracking polls repeat questions to monitor change, and exit polls survey voters after they cast a ballot; each serves a different purpose and none is a perfect census of opinion.
Traditional political socialisation includes family, community, school and religious institutions; modern channels include digital media and online networks. Sources can reinforce or challenge prior beliefs, so exposure alone does not prove persuasion.
Evaluate a poll by asking who was sampled, how respondents were selected, whether the sample represents the target population, when the poll was conducted, how questions were worded, which response options were offered, the response rate and whether likely voters were screened. Sampling error is only one source of uncertainty: nonresponse, question order, social desirability and weighting assumptions can also affect results. A margin of sampling error does not fix a biased sample or guarantee the poll's accuracy.
Sources for this lesson
Pause and explain
Why can a poll's margin of sampling error fail to reveal whether its result generalises?
Worked example
A poll of 500 people who clicked a party website says 70% of likely voters support its candidate. What can the poll establish, and what limits the conclusion?
Show the worked solution
- Identify how respondents entered the sample.
- Compare the sample with the population named in the claim.
- Separate a reported result from evidence about likely voters in the wider electorate.
Answer The poll describes the 500 website respondents, but a self-selected group of party-site visitors may differ from all likely voters. Unless the sampling design represents that wider population, the 70% figure cannot be generalised to the electorate. A large response count does not remove selection bias, and the question wording, timing, response rate and likely-voter screen also matter.
Try a topic test
Choose Easy, Normal or Difficult. Save your answers and open worked review after finishing.
Test this topic →From guided practice to a new situation
Write your answer and reasoning first. Use a hint only if you are stuck. The model solution is for self-checking; your written work is not automatically marked.
Guided: A presidential candidate wins a state's popular vote by 51%. The state awards all of its electors to the winner. Explain the allocation model and one conclusion that cannot be drawn from this result alone.
Give me a hint
Distinguish the statewide result from the national popular vote and from a proportional allocation.
Compare my reasoning
- The state uses a winner-take-all allocation in this example.
- The candidate receives the state's electors under its rule.
- The state result does not establish who won the national popular vote or the Electoral College overall.
The state uses winner-take-all allocation: the statewide winner receives all the state's electoral votes under the stated rule. This single state result does not show who won the national popular vote or whether either candidate reached an Electoral College majority; other states' allocations also matter.
Look for these in your work
- Names the allocation method
- Applies the state's rule
- Limits the conclusion to the evidence
Independent: An incumbent party loses support after a recession. Explain how retrospective and prospective voting could each describe a voter's decision, then give one reason the evidence would not prove either motive.
Give me a hint
Look backward for performance assessment and forward for expected policy.
Compare my reasoning
- Retrospective voting evaluates the incumbent's past record, such as the recession response.
- Prospective voting compares expected future policies or performance.
- The election result aggregates choices and cannot by itself reveal each voter's reasoning.
A voter who judges the incumbent by the recession response is using retrospective reasoning. A voter who chooses the challenger because they expect better future economic policy is using prospective reasoning. The overall vote totals do not reveal individual motives; surveys, interviews or other evidence would be needed, and voters can combine several reasons.
Look for these in your work
- Distinguishes the two models
- Connects them to the example
- Avoids inferring individual motives from aggregate results
Transfer: A campaign cites a poll with a small reported margin of error, but respondents opted in through social media and the question names only the campaign's preferred policy. Evaluate whether the result shows majority support.
Give me a hint
Check sample selection and question wording before interpreting the margin of error.
Compare my reasoning
- An opt-in sample may not represent the target population.
- Leading or one-sided wording can shape responses and omits alternatives.
- A small sampling margin does not repair these design problems; withhold the majority claim until a sounder poll is available.
The poll does not establish majority support among the public. Opt-in respondents may be systematically different from the target population, and a question naming only the campaign's preferred policy may prime or constrain responses. A small margin of sampling error does not address selection or wording bias. A representative sample and neutrally worded question with balanced response options would provide stronger evidence.
Look for these in your work
- Identifies selection bias
- Identifies question wording bias
- Explains why a small margin is insufficient
Common mistakes
- Equating the national popular vote with the Electoral College result
- Treating every state as winner-take-all or confusing Electoral College allocation with a direct proportional national vote
- Calling every district change gerrymandering without evidence of partisan or group advantage
- Treating demographic correlations as fixed traits of individuals
- Claiming a campaign event or poll caused a result without a counterfactual
- Treating margin of sampling error as a measure of every source of poll error
What can you explain now?
Why can a poll's margin of sampling error fail to reveal whether its result generalises?
Compare with the explanation
The sample may be biased or unrepresentative, and non-sampling errors are not captured by that margin
A margin of sampling error describes uncertainty from random sampling under assumptions; it does not correct selection bias, nonresponse, wording effects or a poor target population.
After trying it yourself, choose your next review. This is your self-assessment.
Your review choice appears on Today. Sign in to sync it across devices.
Make it stick
Why can a poll's margin of sampling error fail to reveal whether its result generalises?
If you needed the solution, close it and solve the problem again from a blank page. A correct answer is stronger when you can explain the reason for each step.