The Art of Losing: Me

I was wrong. Here is why.

(Credit for the voiceover for this article goes to the great Samuel Lipson)

When I first decided to write articles titled “The Art of Losing,” I did so with the knowledge that I would eventually have to write one about myself when I got an election wrong. I knew that this was inevitable, so if you went back a few months or a year and told me then that I would be writing this article now, I wouldn’t have been all that surprised.

What would have surprised me, though, is that I would be writing it in reaction to the specific set of results we just got: one in which a Biden administration member underran the rest of their party and lost at the same time that most down ballot Democrats won in the country’s most competitive states. I had feared that exact possibility from the moment I first started writing about the 2024 election, all the way back in the spring of 2023. I didn’t buy the then-existing consensus that Democratic success in the 2022 midterms was necessarily a good sign for the White House, thought that even an economic boom might not be sufficient to fix Biden’s possibly permanently broken public image, and believed that only a nominee from outside of his administration would be actually well positioned to tap into the success Democrats had been seeing outside of him.

Had I stuck with this thinking through the end of the race, I would have ended up closer to the actual results. But when that time came, I did not. Because of that, I was wrong. Full stop. You will never hear me say that I “actually” forecasted the results due to what I said in prior articles or because I didn’t say that Harris was a 100% lock to win. The only reason I mention what I said earlier in the cycle is so I can explain what brought me from saying that there was “no single person on Earth doing more to assist the Republican Party long-term than [Kamala Harris]” to calling her the favorite to win in my final map. That story, in a word, is this: I prioritized a particular list of pre-election signals, all of which correctly indicated the results of previous Trump elections but failed to do so this year. Here is what they were, why they were wrong, and why I thought that they would be right.

Special Elections

During my earliest days of writing about the 2024 election, the first non-polling signals I looked at were special elections: the off-cycle state and federal elections that occur periodically around the country in order to fill vacant seats. Those sorts of elections rarely received coverage from mainstream pundits, but I thought they were worth paying attention to because of my experience following the 2022 elections. That year, a series of Democratic overperformances in Congressional special elections held in the aftermath of the Dobbs decision were what convinced me that the midterms would not be a red wave, which ended up coming to pass. I thought that these races would continue to be useful for the 2024 election for the simple reason that they had been so in every election that their results had been compiled for. Going back to 2018, the average partisan swing of all of the special elections held in the two years before each November election always ended up being very close to the final national results. This was the case in years where the polls were right, like 2018, but also in 2020, when pollsters saw historically large errors.

What made me believe such a trend could be more than just a coincidence was that it fit perfectly within what one would expect as a result of the preeminent trend over the last decades of American politics: polarization. While local races like special elections might have been dominated by local issues during prior political eras, those kinds of unique considerations had become steadily less and less important to how voters were making decisions at all levels in cycle after cycle. Instead, voters had become more purely partisan, voting for candidates of all sorts just based on the R or D next to their name. This had made American politics a lot less interesting, but it also seemed to allow local races to serve as proxies for the overall national environment, with data coming straight from the voters themselves. And over the course of 2023, those of us who had been following the special elections since the end of the 2022 midterms began to notice something quite surprising. The special elections during that year were collectively showing results consistent with a national Democratic advantage, even stretching into the double digits.

In early 2024, however, Nate Cohn of the New York Times would drop a bombshell article that called the usefulness of these races in projecting the national environment into serious question. As he explained it, the data from his polls on the presidential race had found a massive difference in attitudes between the most and least engaged voters in the country. According to his data, there were massive differences in voter preferences between voters who turned out in special elections and voters who only turned out rarely, if at all. While Biden had lost practically no ground with voters who voted in low turnout races like special elections, he was seeing substantial losses among voters who only turned out rarely.

It was a finding that explained practically all of the discrepancy between the strong Democratic numbers in special elections and the strong Republican numbers in polling. While the former included only high-turnout, highly-engaged voters, the latter included voters of all stripes. To the extent that special and off-year elections could be said to represent anything at all, it was just that Democrats were holding strong with voters who always vote—possibly a strong sign for their chances in something like a midterm, but not necessarily bullish for them in a high-turnout presidential race. Cohn would get quite a lot of shit for this analysis, all of it coming from partisan Democrats who had spent the past year hanging on to the results in the specials as a reason to hold out hope that Biden was on track to win despite what the polls said. The posts calling him arrogant and his numbers unrealistic were constant, but none of them were convincing. Cohn, on the other hand, provided a tremendous amount of backing for his claims, even going as far as to perfectly model the result of a strong special election for Democrats in Florida using nothing but data from his polls showing Trump ahead nationally.

Since it was hard to think of any reasons as to why he may be wrong, I would publish a new article arguing that the new best-case scenario for Democrats was that this gap between hyper-engaged and unengaged voters might be expected to dissipate when the election came closer and unengaged voters tuned in. Cohn would publish an article making a similar point that spring. Past that point, I would only reference special elections as a positive sign for the party in conjunction with other indicators presenting similar signs, which would eventually come in spades.

The Washington Primary

Around this time, I would become increasingly convinced that the worst-case scenario I had feared when Biden began his reelection campaign—that he was a uniquely incapable candidate—had come to pass. As I put it in my article going over the state of the presidential race at the start of the year:

Biden, based on all the information we have right now, is a uniquely weak presidential candidate. It’s the only honest conclusion you can come to…the simple fact is that voters want to vote for Democrats, especially against Trump, but they’re also very reluctant to vote for an octogenarian they broadly dislike, see as a failure, and don’t think is capable of governing.

For the next seven months up until Biden dropped out, this analysis was what I used to square the circle between Democratic success in 2022 and Biden’s own polling struggles. The release of the special counsel report calling him an “elderly man with a poor memory” convinced me that practically anyone—even the poorly-polling Kamala Harris—stood to be a better candidate than him. As matchups for the U.S. Senate became set and new polls began to consistently show Democrats down ballot outrunning Biden, I posited that it showed that his age and record of presiding over chaos at home and abroad was preventing him from appealing to a mass of voters didn’t necessarily love Trump, but disliked him. While Trump’s 34 felony convictions in late May seemed to be a possible turning point, it would ultimately only have a marginal impact, raising serious questions as to what might be needed for that voter pool to “wake up” and start shifting left. Although the debate stood as a potential wildcard due to Trump’s past history, it was becoming hard to say what more people needed to see to remember that they hated Trump more than Biden.

Around this time, many commentators would work overtime to explain why Biden looked set to lose without addressing the elephant in the room. Some said that his unpopularity was due to wokeness tainting the Democratic brand, an explanation that flew in the face of the strong numbers that the party’s down ballot candidates were seeing. Others relied on often-cherrypicked data to declare him an innocent victim of a malevolent media environment. In my mind, the most likely explanation was that he was simply too old, and presided over too much chaos, for voters to have confidence in him. At the end of June, the debate (AKA my previous worst prediction) would put this in the sharpest relief possible. While Trump wouldn’t put in a spectacular performance, Biden’s 90-minute medical episode would end up stealing the show and end any chances he had of winning a second term. A push to force him out of the race would begin before he even stepped off the stage.

I went to bed that night feeling the most optimistic I had felt about the election for as long as I could remember. With two months left until the DNC, there seemed to be some runway for the party to nominate a candidate with a profile resembling their well-polling nominees for Senate: young, experienced, and not connected to an administration that had been languishing at Jimmy Carter-level approval for three years. When the consensus became that Kamala Harris was the only possible replacement at such a late date, whether it be over worries about where Biden’s funds would go or outright lies about state ballot access, I disagreed. To me, the benefits of a nominee representing a clear break from the Biden administration outweighed even the prospective chaos of a contested convention. While I ranked Kamala above the likes of Gavin Newsom when considering potential Biden replacements, I also had her below alternatives with a more demonstrable appeal to swing voters.

The next few weeks would be Hell. Rather than facing reality, Biden and his toadies would spend weeks engaging in a doomed battle to stay on the ticket, choosing to let their bid slowly and painfully die in public. When he finally quit the race, the sheer amount of time he had wasted froze out even the most optimistic plans for a non-Kamala replacement. Within minutes of this, the party instantly consolidated around her out of fear of further chaos. It was the biggest, most dramatic move in presidential politics in half a century, and the net effect of it was to replace the losing nominee with someone with all of his problems besides age.

Still, just this one fact would end up making a difference. After being stagnant for nearly an entire year until crashing towards Trump following Biden’s debate, the race would rapidly move leftwards immediately following Kamala’s ascension. It caught me off guard, just as it did practically everyone else. Hardly anyone had taken her seriously as a national figure following her flop in the 2020 primaries and nonexistent presence as Biden’s VP. When she gave her first speech after being named the de facto nominee, it amounted to the first time many of us had heard her speak in years. It was a complete revelation compared to Biden, even if it was only because she could speak in complete sentences and mentioned things besides NATO and AUKUS as campaign issues. Many voters who had been unwilling to back Biden agreed. Polls conducted within days of the swap would show the Vice President suddenly well within contention for an electoral college majority. Despite having done practically nothing to define herself, and despite being connected to her administration at the hip, just the basic fact that she was not Joe Biden appeared to be enough to turn the election into a real race once again.

It was a dizzying two weeks. By the end of it, even the most right-leaning aggregation sites would show Kamala up in their averages, something they hadn’t seen for Biden since September of 2023. Most encouragingly, these gains occurred predominantly among the exact Democratic-leaning voters that Biden had been unable to win over. As Nate Cohn put it following the release of a series of New York Times/Siena polls that showed her leading in Michigan, Wisconsin, and Pennsylvania:

If you drew up a list of President Biden’s challenges this cycle, you could probably find a demographic group corresponding to each one on this list of Ms. Harris’s biggest gains. There’s young, nonwhite and low-turnout voters, and the places they tend to live. There’s the lowest-income voters, who suffered through rising prices. There’s even the TikTok users immersed in the bad vibes of the Biden era. The Muslim and Arab voters angry about the war in Gaza don’t make the list, but only because of their small sample size (just 55 respondents in August) — they would have been No. 1 on the list with a net swing of 49 points toward Ms. Harris.

The sense was that, under this new candidate, the race had turned back to “normal,” with Democrats unburdened by Biden’s unpopularity and now able to benefit from the country’s disdain for Trump. That it happened despite Kamala not doing much of anything at all other than entering the race appeared to be doubly encouraging. With the DNC and an upcoming debate standing as possible opportunities for further gains, the sky seemed to be the limit. Suddenly, Democrats had moved on from hoping that the polls were somehow wrong to hoping that they were right. And fortunately for them, that month of August would provide them with a seemingly perfect test if they were: the Washington primary.

Although only directly relevant to those living in Evergreen state, the few election cycles preceding 2024 had made Washington’s August primary something of a legend among those who follow elections very closely. They fell into a very rare category of pre-election indicators that had both foreshadowed the countervailing misses in 2020 and 2022 without being discredited by Nate Cohn’s analysis. And when the results came in, they showed something incredibly encouraging for Democrats: a down ballot environment better than that of 2020. Even after removing the nationally unrepresentative Seattle metro area from the equation, the rural and predominantly white parts of Washington still put in their most Democratic numbers in years. This was easily the most encouraging aspect of the results for Democrats. While unrepresentative leftward shifts in the Seattle area had caused Washington state at large to miss larger movement towards Trump nationally in 2016, the large gains his party saw in the rest of the state did prove to be representative of his gains among white working class voters nationally. Now, it didn’t show that at all, seemingly providing credibility to the polls showing Kamala up in the election-deciding Great Lakes swing states.

It all gave me a degree of confidence that the Democratic strength we were seeing in those states was real instead of just being a product of 2020-esque nonresponse bias. Given that special elections were said to be no longer representative enough to disprove that that was the case, I had always thought of another nonresponse bias miss as something of a live wire, but this indicator with far higher turnout showing that it wasn’t in the cards stood as a strong sign against it. And, as fate would have it, the primary would end up being quite predictive—but just only in Washington. After years of shifting back and forth in line with the rest of the country, Washington in 2024 trended intensely leftward. Although votes remain to be counted, Kamala Harris’ overall vote share in the state is higher than what Biden received in 2020.

It’s not even just that Kamala Harris would have won the election if the rest of the country behaved as Washington did. It’s that she would have won it easily, carrying perhaps every single swing state. As of now, Washington is set to be one of the single most Democratic states in the entire country, with a result solidly to the left of states like New York and Illinois. As for why this one state specifically bucked every single trend seen in the rest of the country, God only knows. Perhaps the Biden administration’s messaging that funding wars overseas boosted the economy won his party support among Boeing workers but nobody else.

Prioritizing 2022’s Best Pollsters

Despite this promising-if-ultimately-unrepresentative sign for Democrats, the Washington primary results didn’t fully convince me that Kamala was a solid favorite. Although her early moves appeared to indicate that she had something of a sense of what she needed to accomplish to win the election—i.e., establishing herself as something of a generic Democrat separate from the Biden administration—a noticeable shift occurred right after she picked Tim Walz as her running mate. At the DNC shortly thereafter, the entire first day was dedicated to honoring Joe Biden, throwing away what should have been the opening effort to introduce their very new nominee to the country. Things hardly improved from there. Rather than presenting an affirmative message for their own agenda, the party would use its own showcase event to highlight Republican speakers and demonstrate its fealty to Republican immigration policies. When the Vice President gave her acceptance speech, it was heavily oriented around inoffensive biographical details and attempts at what would later be referred to as “vice signaling” meant to show that she was not, in fact, woke.

Centrist pundits—many of whom are now backpedaling to declare that Trump’s win a result of liberals refusing to give up 2019-era wokeness—absolutely loved it. Some said that their only concern was that they liked it too much. But when I finished watching the proceedings, it made me feel extremely nervous. Even assuming that she would get at least some kind of bounce just by introducing herself didn’t stop me from worrying about what her rhetoric represented. Not only was there the obvious rightward swing on policy, but the overall approach of the event was deeply and disturbingly reminiscent of the Biden campaign. The campaign was jumping at any chance to move to the right, no matter how tenuous the justification, while outwardly rejecting gift-wrapped chances to explicitly break from Biden on some of his worst issues. Instead of the new party focused on winning that was promised, we got something far more familiar: a self-indulgent elite putting its own feelings and preferences above all else.

In retrospect, I wasn’t negative enough. Not only did Kamala fail to see a boost after the convention, but her numbers would outright slide downwards. It was a clear sign that the campaign had entered a new stage, one in which the benefit of the purely superficial signaling that dominated the DNC had been exhausted. The remaining voters to be won over were looking at her with a far more critical lens than the demographic that had moved towards her side by virtue of her not being Joe Biden. But instead of shifting into a higher gear, her campaign would behave incredibly sluggishly. She would only sit down for her first sit-down interview in late August, more than a month after she became her party’s de facto nominee. Her website wouldn’t publish a policy page until September 8th, less than two months before Election Day. Instead of a break from the status quo, it appeared to be a literal copy-paste of the sitting President’s agenda, complete with unedited metadata referencing “re-electing Joe Biden.”

These were all small things, none of which might have swung any votes by themselves. But when put together, they added up to a portrait of a campaign that just wasn’t on the ball. My new concern, as I laid out in an article prior to the first and only Harris-Trump debate, was that this approach could cost her the election in spite of all of the positive signs Democrats were seeing by making her fail to qualify as a “generic Democrat” in the minds of voters. This potential outcome—wherein the top of the Democratic ticket failed to replicate the success their party had seen in 2022 and was seeing in down ballot polling throughout 2024—was exactly what I had feared would happen with Biden for more than a year, and it would end up being what came to pass on Election Day. Although the rest of the party wouldn’t reach the strength implied by the numbers put up by Washington Democrats, they would still universally outrun her numbers, ultimately winning at least one election in no less than five Trump-won swing states. Had the party managed to carry those states at the national level, they would have reached 270 electoral votes and won the election.

Of course, just that I warned of this during the campaign doesn’t mean that I got the election “right.” This specific outcome wasn’t my final prediction, and while I would still publish pieces criticizing her strategy over the following months, the bulk of my coverage after the debate was more bullish than bearish. My reasons for this were twofold. First, the first debate ending with Trump ruling out a rematch suddenly turned an election that had just been dominated by countless impending live-wires into one with no scheduled big moments left until November. My understanding of this was that it put us in a position where we generally knew what the main dynamics of the election would be, and that these dynamics favored the Democrats. I thought that, for as flawed as her campaign was, her success at the debate had moved her towards achieving the basic task of her campaign: making herself appear like a credible leader to the country’s most important swing voters. Post-debate numbers—most notably from the New York Times, which I personally held in high esteem after their unique success in the 2022 midterms—seemed to confirm this, at least in the all-important Midwestern swing states.

Second, I was already anticipating that October would inevitably see improvements in Trump’s polling numbers, and this would end up meaning nothing. I had seen exactly that occur two years prior, when firms on record denying the veracity of everything from the 2020 election results to the Holocaust had pumped out surveys showing shock Republican gains, scaring many mainstream pollsters into following their lead. Given that the 2024 race had been defined by stagnancy for most of its existence, I expected that whatever rightward shifts we saw in October were more likely to be the result of right-wingers aiming for Fox News appearances and untrustworthy firms trying to hedge than a reflection of any real shifts in the electorate.

So, if I wasn’t willing to trust the results from Emerson and RacismElections, who did I decide to look to for reliable information? To me, the answer appeared to be the rare pollsters who had been willing to buck the red wave narratives in 2022 and publish surveys that proved to be accurate. Nationally and across the swing states, this meant pollsters like the New York Times and Marist College. At the state level, this included select local pollsters that either didn’t fall to local red wave narratives and correctly depicted the environments in their states. I chose to prioritize this list of pollsters not because I thought they had some kind of secret sauce, but because the numbers produced by GOP-aligned pollsters in 2022 were so egregious, and the ways in which many nonpartisan pollsters played me-too with their numbers were so blatant, that I found it hard to trust the veracity firms without a proven record of post-2020 accuracy.

After a brief polling drought at the start of the month, these pollsters would finally begin to start publishing new surveys. And as fate would have it, these good firms showed largely strong numbers for Kamala, starting a pattern that would persist until the end of the cycle. The Times national poll, published on October 8th, found her up by four points among likely voters, the first national lead the paper had found for any Democratic candidate against Trump the entire year. Most importantly, they said that this was because their results found her improving exactly where I always believed she needed to improve: representing change. The Times would follow this up with another poll showing her up by four points in Pennsylvania, seemingly further validating my read of the race. Around the same time, Marist would publish a poll showing her up by five points nationally.

It was all quite a different story from the numbers and narratives coming from the rest of the political world. After presenting a tied race for months, betting markets started swinging intensely to Trump in the middle of October. Leading statistical forecasts, where her chances had been dropping since the end of September, started to show Trump as the nominal favorite once again. It gave me a strong sense of déjà vu. The narrative shifts we were seeing were the exact same that I had seen in 2022, almost down to the day that they started to come out. My take on it was that it primarily reflected an industry scared of ever going out on a limb for Democrats again following their 2020 misses—something that, if still true, would mean that all of the movement we were seeing for Trump meant practically nothing.

This was the argument that I made in my article presenting the case that the polls were on track to underestimate Democrats. Forming the basis for this beyond just precedent in 2022 was another series of articles by Nate Cohn that went over the state of play in the polling industry and how firms had changed since the 2020 election. All of the changes he said he had made had the effect of increasing Trump’s support, which I thought could have stood to overclock his numbers if it turned out that they were being too reactive to pass misses. It was a phenomenon with global precedent; as Nate Silver brought up in his newsletter, this exact fear had caused British pollsters to dramatically underestimate the left in the 2017 parliamentary election in the United Kingdom.

After putting this case together, I decided to enter wait-and-see mode before publishing my final predictions, keeping an eye on the upcoming final results from the country’s most reliable firms as the best way to accurately gauge Kamala’s strength in swing states. As this drip-drip-drip of polling came out, things were generally very positive for the Democratic side, especially in the must-win Midwest. Marist’s final slate of midwestern polls had Kamala up between two and three points in Michigan, Wisconsin, and Pennsylvania. Marquette’s final poll of the Badger state had her up there by one, while two polls of Michigan from firms that nailed the state’s 2022 results found her up by three. At the same time, all the polls giving Trump his best numbers were the usual suspects, either aligned with the GOP outright or with strong post-2020 records of overcautiousness that had often resulted in them overestimating Republicans. The same split-screen that we had seen in 2022 looked as if it was appearing once again.

And that was before the Selzer poll. Of all of the polls I had paid special attention to because of their success in 2022, there was always a lingering worry surrounding them: the fact that they had underestimated Trump during his runs before. For as much as I believed that those errors were best explained by specific factors unique to those years (something that it seems was, in fact, the case—the errors this year were not nearly the size they were in 2016 and 2020 even among pollsters that didn’t weight by recalled vote), the specter of this happening again was always a lingering worry of mine. This was exactly why I had always put a premium on things like the Washington primary—they had served as early signs that the polls would be off multiple times in the past. And in this opera of canaries in the coal mine of the polling world, none had been quite as good, or had such illustrious of a reputation, as J. Ann Selzer. She had managed to find Trump strength in her state that the averages missed in 2016; when they missed his strength again in 2020, she did so again. And unlike the Republican-aligned pollsters who had clearly only managed to find numbers in line with the results by putting the thumb on the scale for their party, Selzer had managed to near-perfectly estimate the results in her state in 2022. There was no sign of any systemic bias, no history of missing silent Trump voters or bedevilment by nonresponse bias.

It was the last plausible gold standard we had. And when it came out, it was the best individual poll result Democrats had seen in decades.

With a D+3 final result, Selzer didn’t just provide no evidence that pollsters were missing Trump support, as she had in the past. She strongly indicated the exact opposite was happening, and it hit the political world like an atom bomb. Democrats celebrated a potential landslide. Republicans accused her of being bribed by P. Diddy to demoralize their voters. One pollster that had shown strong results for Trump published a preemptive mea culpa for underrepresenting core Democratic constituencies in his likely voter model. My take on it, as reflected in my final ratings article, wasn’t that it represented proof of a Democratic landslide or even a win in Iowa, but that it was a strong sign for the party in the Midwestern states that they needed to win. The following day, the New York Times would release a set of surveys showing Kamala leading or tied in every swing state but Arizona, while Jon Ralston would project her to be the narrow victor in his state of Nevada.

All of that was what got me to my final map: one that I said leaned in the Democratic direction. It didn’t predict a blowout or rule out a Trump win by any means, but it was still wrong all the same.

---

If you’re still here, thank you. Even if you’re hatereading. As for what this newsletter will become, I hope to return to the long, retrospective pieces a la the Stacey Abrams article from 2023. I can finally start working on that long-awaited Bernie piece. More analysis of the failures of the Biden years, too. If it’s not too depressing, maybe some analysis of Trump’s cabinet and legislative priorities. And if nothing else, I hope to be a counterweight to the centrist media that are already salivating about returning the Democratic Party to 1997, if not 1897. 

Regarding forecasting, things like the Washington primary can no longer be assumed to be consistently useful outside of Washington, and the gold standards—Selzer, Marist, the New York Times, Ralston and the like—are a lot less gold now. Broadly, polling is in a strangely familiar state. The best did the worst, the worst did the best, and the industry as a whole did basically fine. There is no existential crisis on alert for either polls or pundits, so it's best to start following aggregates again. Congrats, Nate Cohn; RIP, J. Ann, especially with how awful her explanations for her miss this year have been. 

As for what the past four years taught us, the main story is this: the Biden administration was a test case for the viability of establishmentarian, hawkish liberalism, and it failed. He, Kamala Harris, and their White House saw an army of reactionaries and a completely captured judiciary gut the rights of people across the country and either did nothing or outright capitulated, and lots of people who praised both of their strategies for representing actually existing sanity are now backtracking to pretend that they ran on defunding the police and open borders. It’s just further proof of what we’ve already known: that the strategy offered by the Atlantic writers of the world is center-right ideological hackwork disguised as Hard Truths, as well as a way to guarantee Democrats hurt as many people as possible on the way to their next defeat. Even if there isn’t a clear ideological throughline for which Democrats managed to outperform the top of the ticket the most, one side was the one in charge here, and they need to own it the same way they were positioning to own a potential victory. It begs the question: did they care more about winning or winning on their own right-wing terms? I’ll let you think about Bill Clinton’s speech in Michigan, Walz’s debate slip-up about Iran, and all the campaigning with the Cheneys to reach your own conclusions there.

Going forward to a fascist regime that will destroy the country at worst or just keep chugging along against a Biden-less Democratic party at best, the deciding question of the future will be what shape the opposition takes, just as it was the deciding question going into 2020. There will probably be no #Resistance this time around, which is depressing for the center-left but liberating for the rest of us. People are more introspective than outraged, which is a good thing. We have a window of opportunity to change what has facilitated the success of the self-destructive centrist class that brought us to this moment: their ability to establish a narrative that they are always the best bet at the ballot box. Their failure this year—and it is indeed their failure—needs to change that. Those of us who want to create a better world will need to reiterate that their vision of the world is neither popular nor “common sense.” We’ve been able to show how they’ve been wrong morally; now, it’s time to break the spell that their approach is not only good electorally, but acceptable at all.