T407016 // Autumn 2026 // Université de Genève
Comparative
Politics
Session 08. Interest groups and who gets the policy they want.
Last time, this time
- Katz and Mair on how parties moved from civil society into the state
- Why the cartel breeds its own challengers
- Gilens and Page on whose preferences predict US policy
- Baumgartner and colleagues on who wins and who loses in Washington lobbying
- One group exercise
Katz and Mair saw interest organizations take over the job of carrying demands to the state. This week asks which organized interests get the policy they want, and whether ordinary citizens do.
Think of a policy most people in your country want
Who gets the policy they want?
Eight in ten citizens favor a change. Guess how often it happens within four years.
Does the side that spends more on lobbying usually win?
Who is missing from the lobbies in your capital?
Who governs?
The average citizen
Vote-seeking parties converge on the median voter, so policy follows what the average citizen wants. Downs, and what Dahl called "populistic democracy."
The wealthy
People with high income or wealth dominate key issues despite elections. Beard, Domhoff, and Winters's "oligarchy."
Groups that stand in for everyone
Struggle among many groups yields policy close to what citizens need. Madison's factions, Truman, Dahl's New Haven.
Business groups
The universe of groups is tilted, and policy leans toward corporations and business and professional associations. Schattschneider, Stigler, Lindblom.
Each tradition has a large literature and a great deal of evidence behind it. Until recently none had been tested against the others in a single model.
Will the missing groups form?
Potential groups
Every interest has some influence. Officials must heed the groups that would form if those interests were trampled.
Free riders
Large, dispersed interests rarely organize. Each member's stake is small, everyone lets someone else do it, and officials can ignore them.
An upper-class accent
The heavenly chorus of groups sings with an "upper-class accent." Business prevails when the scope of conflict is narrow and visibility low.
Most interest groups and lobbyists represent business firms or professionals, and few represent the poor. If business wins, is it because business groups are many or because they are rich?
Four theories, four predictions
| Theory | Average citizens | Economic elites | All groups | Mass groups | Business groups |
|---|---|---|---|---|---|
| Majoritarian electoral democracy | |||||
| Economic-elite domination | |||||
| Majoritarian pluralism | |||||
| Biased pluralism |
Two filled squares mean substantial independent influence, one means some, none means little or none. Each theory makes a prediction the data could prove wrong.
1,779 policy questions
Share in favor at the 50th income percentile, a stand-in for the median voter.
Share in favor at the 90th percentile, a household income of about $146,000 in 2012.
Groups from Fortune's "Power 25" lists plus the ten industries that spent most on lobbying, each coded for or against.
Net = ln(F + 1) − ln(A + 1)
The bars show ln(n + 1) for n groups on one side. Each added group raises it by less than the one before.
F and A count the groups for and against, with "somewhat" counted as half. The logs make a second group count for more than an eleventh.
Which way do the errors cut?
The 90th percentile is "neither very rich nor very elite." Still, the top 2 percent's preferences correlate .91 with the top 10 percent's and .69 with the average citizen's.
Only a small fraction of active groups is counted. Baumgartner's team found that the Power 25 count correlates .73 with their full index of group resources.
Only proposals a national poll asked about. They lean toward high salience and cover only a slice of all possible policies.
All three predictors together explain little of the variation in outcomes (R² = .074). Gilens and Page argue the flaws make elite and group influence look smaller than it is.
Alone, everyone looks powerful
Coefficient on policy change, each actor alone and all three together
- Taken one at a time, each actor's preferences predict policy change. Each tradition has its adherents for good reason.
- Entered together, the average citizen's effect falls to near zero. Elites and groups keep almost all of theirs.
- Average citizens and the affluent agree often (r = .78), so a link that shows up alone can be spurious.
All predictors run from 0 to 1, and the outcome is whether the change happened. Pure majoritarian electoral democracy "can be decisively rejected."
What moves the odds of a change?
- Low support means one in five in favor, or five groups strongly opposed and none in favor. High means four in five, or five groups strongly in favor.
- Each line holds the other actors at neutral, half in favor and a net group score of zero.
- The average citizen's line stays flat at about .3. A tiny minority and a large majority get the same odds.
Roughly 15 percent of proposals fall beyond these cut-offs. The low and high points are realistic values near the edges of the data.
Democracy by coincidence
Average citizens and the affluent want the same thing much of the time (r = .78). Citizens then get the policy they want as "coincidental beneficiaries rather than causes of the victory."
The disagreements are important. They include trade restrictions, tax policy, corporate regulation, abortion and school prayer. On these, economic elites "generally get their way."
"When push comes to shove, actual influence matters."
Earlier work found policy matching majority opinion about two-thirds of the time, and as description that still holds. Read as causal impact, those findings "appear to be largely or wholly spurious."
Eight in ten is still too few
Of changes backed by a narrow public majority that happened.
Of changes backed by 80 percent of the public that happened.
Of changes strongly backed by both the affluent and interest groups that happened.
Of cases with groups engaged where most groups opposed the change. Most favored it in 36 percent.
Federalism, the separation of powers, bicameralism and congressional rules all protect the status quo. A majority that opposes a change usually gets its way, and a majority that wants one usually does not.
Whose groups?
Correlation with average citizens' preferences
Coefficient on policy change, all four together
Business groups take stands slightly against what average citizens want, and their effect is nearly twice that of mass groups. Group for group the two are about equally influential. Business wins on numbers and cohesion.
Does the majority rule?
| Theory | Verdict |
|---|---|
| Majoritarian electoral democracy | "Nearly total failure" |
| Economic-elite domination | Does "rather well," though elites share influence with groups |
| Majoritarian pluralism | "Looks very shaky, indeed" |
| Biased pluralism | "More descriptive of political reality" |
"In the United States, our findings indicate, the majority does not rule."
The data speak to winning on contested issues and partly to setting the agenda. Both readings study Washington. Would the verdict hold in your country?
Who wins in Washington?
Randomly drawn issues, 1999 to 2002.
Lobbyists and officials interviewed.
Advocates on 214 sides, about 23 per issue.
Of follow-up to see who got what they wanted.
Journalists and many scholars study captivating cases and roll-call votes. This is a random sample of the issues lobbyists work on.
Who shows up, and who gets noticed
- The issue identifiers were drawn in proportion to lobbying activity, so they mirror the lobbying community. Corporations, trade and professional groups are 74 percent of them.
- Among the major participants on the same issues, those three fall to 48 percent and citizen groups rise from 15 to 26 percent.
- Citizen groups are spread thin. The American Petroleum Institute has about 270 staff and a $42 million budget for petroleum issues alone.
Unions are 6 percent of interest-group participants, and six unions account for more than half of those mentions. The identifiers, who set the agenda, are systematically wealthier.
Advocates inside government
Issues with at least one advocate of each type, of 98
- About 40 percent of the major advocates were government officials. Very few officials stayed neutral.
- Party leaders seldom push an issue themselves. They matter mainly as gatekeepers.
- Policy making is "a bottom-up process that is relatively porous for advocates."
That is why the book says advocate rather than lobbyist. Many advocates are already inside the government.
Two agendas
| Policy topic | Lobbying issues, of 98 | Public's top problem, % |
|---|---|---|
| Health | 21 | 8 |
| Environment | 13 | 2 |
| Transportation | 8 | 0 |
| Banking, finance, commerce | 7 | 0 |
| Law, crime and family | 5 | 26 |
| Macroeconomics and taxation | 2 | 19 |
| International affairs | 1 | 10 |
- Ten of the 98 issues had no news coverage of any kind, and more than 70 percent never made TV news.
- Half the issues with newspaper coverage drew 15 or fewer stories in 29 major papers.
- Only 51 reached a floor vote in either chamber, the stage most earlier studies looked at.
Most lobbying concerns professions, industries and businesses. Where citizen groups take part, they mostly carry environmental values. The worries of the poor and the unemployed rarely appear.
The status quo already reflects power
Lobbying aims to change a policy already in place, the product of earlier rounds of bargaining.
Every policy studied touched many constituencies. Trade with China could be argued in terms of farmers, textile makers, human rights or national security.
The sides that defend or attack a policy are rarely homogeneous. "Lone ranger" lobbying for a single favor is far from the norm.
Thousands of bills and hundreds of hearings compete for space. A reasonable solution is usually not enough to get on the agenda.
"If the wealthy are so certain to win, they should already have won in a previous iteration of the process."
What the four observations imply
- The link between material resources and getting what you want is close to zero.
- Resources mobilized to change the status quo and to protect it are roughly equal in most cases.
- Sides are heterogeneous, with rich and poor groups on the same side.
- Resources matter more for whole sides than for single lobbyists, and not much for either.
- Policy is usually stable. When it changes, it tends to change a lot.
- When mobilization is lopsided the richer side wins, above all when it defends the status quo.
When the goal is change, interest groups are "a surprisingly ineffectual lot." One of the best single predictors of success is whether a side is protecting the policy already in place.
Same city, different answers
| Gilens and Page 2014 | Baumgartner et al. 2009 | |
|---|---|---|
| Question | Whose preferences predict which proposals are adopted? | Who wins when organized interests try to change policy? |
| Cases | 1,779 proposals asked about in national polls, 1981 to 2002 | 98 issues drawn from lobbying reports, 1999 to 2002 |
| Salience | Leans high, since pollsters asked about them | Mostly low, and some had no news coverage at all |
| Business | Business groups end up on opposite sides less than 5 percent of the time | "The wealthy often oppose the wealthy" |
| Status quo | Groups oppose change more often than they back it | Defending the status quo is one of the best predictors of success |
| Bottom line | Economic elites and business groups dominate | Resources show a weak link to policy change |
Gilens and Page built their group measure on a proxy Baumgartner's team had validated. Where exactly do they disagree?
Can both be right?
Gilens and Page study the public's agenda, proposals worth a poll question. Baumgartner's team studies the lobbying agenda, mostly issues the public never hears about.
Baumgartner's team asks whether a side's money and staff predict its success, and expects past power to be built into the status quo already. Gilens and Page ask whose stated preferences predict which changes happen.
Both find a strong status quo bias, and both find mobilization skewed toward business and the professions. Baumgartner's team calls the status quo "a quite unfair equilibrium in many cases."
Baumgartner and colleagues see two ways to change the skew. The missing constituencies mobilize, or elected officials learn whose voices are muffled and whose are absent.
Which of four proposals will pass, and which one should the Consumers' Federation fight for?
Your client is the Consumers' Federation, a citizen group with staff for one campaign this year. It favors A and B, opposes the rail cut, has no position on C and is not yet active on any of the four. Each proposal would change an existing policy. Support is the share in favor among median-income and top-decile poll respondents. Groups are strongly for or against unless marked.
| Proposal | Public | Affluent | Groups for | F | Groups against | A | News stories |
|---|---|---|---|---|---|---|---|
| A · Raise the top income tax rate from 40 to 45% | 63% | 28% | Union federation, an antipoverty charity | 2 | Employers, bankers, chamber of commerce, insurers, retailers, small business | 6 | 140 |
| B · Cap prescription drug prices | 79% | 70% | Insurers, retirees, patients' association | 3 | Pharmaceutical industry, pharmacists' association | 2 | 85 |
| C · Allow night flights at city airports | 24% | 41% | Airlines, airport operators, hotels | 3 | Residents' association, somewhat | 0.5 | 2 |
| D · Cut the regional rail subsidy by a third | 34% | 52% | Taxpayers' union, road hauliers | 2 | Rail workers' union, commuters, rail equipment makers, regional governments | 4 | 30 |
Net group alignment · "somewhat" = 0.5ln(F + 1) − ln(A + 1)
Figure 1 · chance a change passes
- Affluent support
- 20% → .18 · 80% → .45
- Groups, strongly
- 5 against → .16 · 5 for → .47
- Public support
- any level → about .3
- Affluent and groups
- both strongly for → .56
- Predict with Gilens and Page. Compute each proposal's net group alignment and rank the four by their chance of passing. How would a median-voter theory rank them?
- Reread the file with Baumgartner and colleagues. Which proposals would each study's sample likely include? Where do the wealthy oppose the wealthy, and who defends the status quo? Revise your ranking.
- Advise the Federation. Pick one campaign, the side and the allies inside and outside government it should join, and the biggest risk it faces.
The country, groups and figures are invented.
Before Thursday 19 November
Week 10. Welfare states and redistribution.
Esping-Andersen, Gøsta. 1990. The Three Worlds of Welfare Capitalism. Princeton University Press, chapter 1.
Iversen, Torben, and David Soskice. 2006. "Electoral Institutions and the Politics of Coalitions: Why Some Democracies Redistribute More Than Others." American Political Science Review 100(2), 165-181.
- Complement on hypotheses and how they are evaluatedM2170
- Quiz 4 opens the class. It covers Weeks 07 and 09.
Progress saved
Questions?
Office hours Thursdays 14:30 to 16:00 in M4328
Rate each group member's participation in today's exercise. Required before you leave the room.