Have Democrats Learned Anything?

The Democrats managed to win last November's presidential election, but what about the next one? Given the party’s dependence on white suburban voters and the threat of resurgent Trumpism, they will most likely double down on their risk-averse 2020 strategy. That will only mean inviting further working-class defections.

Hillary Clinton makes a concession speech after being defeated by Donald Trump, in New York on November 9, 2016. (Jewel Samad / AFP via Getty Images)

Despite Joe Biden’s victory in the 2020 presidential election and the Democrats’ razor-thin margin of control over the US Senate, the party’s underwhelming showing in down-ballot races — both at the federal and state levels — has put it on course for an internal reckoning. What lessons will the Democrats learn from their losses in the House, their near-failure to capture the working Senate majority required to pass any meaningful progressive legislation over the next two years, and their forfeiture of multiple state legislative chambers leading up to a crucial redistricting year? And what are the implications of the Democrats’ internal postmortem for progressive electoral politics? While not written to anticipate the contours of post-2020 Democratic politics, Seth Masket’s masterful Learning From Loss: The Democrats, 2016–2020 provides a wealth of useful insights for understanding the Democrats’ likely trajectory over the coming years, as well as the strategic decisions progressives must grapple with as they seek to expand their influence on Democratic Party politics.

First, Learning From Loss makes a forceful case that the Democratic Party is both better organized and, paradoxically, more porous than most commentators on the progressive left allow. In particular, Masket argues that Biden’s victory in the 2020 Democratic presidential primary had much less to do with a concerted effort by Democratic leaders to tip the scales against progressives, and much more to do with a hyperfocus on electability among the party’s base, combined with a powerful (if empirically questionable) assumption that progressive presidential candidates are simply not electable. These findings should cast serious doubt on claims that progressives are structurally incapable of wielding influence within, or even taking control of, the Democratic Party, even as the medium-term prospects of doing so appear weaker in light of the 2020 election results.

Next, Masket makes a series of important contributions to understanding the legacy and potential future of the Bernie Sanders wing of the Democratic Party. Through a detailed analysis of Democratic donor patterns and candidate staff reshuffling between 2016 and 2020, he shows that, although the democratic socialist faction of the party was unable to expand its 2016 ranks significantly in 2020, it has become a distinct and stable force within the party that is not likely to dissipate in the near future. At the same time, however, Masket’s findings suggest that there is no obvious path forward for Sanders’s coalition to meaningfully expand its ranks within the party over the coming years (at least not outside of heavily progressive areas of the country). This poses a major challenge to democratic socialists in Congress and in state legislatures who have experienced impressive gains in recent years (including in 2020) but who appear to be edging ever closer to a relatively low electoral ceiling.

Joe Biden and Sen. Bernie Sanders during the second night of the first Democratic presidential debate on June 27, 2019, in Miami, Florida. (Drew Angerer / Getty Images)

Finally, Masket’s rich analysis of Democratic Party activists’ interpretation of Hillary Clinton’s 2016 loss offers insight into the debates we are likely to see unfold within the party over the coming months and years: Was the Democrats’ weak showing in 2020 a product of poor campaign strategy and messaging (particularly around identity politics), poor candidate selection, exogenous factors related to COVID-19 and the accompanying economic and health crises, or some combination of these factors? To what extent will party activists reevaluate previously held assumptions about the relative competitiveness of progressive candidates against centrist candidates, white male candidates against female candidates and candidates of color, and campaign strategies targeted to working-class voters, voters of color, and wealthy suburban constituencies within the party?

Learning From Loss suggests that post-2020 intraparty debates are likely to mirror — and ultimately produce a similar outcome to — the party’s post-2016 debates. The party will remain highly risk-averse and will have few incentives to update its post-2016 belief that centrist candidates tailored to the median white suburban voter remain the key to party success. I close this essay with reflections on the strategic liabilities of repeating the lessons learned in 2016, which are based on the assumption that any progressive messaging is a political liability for Democrats in red and purple states. To the contrary, I suggest that economic populism remains a potentially powerful, if largely untested, strategy — not only for progressive success but also for Democratic Party success writ large.

Sanders Wasn’t Robbed, He Just Lost

Learning From Loss sets out to explain how the Democrats’ collective assessment of Clinton’s 2016 loss set the stage for, and in many ways determined the course of, the 2020 Democratic presidential primaries. Its core thesis, based on a range of survey evidence and interviews with Democratic Party activists between 2017 and 2020, is that post-2016 political conditions significantly increased the likelihood that a candidate like Joe Biden would end up securing the nomination. Contrary to many analysts on the progressive left, Masket makes a strong case that Biden’s victory over Sanders was not the result of a concerted effort by party insiders to stop a democratic socialist insurgency. In fact, he argues that conditions within the party in 2020 were as amenable to outsider challenges as perhaps they ever had been.

Masket identifies three factors that made the Democratic Party in 2020 more favorable to Sanders than in 2016. First, Sanders’s capacity for successful fundraising based almost entirely on small-dollar donations significantly limited large donors’ and party elites’ control over the primary process. Second, the role of superdelegates was weakened compared to past primaries — in large part thanks to reforms insisted on by Sanders and his allies in the wake of the 2016 primaries. Further, Masket points out that, far from attempting to reign in such reforms, Democratic National Committee (DNC) chair Tom Perez was excoriated by Sanders opponents who accused him of pushing too hard in the direction of Sanders-friendly reforms. In short, the party took all realistic steps possible to ensure the 2020 playing field would be more even than in 2016. Finally, the 2020 Democratic primary field included an unprecedented number of plausible mainstream candidates who served to split the primary vote, thereby increasing Sanders’s chance of victory.

All that being said, however, Democrats did ultimately circle the wagons around Biden in early March 2020, demonstrating an impressive level of coordination and organizational strength. Was this, as many Sanders supporters have argued, evidence that the party was in control of the process all along and would never have let a candidate as far to the left as Sanders secure the nomination? Masket argues to the contrary, suggesting that two interrelated factors explain why Biden prevailed. In the first place, even before March, the party base had largely decided both that Sanders was their least preferred candidate and that Biden was an acceptable alternative to whoever they had initially supported in the primaries. In polls of Democratic primary voters, Sanders was consistently ranked as the most unpopular candidate in the field (i.e., the largest number of respondents reported that he was their least favorite candidate), while Biden consistently ranked as the most popular candidate (he was the top choice of the largest number of respondents) and the least unpopular candidate. As a result, when it became clear that Sanders and Biden were the only two remaining viable candidates, it should not have come as a surprise to anyone that the rest of the party continued to dislike Sanders and opted to support Biden.

Senator Bernie Sanders speaking at a campaign rally on March 1, 2020. (Wikimedia Commons)

In turn, Democrats perceived the stakes of the 2020 election as being higher than any previous election in memory, making them particularly risk-averse during the nominating process. To drive this point home, Masket points out, for instance, that Donald Trump’s approval rating among Democrats was even lower than Barack Obama’s among Republicans. Further, in previous election cycles (and as recently as 2019), Democratic voters reported valuing policy congruence with candidates more highly than candidate electability. Democrats’ calculus of electability and policy congruence changed dramatically in 2020, as they viewed defeating Trump as a higher priority than any other objective. As Masket’s interviews with party activists show, many Democrats expressed a willingness to compromise on important policy goals in order to defeat Trump. In this context, Biden, viewed as the most electable candidate in the field by Democratic voters, had a major advantage over his rivals. This created an insurmountable obstacle for Sanders, whose explicitly democratic socialist ideology and far-left policy agenda were easy fodder for his opponents when trying to make the case that he was unelectable. Masket also shows how electability posed major challenges for other candidates, particularly female candidates and candidates of color, who are often perceived (erroneously) by voters as being more politically extreme and therefore less electable than white male candidates.

Masket’s analysis of why Biden prevailed over Sanders has important strategic implications for progressive electoral politics. It helpfully dispels the notion that progressive, and even democratic socialist, insurgencies within the Democratic presidential primaries are doomed to failure by the supposed class character of the party. Sanders fought in 2020 on a relatively equal playing field — but, given the centrality of electability to voters and party activists, as well as the broadly held belief across the party that Biden was more electable, Sanders simply lost. That fact does not indicate that future progressive insurgents would meet the same fate under conditions where electability was less important in voters’ electoral calculus. Given the likelihood that electability will remain a top concern for Democratic voters in 2024 (especially if Trump seeks the presidency again), and since there will likely be strong support for vice president Kamala Harris’s candidacy if Biden does not stand for reelection, the short-to-medium-term prospects of insurgent success in the Democratic presidential primaries are limited. That said, there is little reason to conclude that progressives should abandon contests for power within the Democratic Party, even (and perhaps especially) at the national level.

The Sanders Faction of the Democratic Party Will Endure, But It May Not Grow

Another important contribution to debates around progressive electoral strategy is Learning From Loss’s empirical demonstration that Sanders’s wing of the Democratic Party is a coherent faction that has sustained itself over multiple election cycles. A party faction, according to political scientist Daniel Silva, has four primary characteristics: ideological consistency, organizational capacity, persistence over time, and intent to affect a party’s political orientation. Clearly, the Sanders wing of the party is ideologically consistent (at least compared to other party factions) and intends to affect the Democrats’ political orientation. Masket explores whether it can also effectively influence a range of electoral contests, and if it has the capacity to do so over time. His analysis helps us to understand how likely the Sanders faction is to both endure and play an influential role in Democratic Party politics now that Sanders himself is no longer a candidate.

Masket finds clear evidence that the Sanders faction is enduring across election cycles as well as organizationally competent. To demonstrate factional endurance, he first examines donor data across the 2016 and 2020 Democratic presidential primaries. He finds that there was remarkable consistency in Sanders’s donor patterns. As shown in figure 1 below, among donors who gave in both the 2016 and 2020 Democratic presidential primaries, virtually none who gave to Sanders in 2020 were Clinton donors in 2016, and, with the exception of those who gave to Elizabeth Warren in 2020, the overwhelming majority of 2016 Sanders donors backed Sanders again in 2020. Warren, interestingly, drew more support from Sanders’s 2016 donors than from Clinton’s 2016 donors — but, given the perceived ideological affinity between Sanders and Warren among many Democratic voters, this is not particularly surprising.

Figure 1.

Candidate’s Share of Clinton Donors from 2016*

*The lower the number, the more the candidate drew from Sanders’s 2016 donors, and the higher the number, the more the candidate drew from Clinton’s 2016 donors. Source: Masket, Learning From Loss, 179.

Masket next examines staffing patterns among 2016 and 2020 Democratic presidential candidates, based on a data set consisting of 260 staffers who worked for Democratic presidential campaigns in both 2016 and 2020. He finds that, while Clinton’s 2016 staffers dispersed across the field of 2020 presidential candidates other than Sanders (with the plurality joining Warren’s campaign), the overwhelming majority of Sanders’s 2016 staffers (79 percent) returned to his 2020 campaign, with around 12 percent moving to Warren’s and a small handful shifting to other candidates.

Masket finds not only that Sanders activists were consistent across election cycles but also that Sanders-aligned organizations — particularly Our Revolution — had a major impact on donation patterns in 2017–18 gubernatorial races. Specifically, in primaries where Our Revolution endorsed a gubernatorial candidate in 2017 or 2018, Sanders’s 2016 donors were 50 percentage points less likely than Clinton’s 2016 donors to support mainstream party candidates. He shows, additionally, that giving patterns of Sanders’s 2016 donors in 2017–18 contests changed significantly after Our Revolution endorsements were made. This indicates that the endorsements themselves, rather than Sanders supporters’ predisposition to self-select into progressive gubernatorial camps, drove 2017–18 donor patterns among Sanders’s 2016 supporters. Such evidence suggests that the Sanders faction has sufficiently strong organizational capacity to meaningfully influence Democratic primary politics in contests where Sanders himself is not a candidate.

Yet the same evidence Masket offers to demonstrate the strength of the Sanders faction also points to important weaknesses. The Sanders faction’s experiences between 2016 and 2020 make it clear that a well-organized and ideologically consistent progressive minority faction can impact the contours of Democratic politics. This can be observed from a range of data points, from the emergence of Medicare for All and the Green New Deal as litmus test issues early in the 2020 Democratic presidential primaries, to Sanders supporters’ influence in the DNC’s 2017 Unity Reform Commission and the 2020 Biden-Sanders Unity Task Forces. For this influence to translate into a successful presidential nomination, large blocs within Democratic legislative caucuses, or significant policy wins outside of heavily progressive states like New York, however, the faction will need to grow. To date, there is little evidence that the Sanders faction can surpass a significant but still relatively low ceiling of support within the Democratic Party. For instance, Sanders was extraordinarily unsuccessful at wooing 2016 Clinton donors in 2020, relying almost exclusively on his own 2016 donors to sustain his 2020 campaign. Further, consistent with Sanders’s polling numbers throughout the 2020 primary season, Masket’s donor data show that Sanders’s share of Democratic primary donors never surpassed 30 percent, and that, among strong Democratic partisans, he never reached 10 percent support.

There are only two obvious paths to growth, and neither appears particularly likely, at least in the short term. On the one hand, the mutual hostility between the Sanders faction and the rest of the Democratic Party has made building allies within the party exceedingly difficult. On the other hand, Sanders’s failure to significantly expand the Democratic primary electorate among low-propensity working-class voters in 2020 makes attempts at a hostile takeover appear similarly unlikely. It is possible that a future progressive outsider candidate running in an election where electability is less salient could be more successful, but that would require either building stronger bridges to the rest of the party or offering a political program or campaign style that appeals more broadly to low-propensity voters who could make up for the candidate’s weakness among traditional Democrats.

Lessons From 2016’s Loss Will Shape Democrats’ Response to 2020’s Setbacks

I turn now to the broader implications of Learning From Loss for the future of Democratic Party politics. Masket’s analysis of post-2016 intraparty debates provides useful insights for understanding the party’s likely trajectory in the wake of its inevitable 2020 postmortem. In particular, he documents a widespread perception among Democratic Party activists and Democratic primary voters that Hillary Clinton was too focused on identity politics and unable to connect with white working-class voters, and that Clinton was a weak candidate, at least in part due to her gender. In turn, Masket conducted two survey experiments examining the extent to which exposure to identity politics narratives (narratives suggesting that identity politics are a political liability for Democrats) affects Democratic voters’ views of presidential nominees. With important differences across race and gender, Masket finds evidence that identity politics narratives make many Democrats less supportive of female candidates, candidates of color, and candidates supporting policies aimed at redressing group-specific inequities (such as combating workplace discrimination against women and people of color), while increasing their support for moderate white male candidates. Though he is careful to note that no consensus view emerged among Democrats to explain Clinton’s loss, Masket’s findings suggest that the combination of pervasive narratives around identity politics and a laser-like focus on electability convinced many Democrats that their best bet in 2020 was a “safe” candidate — specifically, a centrist white male candidate.

The results of the 2020 general election, as well as early reactions by Democratic politicians, suggest that similar dynamics will characterize intraparty debates over the coming months and years. That is, the party’s strategic orientation in 2022 and 2024 will likely evince a similar logic to its post-2016 orientation. Biden’s victory, on the back of his remarkable success among suburban white voters — combined with his stronger showing compared to many (often more progressive) down-ballot candidates in red and purple states — is likely to confirm the wisdom of the party’s decision to nominate a centrist white male presidential candidate. This is especially likely given that electability will remain a central concern of Democratic voters in the post-Trump (and potentially future-Trump) era. To be sure, Biden’s choice of Kamala Harris as vice president, and his commitment to gender and racial diversity in his cabinet, demonstrate the competing pressures party leaders face to deliver for key constituencies in the Democratic coalition (especially women and African Americans). Overall, however, the 2020 election will likely be viewed by most Democrats as a vindication of the party’s risk-averse electoral strategy in 2020.

Learning From Loss studiously avoids the question of which post-2016 loss narratives were the most or least accurate. Indeed, Masket emphasizes that political science models of the election highlighting economic growth, international security, and party incumbency suggested the election would be a toss-up. As a result, he cautions against placing undue causal weight on any contingent factors related to candidate characteristics or campaign messaging. Nonetheless, many of the intraparty debates sparked by Clinton’s loss remain central strategic questions for the party, and it is worth briefly considering their empirical merits in light of the Democrats’ (at best) disappointing electoral performance in 2020.

For instance, Masket recounts debates between progressive members of the DNC’s 2017 Unity Reform Commission (tasked with proposing changes to the party’s rules and procedures that would address complaints raised by Sanders and his supporters during the 2016 presidential primaries) about the relative utility of focusing on bringing white working-class voters back into the Democratic coalition versus mobilizing people of color who did not turn out in 2016. With the dust settling from the 2020 presidential election, it is increasingly clear that neither of these strategies has been particularly effective. Indeed, despite impressive turnout gains in communities of color during the Georgia Senate runoff elections of January 5, 2021, Democrats’ most important support gains in 2018 and 2020 came from the ranks of affluent suburban voters. As figure 2 shows, across key battleground states, Biden’s advances over Clinton in wealthy suburban areas were consistently higher compared to areas with a large African American population.

The implication is clear: not only are Democrats likely to believe they should double down on centrist candidates who can appeal most successfully to the median white suburban voter, but they appear to be objectively correct in that assessment. Yet, as many commentators have pointed out, the prevalence of split-ticket voting in the suburbs suggests that much of Biden’s support in these areas reflected a rejection of Trump, not an endorsement of Democrats. With Trump off the ballot in 2022 and potentially in 2024, will a red-state and purple-state strategy based on minimizing programmatic differences between the parties be enough to hold erstwhile conservatives in the Democratic fold? Beyond this, will the strategy be up to the task of defeating vulnerable Republican senators and defending the Democrats’ House margin in 2022, when it was barely able to do so in 2020 with Biden at the top of the ticket? Perhaps not — especially if Republicans can effectively weaponize connections between prominent members of the party’s left flank and centrists struggling to hold on in competitive congressional districts.

Figure 2.

Shift in Clinton − Biden Vote share by County Type (Counties where Biden outperformed Clinton)

Source: Author’s calculations. Electoral data from Dave Leip’s Atlas of US Presidential Elections (uselectionatlas.org); demographic data from the American Community Survey.

Despite these problems, it may be the case that Democrats simply have no promising strategic options in the short term, and that repeating the lessons learned from 2016 in the post-2020 period amounts to the least bad alternative. I want to suggest, however, that there is one approach the party has largely bypassed, and that such an approach might be an effective response to the supposed trade-off between progressive policy and electability explored in Learning From Loss. There is substantial evidence that candidates perceived by voters as being extreme tend to underperform candidates perceived as moderate, but it is less clear how voters’ perceptions of extremism vary across different policy positions and message framings offered by candidates. For instance, the results of Masket’s survey experiments suggest that many voters perceive progressive policies aimed at addressing group-based inequities as being more extreme than, say, a non-group-based policy to increase high-paying jobs through economic growth, and subsequently punish candidates who hold the more progressive position.

While not tested in Masket’s experiments (since supporting economic growth is not an unambiguously progressive policy), a progressive economic policy platform presented in relatively neutral terms — and one that is relatable to different working-class constituencies (urban/rural, white/black/Latino) — could be perceived as less extreme compared to alternative progressive platforms. Consequently, such a platform could be more effective in appealing to disaffected white working-class voters as well as low-propensity working-class voters from all demographic groups. There is little doubt that large majorities of voters — including in red and purple states — support a wide range of progressive economic policies, from free universal health care to a $15 minimum wage. Yet, as the decisive victory of progressive ballot initiatives in states carried easily by Trump in 2020 suggests, Democrats are not effectively translating these preferences into votes. Part of the explanation for this is likely attributable to the fact that many voters associate Democrats with progressive policies they perceive as less appealing than bread-and-butter economic policies. Alternatively, some voters may not trust Democrats’ progressive policy commitments because they associate Democrats with elite interests opposed to those of ordinary Americans.

A fascinating recent report on public opinion in rural America suggests that part of the Democrats’ problem reaching working-class voters lies in messaging and style. The report tests the appeal of different messaging frames on rural voters’ political preferences. It finds that, by relating progressive policy items to the experiences and values of working-class communities (in this case, rural communities), progressives might garner higher levels of support among constituencies they would otherwise struggle to reach — without losing support among their traditional base constituencies. Future research (and future electoral campaigns) are needed to better understand which types of policies and messaging frames can be deployed most successfully by progressive candidates among working-class voters, but there is reason to believe that the trade-off between progressive policy and electability may not be as stark as it is presented in Learning From Loss. Democrats in general, and progressives in particular, may yet have a path to electoral success in red and purple states that does not depend primarily on sustaining and growing their support among affluent suburban voters.