Ever since the New York Times needle faltered on election night 2016, the institution of polling has had a bad rap. From skewing too far in favor of Democrats to failing to accurately represent the opinions of members of non-white demographic groups, the polling community has made significant errors that have caused many people to lose faith in the veracity of opinion polls. Progressives in particular have had a contentious relationship with polling — the role of progressive organizers is to stretch the bounds of public consensus, and the role of polling is to reflect where that consensus currently is. But as our nation becomes increasingly polarized, it is crucial for decision-makers and political leaders to understand how people across the country think about the important political issues of our time. Like the Democratic Party itself, pollsters need to build trust with the American people once more.
At Data for Progress, where we work, we don’t see polling and progressivism as at odds. In fact, we believe polling is one of the most grounded, democratic aspects of politics. Indeed, when leveraged effectively, polling data is a useful tool for the progressive movement — it reflects what the average voter supports and what their policy preferences are. Data for Progress is committed to building progressive organizers’ trust in polling and using polling to inform and guide the progressive movement. In our day-to-day work, we use polling to support candidates and movement partners in three major ways: to help persuade voters that our agenda is popular, to shape effective public messaging, and to inform electoral consolidation to ensure progressive victories.
Politics is a game of addition. To pass policies and win elections, you need to persuade people to join your cause — everything from a local school board race to a presidential campaign requires you to build a coalition of supporters. By demonstrating through polling that certain progressive policies are popular, movement leaders can use this data to persuade elected officials conscious of reelection to get on board with their causes. Data for Progress polling regularly shows that progressive policies — from creating a Civilian Climate Corps to legalizing marijuana to reducing prescription drug prices — have broad support across the country. From frontline districts to deep-red states, polling indicates that voters want their legislators to back some key progressive priorities. Movement leaders and allies can take this polling data to hesitant elected officials and show them that leaning into progressive policies is not a political risk — it’s politically wise. And when elected officials fail to get behind progressive policies that our polling shows are popular, it can provide organizers a compelling and persuasive line of argument against an out-of-touch incumbent they’re seeking to replace (the successful playbook leveraged by many members of the Squad).
While we might hope that our progressive policy priorities always enjoy wide margins of support among voters, that’s not always the case. When our data show that a policy is unpopular, this helps organizers and elected officials identify where there is still more work to do in shifting public opinion. By identifying the most and the least popular policies, we can help social movements allocate their limited resources effectively and address the areas and constituencies with the most potential for grassroots support.
To engage skeptical constituencies and expand their base, social movements need to optimize their public messaging to maximize support. Time communicating with the public is a limited and precious resource — movement leaders and allies must be wise about what they say when they have the public’s attention. It’s important not only to tailor the issues discussed on a campaign to the targeted portion of the electorate but also to frame those issues in a way that will resonate with voters.
For instance, our polling has shown that messaging climate policy through the lens of job creation and clean air and water is one of the most effective ways to win over marginal voters. That’s not to say that messaging about other components of climate policy, like reducing carbon emissions or improving public health, is not effective in swaying voters. It is. But, if you only have a few minutes to connect with a voter, be it during a debate, through a TV ad, or at their door, you’d be wise to use that limited time to talk about issues and policies in the most popular way.
Data isn’t just helpful for persuading skeptical audiences and shaping public messaging — it’s also an invaluable tool for helping to consolidate around the candidates who have the best chance to win elections. Horse-race polling isn’t only for generating headlines and changing electoral narratives. When applied effectively, it can help inform strategic decisions about how progressives should consolidate their electoral power to maximize their chances of winning.
Take the Massachusetts 4th congressional district as an example. In August 2020, Data for Progress conducted a poll of the open-seat race happening in the district. We found that the most conservative candidate in the field, Jake Auchincloss, a former Republican political organizer, was likely to win by an extremely close margin, with Jesse Mermell, a strong progressive candidate, trailing closely behind. Our data showed that if the worst-polling progressive candidates dropped out in order to consolidate the progressive field, we might have been able to eke out a win. However, the field did not consolidate, and the conservative candidate won the race in a result that almost exactly mirrored our polling predictions.
On the flip side, take our recent polling of the Manhattan DA’s race in NYC. We found that Alvin Bragg was the most viable progressive candidate, neck and neck with a corporate Democrat who tried to buy herself the seat via self-funding. After our poll showed that Bragg was the leading progressive contender, he gained tremendous momentum and pulled away from the pack — culminating in his resounding victory. It’s incredibly encouraging that so many progressive candidates are running for office, and polling can help inform pragmatic decision-making to ensure that progressives in fractious primaries have a clear path to victory, especially in deep-blue districts.
While polling is a critical tool at our disposal, it isn’t a panacea for all of our political problems. Sometimes we find that a certain policy priority isn’t very popular or that a certain way of speaking about progressive priorities doesn’t resonate. In the grand scheme of our progressive mission, that’s okay, and it doesn’t mean that we shouldn’t pursue those priorities. Republicans don’t run on unpopular policy — rather, they campaign on the most popular items in their agenda and then enact the unpopular ones once they hold power. We believe Democrats should adopt a similar approach: campaign on the most popular parts of our agenda and then use the political capital we gain from implementing popular things to also push for potentially less popular policies that are important, nonetheless.
By the same token, Data for Progress hasn’t always gotten it right. Even though we were rated the most accurate pollster in the 2020 Democratic presidential primary, we still made mistakes in our 2020 polling. In our recently released 2020 polling retrospective, we reflected on those errors and identified gaps where we can improve for the next election cycle. We realized that there were three key phenomena we didn’t take into sufficient account: the relative enthusiasm of Democrats to take our surveys, the strong effect of where voters live on their behavior, even when you control for other demographics, and the role of urbanicity in determining the degree of electoral swings. We’re fixing these issues by improving our weighting with zip codes, using more language that appeals to conservative-leaning voters, and limiting likely Democratic activists from taking our polls.
The hard, transformative work that progressive organizers and elected officials are doing to enact systemic change requires reliable and actionable data to support it. In a political climate in which the odds are stacked against grassroots movements and the people who power them, it’s crucial that we don’t just work hard at persuading people to join the progressive movement but that we also work smart. Polling helps us do just that. Instead of a vertigo-inducing pollercoaster, we see data and polling as the tracks for the progressive movement’s bullet train, taking us into the bolder and brighter future we deserve.
