Governing Immigration by the Rules

Legal Procedure, Equal Treatment, and Public Support Across Four Countries

Steven Denney

Leiden University

American Political Science Association Annual Meeting · Boston · September 2026
Public Attitudes towards Immigration Policies · Division 52
Funded by the Dutch Research Council (NWO)

Do citizens judge immigration policy only by what it does and whom it affects? Or also by how the government makes and applies it?

Most of what we know is about the first two.

Recent backlash is also about how governments handled immigration, not only what they decided.

United States

Presidents set deportation and border policy by executive order, from DACA in 2012 to the 2025 deportation orders. The court fights were over the authority to act and due process.

the dispute is who may decide, and how
Tichenor 2002; Christenson & Kriner 2017
Germany

In 2015 the government admitted asylum seekers by executive decision, bypassing the EU's Dublin rules. The backlash asked whether it had acted lawfully or lost control (Kontrollverlust).

was the law followed, is the government in control
Joppke 1999; Solodoch 2021
Yet most explanations of immigration attitudes ask who gets in and at what cost, not how the government acts.

Existing explanations, from groups to order

Immigrants are judged by whether they follow the rules.

The immigrant
A newcomer
→
Their conduct
Follows the law, works, contributes
→
Public judgment
Deserving
The state & government
Deciding authority
→
Its conduct
Follow established rules & norms?
→
Public judgment
?
Citizens expect immigrants to follow the rules. Do they expect the same from governments?

For governments, it is about procedural legitimacy.

The state & government
Deciding authority
→
Its conduct
Decides through fair and proper procedures
→
Public judgment
Accepted, even by those who dislike the decision
This study tests whether governments are held to a procedural legitimacy standard in immigration policy.

Must the government follow the rules too?

What we know
Immigrants are judged by whether they follow the rules.
Aviña et al. 2026; Levy & Wright 2020
What this study tests
Are governments?
Design
Pre-registered conjoint
Countries
4
Respondents
≈10,000

Fielded in the United States, Germany, South Korea & Singapore in August 2026 using quota-matched online panels.

A single-profile factorial conjoint experiment

  • Decision process lawful baseline · executive overreach · external imposition · procedural neglect4
  • Application even-handed · discretionary · unequal3
  • Policy area skilled labor · asylum · family reunification3
  • Policy direction expand · restrict2
  • Immigrant composition five country-specific origin bundles5
360profiles, fully crossed · ≈100,000 policy evaluations
10 policy cards per respondent + first task repeated (Clayton et al. 2025)
Outcome
Support or oppose the policy.
Then
“Regardless of how I personally feel about this policy, I would accept the government's decision as properly made.”
Estimation
AMCEs, respondent fixed effects; errors clustered by respondent.

Four countries, one design

United States n = 2,498

Established democracy. Immigration highly salient and rigidly partisan; executive power at the center of the fight.

Germany n = 2,502

Established democracy. Immigration highly salient; legal-procedural and multilevel, with the EU as a real outside authority.

South Korea n = 2,500

Newer democracy. Immigration growing but less politically salient; politics recently turned on the limits of executive power.

Singapore n = 2,500

Electoral autocracy. Managed, technocratic migration regime; legitimacy rests on administrative competence.

Regime type and salience vary. Design localizes for institutions and context.
Ogura, Miwa & Kobayashi 2026

One policy card

Policy
The government announced it will reduce the number of skilled worker visas issued each year to prioritize domestic workers.
Affected groups
This policy would primarily affect immigrants from India, Pakistan, and Bangladesh.
Decision process
The decision was made by the President through an executive order, bypassing congressional approval and judicial review. The White House argued swift action was needed.
Application
Administration officials signaled the policy would be applied unevenly, with some applicants exempted and others in comparable situations strictly enforced, with no stated basis.
Do you support or oppose this policy?
Support
Oppose
Shown after the respondent answers
“Regardless of how I personally feel about this policy, I would accept the government's decision as properly made.”
Strongly agreeAgreeNeitherDisagreeStrongly disagree

Bypass the rules, lose support.

United States −10 −20 0 Germany −10 −20 0 South Korea −10 −20 0 Singapore −10 −20 0 DECISION PROCESS Lawful legislative process (reference) Executive overreach −13.4 −8.5 −5.4 −6.6 Procedural neglect −15.3 −7.7 −11.6 −12.7 External imposition −1.3 −0.9 −0.2 −1.2 APPLICATION Even-handed (reference) Discretionary −7.2 −5.9 −11.0 −3.7 Unequal application −19.3 −9.1 −14.2 −10.3

The outcome matters. A lot.

Share supporting the policy when it expands admissions and when it restricts them.

policy expands admissions policy restricts admissions 20% 40% 60% 80% 0 share supporting the policy United States 57% 41% the only country where restriction is less popular Germany 37% 56% South Korea 43% 60% Singapore 56% 68%

And yet the process penalty does not depend on it, pooled and in three of four countries.

Change in the penalty for bypassing the rules when the policy restricts rather than expands admissions.

pre-registered bound for “about the same”, ±5 points United States +4.5 +6.7 Germany +0.1 −0.8 South Korea +1.4 −2.0 Singapore +2.5 +1.9 −5 +5 0 ← larger penalty when restricting smaller penalty when restricting → executive overreach procedural neglect

Opposed the policy. Still accepted the decision, unless the procedure was violated.

United States 28% still accept a lawful decision −5 −10 0 change in that share, percentage points under executive overreach −10.1 under procedural neglect −10.9 Germany 21% still accept a lawful decision −5 −10 0 under executive overreach −8.5 under procedural neglect −8.2 South Korea 17% still accept a lawful decision −5 −10 0 under executive overreach −4.9 under procedural neglect −5.3 Singapore 20% still accept a lawful decision −5 −10 0 under executive overreach n.s. under procedural neglect −2.9

Reactions to immigration policy are not only reactions to immigrants or outcomes. They are reactions to how governments exercise authority.

Steven Denney · Leiden University · s.c.denney@hum.leidenuniv.nl
These slides
scdenney.github.io/assets/slides/governing-immigration/

How a policy is governed outweighs which group it affects.

United States GOVERNANCE GROUP Immigrant composition1.1not distinguishable from zero Germany GOVERNANCE GROUP Immigrant composition1.3not distinguishable from zero South Korea GOVERNANCE GROUP Immigrant composition1.5not distinguishable from zero Singapore GOVERNANCE GROUP Immigrant composition3.4 Decision process15.1Application19.0Decision process7.6Application9.2Decision process12.1Application13.8Decision process13.0Application10.3
Read narrowly

Both governance penalties exceed the registered composition contrast in all four countries.

Restriction is the popular outcome in every bloc but two.

Support for a policy that restricts rather than expands admissions, percentage points, by partisan bloc. Singapore's blocs are governing-party lean and opposition lean, not left and right.

−40 −20 +20 0 ← expansion more popular restriction more popular → United States −43.1 +24.5 Germany −3.5 +34.0 South Korea +11.6 +19.9 Singapore +8.1 +18.5 left-leaning in the US, Germany and Korea · governing-party lean in Singapore right-leaning in the US, Germany and Korea · opposition lean in Singapore

American partisans relax the standard when the policy restricts.

Penalty for bypassing the rules, percentage points, when the policy expands vs. restricts admissions. Exploratory, not registered.

Left-leaning n = 1,385 support for a lawful policy: 78% when it expands, 30% when it restricts 10 20 0 penalty, percentage points Executive overreach 17.9 8.8 Procedural neglect 17.9 10.6 Right-leaning n = 850 support for a lawful policy: 50% when it expands, 70% when it restricts 10 20 0 penalty, percentage points Executive overreach 13.4 10.1 Procedural neglect 18.4 10.7 policy expands admissions policy restricts admissions

Pooled, the process penalty is the same under expansion and restriction.

Change in the penalty when the policy restricts rather than expands, all four countries pooled, with 95% intervals.

registered equivalence bound, ±5 pp −5 +5 0 ← larger penalty when restricting smaller penalty when restricting → Executive overreach +2.2 External imposition +0.2 Procedural neglect +1.2 pooled estimates, 95% CI
← → navigate · R reset · F fullscreen · N notes · B blackout · P print preview