No, the Senate isn’t rigged
The data refutes a widely-circulated claim that Republicans have an inherent advantage in the chamber. It also signals to Democrats they need to pick up their game — including on their own turf.
After Brett Kavanaugh’s confirmation to the Supreme Court, fringe proposals to abolish or reform the Senate received newfound attention on multiple left-wing websites. One of the key reasons is based on a data point observed on Twitter by Norman Ornstein, a scholar at the American Enterprise Institute and critic of the Republican Party, that circulated among several mainstream news outlets: “By 2040, 70 percent of the country will live in 15 states. 30 percent will choose 70 senators. And the 30 [percent] will be older, whiter, more rural, more male than the 70 percent. Unsettling to say the least.” The logic goes that whiter, older, smaller, more male and (ostensibly more Republican) states could determine up to 70 senators. The assumptions Ornstein and others make are wrong generally because of a common over-dependence on national trends to understand state dynamics. Apart from this, any discussion of the Senate should be prefaced with recognizing that equal representation is a vital feature, not a flaw, toward upholding the functional premise of America’s existence: to impede the acquisition of central authority.
If anything that premise — the unusually high rank it holds in American governing priorities relative to other countries — is what makes America exceptional. Conversations, even provocative ones, about how to enhance or reform American democracy will be better served by acknowledging this premise — that overwhelming consensus is required for legally-binding power to apply to everyone. But this isn’t how some writers with sizable bullhorns have sought to frame the matter. Instead, in the sprited journalistic trend of hunting for benign disparities and attributing malice to them, some have sought to undermine a democratic institution by spinning the inevitable disparities the Senate creates as inherently problematic or another kind of injustice to which only the Trump era has awoken us.
Why is this worth addressing? Why spend energy rebutting critiques that don’t reveal any unknown truths and are fundamentally a product of visceral frustration with the current political composition of the country? The politically active among us might know why we have a Senate and the purpose that it serves, but the lack of civic knowledge among the public suggest that most do not. The two-thirds of Americans who can’t name the three branches of government nor understand their constitutional system cannot maximize their self-governing influence that the Senate preserves by impeding central authority enough so that citizens can have a greater probability of exerting influence in proximate political bodies. These Americans, particularly younger and more liberal audiences of digital media outlets buttressing a Senate skeptical narrative, might be vulnerable to falsely believe claims or signals that the Senate is politically rigged just like some white Americans might believe the hysteria Fox News personalities have incited over demographic change. Misleading data over the Senate no less undermines American democracy than a president who refuses to accept election outcomes, makes unsubstantiated claims about mass voter fraud, or mocks the rule of law.
After Ornstein sent out his tweet, FiveThirtyEight briefly touched on a variant of it: “By 2040, eight states will be home to nearly half (49.5%) of the country’s entire population.” The Washington Post, Bloomberg, and Slatecited either the 70 percent or 50 percent data point. The Atlantic even ran a piece by former Democratic congressman John Dingell calling for the Senate’s abolition and citing Ornstein’s data. What is not acknowledged is that today 66 percent of the country live in 15 states and 48 percent of the country live in eight states. Ornstein implies the data is skewing the Senate, ostensibly on political grounds. Ten years ago, Democrats won an effective 60 seats in the Senate under virtually the same population proportions — of which have remained virtually unchanged since 1960, the last major party realignment.
Washington Post reporter Philip Bump has repeatedly criticized the Senate. The day of Brett Kavanaugh’s confirmation, Bump signaled the vote was crooked because an effective majority of the senators who backed Kavanaugh represent a cumulative minority of the country’s population. Bump didn’t discuss why the Senate exists. Instead, Bump chose the data he liked and reviewed the Senate not as it is, but as it would be if representation were perfectly proportional to a non-existent Senate popular vote. By Bump’s logic, hypothetically, if one additional senator who represented a large state like Illinois or Florida voted for Kavanaugh resulting in a majority of the senators representing a proportional number of the population, would that have made Kavanaugh’s confirmation any more just?
Based on the incoming 2019 class of senators, Republicans hold seats in states that represent 55 percent of the population and Democrats 60 percent (nine states have a senator of each party). If we look at the data as a 55–60 divide, it doesn’t seem dire in the least. In fact, it’s appropriate — given that we know representative equality in the Senate inevitably creates disparities relative to population. At the same time, again by Bump’s logic, the states Donald Trump won in 2016 represent a majority of the U.S. population (51 percent to 48 percent for Clinton). Is this a fair assessment? I don’t think so.
Like many things today, the obvious bares repeating: Senators are elected by the people to represent their state, not the country.
The data point Ornstein first cited received attention by other progressives. At Vox, Ezra Klein cites the 70-percent and characterizes American democracy as “rigged” and “illegitimate”. Klein framed the issue this way, “the constitutional system is distorting the political competition between parties.” In other words, it isn’t incumbent on Democrats to adopt a campaign strategy to earn more seats in the Senate. Instead, the rules should change to suit hyper-momentary politics — rules that haven’t changed in at least 60 years, during which Democrats held the Senate for 42 years. Klein does not acknowledge that the Founders eschewed perfectly reciprocal political competition for equal representation as an impediment to central power. History shouldn’t be revised to contend that the country is failing to meet a goal it never sought to achieve by design.
Based on the political composition of the incoming Senate as well as four factors across all states — population size, the proportion of white Americans, male-to-female ratio and age — I ranked every state by these factors individually and then averaged those rankings to create a single master rank that captures all of these factors combined.
Let’s start with population since that is the most salient. In the largest 15 states, Democrats hold an effective four-seat advantage over Republicans, 19–11. In the bottom 35 states, Republicans hold a seven-seat advantage, 42–28. But it isn’t in the smallest states where Republicans trounce Democrats; in the smallest quartile of states, both parties hold an equal number of seats. Instead, it’s in the second-least populated quartile of states — like Utah, Iowa, and Mississippi — where Republicans gain five of the seven seats in their bottom 35-state advantage.
Among the most-white quartile of states, Republicans hold an effective four-seat lead over Democrats, 10–18. In the second-most white quartile, the parties draw at 12 seats. In the third, or second-least white quartile, Democrats lead by one. In the least-white quartile, both parties draw at 11. Republicans hold more seats in the whitest states, but earn their Senate majority by remaining competitive in every other quartile of states. They lead by five seats in the most-white 25 states and trail by two seats in the least-white 25 states.
Let’s talk about sex. 10 states are majority male. Republicans hold 13 of those 20 Senate seats, a practical three-seat advantage. In the following, second most-male quartile, Democrats hold a one-seat lead. In the third, second-least male quartile, Republicans lead by three seats. In the least-male quartile, Democrats lead by two. In the top 25 most-male states, Republicans lead by four. In the least-male 25, Republicans lead by two. The story is the same: Republicans win where national trends indicate they will and remain competitive in states where they stereotypically are least expected.
The age gap proves to be the most surprising data point given the stereotypes about Republican voters being older. It turns out the younger the state, the more likely it is represented by Republicans. Overall, Republicans hold their largest Senate advantage over Democrats. In the 25 youngest states, it’s a nine-seat lead, 34–16. In the 25 oldest states, Democrats lead by six, 31–19.
The final data set is a master ranking of sorts. It aggregates all of the state rankings based on the four individual factors. The higher the rank, the more representative a state is of those four characteristics — smallness, whiteness, maleness, and oldness — combined. Interestingly, the two highest-ranking states in the master rank are Montana, represented by one Republican and one Democrat, and New Hampshire, represented by two female Democrats. Overall, in the top-quartile, Republicans hold an expected lead of four seats, 8–16. In the second quartile, Democrats hold a 2-seat lead, 15–11. In the third quartile, both parties earn 13 seats each. In the overall least small, least white, least male and youngest quartile, Republicans lead by one seat, 13–11.
(the chart above was added in an update on Jan. 6)
Generally, Republicans perform better in smaller states but, again, the smaller red states — like Wyoming, Utah and Idaho and Mississippi — alleged to be skewing the Senate toward Republicans are practically no more a barrier for Democrats than Hawaii, New Jersey, Maryland and New York are for Republicans.
The bottom line is Republicans perform exceptionally well where national trends would expect — whiter, smaller states — but they appeal to a wider variety of states and remain overall more competitive than Democrats regardless of population, the proportion of whites, male-to-female ratio or age. Republicans only have more to lose as long as Democrats pick up their game. But a Democratic Party that chooses to move further left — and depend on power centralization as a tool in executing a more progressive agenda — could more inclined to surround itself with people who frame the Senate as an unjust barrier toward not so much a faithful democracy but idealistic uniformity.
Ornstein’s omission is a product of political media’s obsession with nationalized macro-analysis — the tendency to understand the country’s political dynamics by superimposing static national narratives and data over fifty dynamic states. Montana and Wyoming seem like similar states, don’t they? Except, over the last 40 years, Montana has elected Democrats to the Senate 5–1, while Wyoming hasn’t elected a Democrat to the Senate in nearly 50 years. Just as Republicans need to understand changing national demographics to retain the presidency in future elections, Democrats will need to adapt their message to the states if they seek to represent more Americans.
With identity politics trending, the same one-dimensional oversimplifications that failed to foresee Trump’s victory take to new heights. That said, it is also often more expensive and more time-consuming for political journalists to collect and crunch statewide data and conduct statewide public opinion polling than it is to grab national data and draw inferences from that. However, if more accurate political journalism is the goal, understanding the Senate (and the House) from a more micro-analytical standpoint will capture the electorate’s views more accurately. It will also provide realistic expectations around what policies are plausible in a way national (polling) data frequently fails to capture since nationalized majority-centric polling often doesn’t scale-down in a vast representative democracy.
Another flaw with macro-analyses like Ornstein’s is that it assumes political coalitions will not change by 2040. In the midst of what is shaping up to a politically, technologically, culturally and environmentally disruptive era, there are plausible cases to be made that coalitions may fragment, new ones will emerge, and that the white-rural/diverse-urban dichotomy may become less stark. Both parties will need to diversify their coalitions at the state level. Senate Republicans can retain victories in states like Georgia and Florida by continuing to narrow their national gap with black and Latino voters at a state level. In 2016, Sen. Johnny Isakson of Georgia won re-election with a comparatively large 18 percent of the black vote and performed 11 points better with Latinos than Trump. This year in Florida, Republican Rick Scott edged out incumbent Democrat Bill Nelson by 10,000 votes, securing 60 percent of whites and two percent more of the black vote than Trump, who outperformed Mitt Romney among blacks. Scott also performed 17 points better among Latinos than Trump. The Florida gubernatorial race was an adjacent example of the GOP’s inroads with black voters: Republican Ron DeSantis won 18 percent of black women over Democrat Andrew Gillum. Democrats will need to narrow their gap with white men and retain their 2018 advantage with white women. More importantly, Democrats need to build a coalition from the ground-up the way Republicans had over the past twenty years. This will raise greater awareness of Republican successes with chipping away at the non-white Democratic coalition.
It isn’t as though Democrats can’t compete or galvanize people in red states. It’s whether they have the will or interest in representing more of the country. Progressives have organized teachers in red states like Arizona, Oklahoma and West Virginia for bigger budgets, legalized marijuana in California, decriminalized it in Mississippi, Missouri and Illinois, expanded taxpayer-funded college tuition in New York, expanded Medicaid in Montana, Louisiana and Kentucky, repealed the “bathroom bill” in North Carolina, flipped a U.S. Senate seat in Alabama, found conservative allies in Texas on prison reform, and established “sanctuary” ordinances for undocumented immigrants in dozens of cities and counties. Democrats need to prove they can convert the energy behind those initiatives into victories in Senate contests. They need to be competitive in open-minded red states like Iowa, Arkansas, North Carolina and Louisiana without depending on an attractive presidential candidate’s coattails. Much to the dismay of progressive voters and journalists, that’s going to involve electing a few Joe Manchins or Jon Testers.
We’re accustomed to evaluating the efficacy of representative democracy on its efficiency or based on how reciprocal and proximate voting is to direct action. The idea goes that providing the people one clear avenue toward self-governance will create an easier democracy to navigate and a more uniformly desirable society. However, this ignores a universal truth about diverse and complex systems: they fragilize when risk concentrates in fewer places.
This phenomenon has parallels in domains from human development to thermodynamics to finance to physiology — and authoritarianism. Fundamentally, America’s constitutional architecture is a risk management system that is instrumental in the realization of American virtues. It reduces risk by dispersing power. It disperses power by taxing the sweeping change that comes with centralized majority power and redistributes that influence downward to the people. This dispersement increases the accuracy with which people are represented and secures multiple medium-risk access points to power for the people. The result is an immune system that can adapt to short-term volatility while remaining resilient in the long-term. In this sense, the efficacy of American democracy bases itself less on efficiency but accessibility — a system where incurring more risk requires buy-in from a consensus of Americans, not a simple majority. If done properly, denationalizing politics can empower more people who are willing to organize and self-govern to partake in American democracy at various levels than it otherwise would in an unregulated, free-market of dog-eat-dog tribal majoritarianism.
Despite his vehement opposition to equal representation in the Senate, James Madison nevertheless recognized its redeeming qualities writing that “this complicated check on legislation may in some instances be injurious as well as beneficial.” Madison goes on to write how equal representation acts as an “impediment” that improves protection “against improper acts of legislation.” Given that the “excess of law-making seem to be the diseases to which our governments are most liable” it is conceivable that equal representation may be “more convenient in practice than it appears to many in contemplation.”
The American public will be better served by fewer petulant attacks on American institutions and more credible media coverage and pressure on the political establishment to address what is broken. This includes the disproportionate influence of the lobbying class; the excessive amount of time Congress spends fundraising; the lawmaking happening in undemocratic bureaucracies; the former and future corporate executives who write and then game federal regulations; the frighteningly unilateral power allotted to the presidency. Ultimately, what is most dysfunctional about American democracy is the misplacement of influence and responsibilities within a system that serves its intended purpose and, as a result, remains improbably resilient. It will remain so as long as the people understand the division of power and how they can engage in self-governance effectively. Equally, democracy will depend on ensuring those we elect have skin in the game and that they faithfully perform the duties of their office by representing their constituents first.