
David Brady on Health Care Reform, Public Opinion, and Party Politics 8/24/2009
What this covers
David Brady of Stanford joins Russ Roberts for an extended conversation about American health care politics, public opinion, and the structural forces that shape legislative action. The discussion centers on why popular support for health care reform collapses once respondents are told what it costs, and traces this pattern to a deeper truth: politicians' votes are driven not by ideology or constituent demand but by calculations of reelection risk. Brady walks through contingent valuation surveys in which abstract enthusiasm for reform evaporates when a tax price is attached, and argues that legislators hold private ideological views but vote on what he calls "induced preferences"—positions calculated to maximize their survival at the polls.
The conversation ranges widely across political economy. Brady explains how supermajority institutions—the presidential veto and Senate filibuster—create a gridlock zone between the 40th and 60th senator within which the status quo cannot move. He traces the rise of the "personal vote" in the mid-1950s, a measurable shift in which incumbents built loyalty to themselves rather than their party, a phenomenon he dates to scientific polling's forcing politicians to position themselves independently of their president's fate. He argues that what appears as rising polarization is actually "sorting"—liberal Republicans and conservative Democrats switching sides—leaving the underlying distribution unchanged but visually separated. On health care specifically, Brady examines why universal coverage succeeded elsewhere when medicine was cheap but cannot work in America now, and why the 1993–94 Clinton effort failed despite a Democratic president: there simply weren't the votes, and the dilemma remains that letting people keep their plans makes cost control impossible. He closes with skeptical reflections on how much of political science qualifies as science.
Brady argues that public support for health care reform collapses once respondents are forced to confront real dollar costs, and that legislative gridlock and politicians' positions are best explained by induced preferences—legislators calculating reelection consequences—rather than by ideology or genuine demand for change.
- Contingent valuation surveys show abstract support for reform evaporates once a tax cost is attached
- Legislators hold private ideological views but vote on induced preferences shaped by constituents and reelection risk
- Gridlock arises from supermajority institutions (filibuster, veto) keeping policy stuck at the status quo between the 40th and 60th senator
This asset isn't compiled yet
You're seeing its claims, ranked. Compile it to build the argument threads, weight them, and check each claim against your library — the full view.
The 'personal vote' worth about 8% emerged in the mid-1950s (documented via the O'Hara 'Democrat'-to-'Chicago' signage example), enabling candidates to build loyalty to themselves rather than their party, which explains why presidents from 1948 onward routinely faced opposite-party Congresses—a phenomenon that had not occurred since Zachary Taylor in 1848.
“this is called the personal vote and if we can document that the personal vote begins at about this time and what that meant was that candidates were able to have some personal loyalty to them not to their party... it a ver just about 8%”
The personal vote is measurable via 'retirement slump' (a party loses about 7-8% of vote share when an incumbent retires) and 'sophomore surge' (a new incumbent gains about 8% in their first reelection), both of which were random before the 1950s, appeared thereafter, and remain around 6% today.
“when a member retires what... happens to his party share the vote... it's random until the 1950s after that retirement slump average is about 7%... we have a clever term sophomore surge and on average the sophomore surge is about 8%”
Pre-1932 incumbency was strong but attached to party and economic success (98% of Republican incumbents won 1920-30 amid prosperity, but they lost heavily when the economy collapsed in 1932); the modern personal vote shifted incumbency advantage to individuals, eliminating coattails and allowing opposite-party incumbents to do better by voting with or against the president as needed and explaining it as principled independence.
“the Republican incumbency... between 1920 and 1930 to 98% of incumbents who ran for re-election won but the difference was they were ran and won because... the economy was doing very well... in 1932 when the economy went to hell incumbents lost like crazy... those coattails disappeared with the rise of the personal vote”
The exogenous innovation of scientific polling (better than the 1930s practice of party bosses estimating popularity) forced politicians to behave differently: by 1956 a Democratic congressman who knew Stevenson would lose had to position himself independently of the presidential ticket, giving rise to split-ticket voting and the modern personal vote.
“what happened was the exogenous innovation of polling forced people to start to behave differently... as polling got better you're a Democratic congressman in 1956 you've already been beaten by Eisenhower... you begin to think about how do I position myself differently”
The great issues in American politics have always been solved by partisanship, not bipartisanship: one party takes the progressive position (Republicans ending slavery and keeping the gold standard, Democrats building the welfare state) while the other holds the old position, until the losing party is forced to move over and the old question disappears—e.g., the welfare debate shifted from 'welfare vs. no welfare' to 'how much welfare' after the New Deal.
“the great issues in American politics have always been solved by partisanship... in each case what happened was one party took the progressive position the other party stayed with the old position”
The apparent increase in party polarization is actually a 'sorting' rather than a change in the underlying distribution of views: liberal Republicans became Democrats and conservative Democrats became Republicans, so the same marbles are now separated into distinct jars, while the 1974 reform acts gave leadership power to structure votes along left-right lines.
“it's a sorting now and the people who were kind of liberal Republicans are now Democrats and the people who were conservative Democrats are now Republicans... think of a big blue marble... they used to be intermixed now the Reds are over here... so that's a sorting and not a change fundamental change in the distribution”
Not every voter or issue counts equally; over time representatives build 'slack' or trust with voters that gives them leeway, but the Brady-Cogan-Keynes research shows that voting too far from one's district eventually gets a member defeated, though it is a slow, uncertain process.
“there's another economists call it slack... representatives view build trust or something so that they have some slack and voters give them some leeway in how they vote”
Legislative gridlock can be defined objectively using median-voter theory and supermajority institutions: because the presidential veto and the Senate filibuster (Rule 22, requiring 41 senators to block) create a gridlock region between the 40th and 60th senator, any status quo policy located in that region cannot be moved without an election changing the pivotal members.
“there are supermajority institutions... the presidential veto... and in the Senate rule 22 which is the filibuster which says if you have 41 senators willing to stick together you can't vote... if legislation is stuck at the status course between the 40th and the 60th senator it's hard to move it out and only an election can change that”
Standard surveys mislead because they don't properly value or ask about the trade-off, analogous to environmental willingness-to-pay studies where people claim they'd pay a lot to clear up a view because the public good seems costless to them.
“most surveys don't properly value or ask the question about what's the trade-off... studies of the grand when it's cloudy... if you ask people well why are you willing to pay to clear the ground oh really to pay a lot... if it's a public good and it doesn't cost anything”
Despite roll-call polarization, bipartisanship has actually increased in co-sponsorship of bills (introduced in 1974) because members from evenly divided districts must show constituents they are moderate, doing so through cross-party co-sponsorship rather than crossover floor votes.
“they introduced this issue of joint sponsor co-sponsorship of bills and if you trace co-sponsorship of bills there's been a lot more bipartisanship on co-sponsoring... if you come from one of those districts that's got evenly divided... you can't win reelection being either fully Republican or Democrat”
Bipartisanship cannot resolve fundamental disagreements rooted in opposing worldviews (e.g., whether markets or government better solve problems like climate change), so the proper democratic mechanism is elections that give a party the ability to govern, after which voters judge results and can vote them out.
“I'm supposed to give up my view that markets work better than government you're gonna give up your view the government's work better than Markus I don't think so and therefore as long as those are the sorts of sets of issues why would we expect bipartisanship... what we expect is elections that give parties an ability to govern and then it works or it doesn't work and you get to vote them out”
Political science has made genuine scientific progress in areas that borrow economic assumptions (e.g., that legislators want reelection, treated like a price variable) and game theory, plus psychology/behavioral economics (framing analysis à la Tversky and Kahneman), but much of the field—political theory and historical analysis—remains art rather than science, paralleling economics which has more science but still much art.
“there are certain areas of political science that have made progress and are more scientific... where you're able to quantify and to take assumptions from economics... essentially they've used economic models and economic thinking and apply them to politics and then there's sort of the psychology... behavioral economics making some progress... but there's still a lot of art a lot of art there as well as the science”
What legislators are uncertain about is not the policy effect of a plan but the reelection consequence; the 1993 'Harry and Louise' ads were effective because they increased uncertainty—warning satisfied voters their doctor might be taken away—making Democrats fear a future opponent who would blame them, which pushes legislators to fall back on the status quo under uncertainty.
“the Harry and Louise commercials were very effective but only because they increased uncertainty... Democrats in this case who wanted to vote for the legislation are thinking if I vote for this and it doesn't work out... they've taken the doctor I'll get an opponent who will say... we took their doctor”
Legislators' votes reflect 'induced preferences' rather than personal ideology: a thought experiment where half of members vote unrecorded each time shows the unrecorded outcome would be more left for Democrats and more right for Republicans than the public votes, because recorded votes are calculated to maximize reelection given constituents' positions.
“imagine all the interest groups... on any given vote half of the members of the House and half of the members of the Senate... get to vote but it's not recorded... would the legislation be more left or right my view is that representatives and senators have private views and those private views for Democrats are further left than the legislation they passed and for Republicans for the right”
Universal-coverage systems in Britain, Italy, Germany and elsewhere succeeded because they were established when the cost of medicine was low, allowing creation of public boards with authority to deny treatments; the US cannot replicate this because current expectations and advanced technology are so high that no minimum standard denying treatments could ever win agreement.
“when those systems started the cost of Medicine was quite low... they were able to create public boards they were created institutions that actually have authority and make decisions which deny certain treatments etc in the United States at the present time the expectations are so high”
Average voters exhibit a schizophrenia about taxes—treating income tax as funding the government while believing payroll taxes go into their own Social Security/Medicare 'lockbox' account—when in reality payroll taxes (near a flat ~15-16% counting the employer share economists believe workers bear) fund general spending, distorting perceptions of who pays for government.
“I think a large portion think they don't pay anything for the government which is the income tax part... they often think that their side of it the eight and seven to eight percent... is going into their own account in the proverbial lockbox so of course it's not”
Americans who pay zero federal income tax are markedly more in favor of health care reform because, under the survey's static assumption that they won't have to pay, reform costs them nothing—making this a conservative test that likely understates resistance.
“a lot of Americans pay zero income tax right so not surprisingly they're really enthusiastic about health care reform... it's rational it's not going to cost them anything”
Polls show persistent abstract desire for health care reform yet little change occurs because most Americans (about 87% in the early 90s, 83% currently) are satisfied with their own health care, so any shift from the status quo is politically problematic.
“in the 90s before the Clinton period it was about 87 percent thought that their health care was either very good or good... and even at the present time it's 83 percent are pretty satisfied with their health care therefore when the status quo is you're pretty satisfied anything that shifts the status quo is problematic”
Two-thirds of Americans consistently favor lower taxes, more spending, and a balanced budget simultaneously because politicians tell them there are no trade-offs—that costs are short-term and reforms save money in the long run—so voters have no incentive to weigh trade-offs in polls.
“any time you do the polling two-thirds of Americans are in favor of lower taxes more spending and a balanced budget... I think people are told by politicians that there are no problems there are no trade-offs”
Politicians build the personal vote through constituency service—maintaining many district offices (especially the electorally insecure), spending Friday-Monday at home, sending franking mail, baby books, constituent surveys, and Capitol-flown flags, and using local press facilities—techniques analogous to product brand loyalty that can be squandered.
“the more insecure you are electorally the more people you have back home... members are back home Friday Saturday Sunday Monday... they send out franking... getting a flag that's flown over the Capitol United States that's passed out by members of Congress”
When survey respondents are asked via contingent valuation to attach a real dollar tax cost (based on their actual estimated taxes) to a reform option, support falls away rapidly—everybody favors reform but no one is willing to pay for it; only coverage for people with chronic/continuing illness retained borderline (~50-52%) support.
“it turns out that everybody sort of in favor reform but no one's willing to pay for it”
Smart, election-winning politicians repeatedly overestimate what they can accomplish (Clinton in 1992 on gays-in-the-military and health care, Republicans in 1994 on the Contract with America, of which only one item passed the Senate), an unexplained pattern likely tied to 'the flush of winning.'
“why is it that relatively smart politicians always over[estimate]... in ninety-two Clinton comes in absolutely over estimates what he can do at gays-in-the-military health care etc... the Republicans get in in a 1994 absolutely overestimate what they can do in terms of the contract with America”
The 1993-94 Clinton health care reform failed because, despite a Democratic president shifting the veto pivot leftward, the actual votes weren't there: the House bill never cleared committee, and as Cooper/Moynihan/Chafee dropped coverage from 100% to 85% to lower costs they got closer but ran out of time, after which Democrats lost the 1994 election.
“the bill mrs. Clinton wrote and submitted to the House never even cleared the committee... they started dropping coverage going from a hundred percent eighty-five percent and therefore decreasing the cost and that got closer but by the time the session ended they... couldn't get it done”
All current reform proposals adopt the 'if you like your plan, keep it' lesson learned from the Clinton failure, but that creates the core dilemma: letting people keep their plans makes it impossible to deal with cost, so members of Congress know the cost will likely go out of sight.
“all of the proposals before the Congress today our proposals where if you like your health care plan keep it but that's the problem if you keep your health care plan you can't deal with the cost”
Decades of tax changes have removed more and more families from the income tax, creating an unhealthy political economy in which roughly 40% support any government expansion because it costs them nothing out of pocket, while ignoring the long-run economic effects of rising tax burdens.
“we have removed more and more families from the income tax so the political economy that is a pretty unhealthy thing... for any type of expansion of government with pretty much a 40% support because doesn't cost you anything out of pocket”
Insurance for routine health care distorts behavior the way insuring oil changes would: because the marginal cost to the user is near zero, people overuse the service (the well-to-do get oil changed every 500 miles), and the political process then expands mandated coverage to include ever more items ('pinstriping'), driving costs up.
“if you're well-to-do you get your oil changed every 500 miles rather than every 2,000 because it wouldn't cost you anything which is what happens with insurance right people overuse the doctor because it doesn't cost them anything”
State coverage mandates (e.g., California requiring acupuncture and massage therapy versus Arizona's bare-bones plan) ratchet up the minimum cost of insurance, so the political dynamic always expands coverage, making a basic low-cost minimum plan hard to achieve and underlying the rationale of McCain's proposal to allow cross-state plan purchase.
“the Arizona plan is pretty much bare-bones minimum where the California plan includes acupuncture and massage therapy etc... the political dynamics that is to always expand it”
The US already approximates a hybrid public-floor/private-choice system: Medicaid covers the poor and Medicare the old with generous use of best technology, and the genuine gap is the group not poor enough for Medicaid who choose jobs trading higher wages for fewer benefits, plus eligible-but-unenrolled people and young high earners—so the country is 'not far' from the minimal-floor model.
“there is a floor it's pretty generous they use the best technology poor people get heart attack prevention and cure... the gap is that group of folks who aren't poor enough to qualify for Medicaid and choose often employment that has higher compensation and wages and salary and lessen benefits”
Cost matters most to the ~36% of independent voters, who are starting to trend against reform as facts emerge, which drives senators from Republican-leaning states (like Nelson of Nebraska) to oppose high-cost legislation regardless of party.
“cost matters to people it seems to matter most to independents... the 36% of people who are independent they're starting to trend down independence of always worried more about costs as the facts come out”
The 2009 stimulus and GM/Chrysler bailouts passed despite weak popular support because of timing: a popular president, committee chairs eager to act after years out of power, little organized press opposition, and the ability for representatives to lay off responsibility on the president if it failed—while support fell away over time as the GM/Chrysler bailout and then health care followed.
“the president was in favor of it the Democrats in Congress the committee chairs have been sitting there out of office... they were anxious to do what they hadn't been able to do... there was nothing not much in the press that took it on... the president's very popular and if it didn't work they could lay it off”
Brady predicts the 2009 reform will not pass the high-cost ~$1.2 trillion Waxman bill because moderate 'blue dog' Democrats and conservative-state senators (Landrieu, Pryor, Lincoln, Nelson) cannot vote for it; instead something like a ~$600 billion state-based, non-entitlement, budgeted bill (like the Dorgan bill) that increases coverage is what can pass.
“my view is they will not get that bill because the blue dog or moderate conservative Democrats... will not be able to vote for it... I think what they're going to get is something like a six hundred billion dollar bill that will be more like the Dorgan bill which is going to be through the states and not an entitlement”
Even though reconciliation would let a bill pass with 50 rather than 60 votes (avoiding filibuster), Democrats with 60 senators still lack 50 votes for a high-cost plan because conservative-state Democrats won't vote for measures that endanger their reelection, and the absence of real cost controls (which the CBO will expose) further undermines the bills.
“they can't be filibustered because it's under reconciliation which means that you only need 50 votes rather than 60 votes but the problem is they don't have 50 Democratic votes... these people are not going to vote for stuff that causes them problems in their next election”
Coverage for people with chronic/continuing illnesses retained borderline support even when cost was attached because such people cannot be held responsible in any way for their condition.
“these are people who can't be in the held in any way responsible for their condition and even that we had done narrowest of margins”
The internet panel survey method (YouGov/Polimetrix) is more reliable than phone polling for cost-attached questions because the same individuals are repeatedly sampled, so detailed background like income is known reliably, allowing accurate estimates of the taxes respondents would actually pay.
“this is a panel of individuals who are sampled on particular questions where unlike a question... that's implemented over the phone because you're often interviewing the same people you have a great deal of information about them so you know their income pretty reliably”
Measured by roll-call voting, bipartisanship has effectively disappeared (Obama's Republican approval fell from ~48% to ~20% after the stimulus), but this is partly an artifact of measurement—real bipartisanship now occurs through other channels like co-sponsorship.
“he started out with about 48% Republican approval and with the stimulus package that went immediately down to 20% approval... if the only way you're measuring that is on roll call voting then you're not going to find there's not gonna be any bipartisanship”
Welcome to econ talk, the website and email address where listeners can subscribe and comment.
“welcome to econ talk part of the library of economics and liberty I'm your host Russ Roberts”