In 2008, I played a small part in coordinating a multimillion-dollar prime-time infomercial, the visual “closing argument” for then-Sen. Barack Obama’s presidential campaign. Although Obama was a skilled campaigner running a historic race with a strong message, the thinking behind the ad was that it would help seal the deal with voters. You might remember it: With sweeping camera shots and meticulous lighting, Obama was presented simultaneously as a barrier-breaker and a bastion of American tradition. Indeed, he was. It was compelling television, and the presentation was a staggering display of financial wherewithal in an era in which campaigns had unalloyed faith in TV’s power to move votes.

Those days are over — or at least they should be. As Internet and mobile media have blossomed, fewer Americans are seeing broadcast and cable TV ads, putting large segments of the population out of traditional TV’s reach. But because of the way campaign ad-makers are paid, and because parties, candidates and campaign professionals are conditioned to prioritize television, the core expense of political campaigns is often still pricey TV ads.

At the heart of this problem is a warped set of economic incentives. TV consultants employed by campaigns typically earn 10 to 15 cents for every dollar of television ad time their candidate purchases. The more ads, the more income. Although this structure helps address the industry’s unusual risks — campaigns fold, paydays are sporadic — it creates a push to generate the most TV ads possible. Adam Sheingate, a political scientist at Johns Hopkins University, put it simply: “The economic incentives of the consulting industry are driving up the cost of campaigns.”

Meanwhile, recent research has sharpened our understanding of what political TV ads do — and, more important, what they do not do. One study from 2013 found that the “persuasive impact” of campaign communication “may end almost as soon as communication ends.” A 2013 Washington Post analysis found that even a significant increase in strategically chosen television spending would still have left Mitt Romney far short of winning the White House. In 2017, a pair of U.C. Berkeley and Stanford researchers concluded that traditional campaign persuasion tactics — including television advertising — have virtually no impact on voter choices in general elections. And a 2016 study determined that political attack ads are “never efficacious” at improving a candidate’s vote share.

Candidates, meanwhile, are running more attack ads than ever before and have increased spending on ads running long before Election Day.

This means campaigns are judged, partly, by how much money they have for this sort of campaign. In an August report, Michael Beckel, an advocate of campaign-finance reform, told the New York Times that money is one measure “party officials and activists use to assess candidates’ viability.” Also, “If you’ve got more money in your war chest, you’re in a position to make fewer tough decisions,” he said. “Do you go up on TV with one more ad? Do you go on the radio? Do you hire 50 canvassers? You’re in a position to decide how to get your message out.” He acknowledged that, “at the end of the day, any candidate is worried about being outgunned by an outside group that’s going to come in with an 11th-hour negative attack,” including on TV.

But reality is more marbled than conventional wisdom suggests, and fundraising advantages do not always translate into victory. This year, underdog Andrew Gillum, the mayor of Tallahassee, won Florida’s Democratic gubernatorial primary despite being outspent on TV by several rivals. In Massachusetts’s 3rd Congressional District, Lori Trahan was outspent by two competitors before narrowly winning the Democratic primary. In another Massachusetts Democratic congressional primary, Boston City Councilor Ayanna Pressley upended Rep. Michael E. Capuano with about half of his spending and, as Politico reported, “Capuano ran two TV ads, both about his efforts to oppose Trump,” while Pressley had just “one TV ad that ran only on Spanish-language television,” for which she paid a relatively modest $17,000. In 2016, Jeb Bush and Hillary Clinton each took a turn in candidate Donald Trump’s maw, despite spending more on TV, in part because of Trump’s dominant share of news coverage online and off.

Meanwhile, years of campaign cost inflation, in large measure a result of spending on television ads, has distorted our political system. First, high campaign costs create a barrier to entry for potential candidates who have neither personal wealth nor wealthy networks. Otherwise capable candidates — nonprofit organization leaders, public interest lawyers, policy experts — are often eliminated from the competition before it begins. Second, the cost of campaigns forces candidates to focus on raising money from special interests, and as Harvard scholar Lawrence Lessig has documented, the result is donor priorities set the congressional agenda at the expense of democratic responsiveness.

Despite these problems with campaign cost inflation, there are several reasons breaking our TV habit is easier said than done. First, high TV spending and its attendant fundraising are mechanisms of control for the political establishment, a way for entrenched interests to try to control up-and-coming politicians and discourage upstarts. I have advised campaigns where party-affiliated consultants discouraged and turned away from otherwise strong candidates because they concluded that the candidate was unlikely to hit an arbitrary fundraising target.

Fretful candidates then remain indefinitely crouched in TV-based warfare, closely watching every ad their opponent runs and responding in kind. This tit-for-tat fulfills a psychological need to simply do something, even if the benefits are unclear. TV spending is also logistically and intellectually easier than making meaningful, in-person voter contacts or building organizations to mobilize communities over the long term, which takes time, requires lots of people working together and bears uncertain prospects of success. TV, though, feels safe, especially given the establishment imprimatur. Mayor Gillum, by contrast, had success focusing on digital outreach, particularly text messaging.

In addition, media consultants dominate campaign budget decision-making — and stress the irreplaceability of TV. For lack of expertise, younger operatives accept media consultants’ conservative advice. Longtime pollsters and other consultants, with experience rooted in the TV era, often share and reinforce media consultants’ view of how to win. The cycle is self-fulfilling.

Even assuming all these people labor for their causes in good faith, the perceived need for millions of dollars of TV advertising yields undesirable outcomes, both from a strategic perspective and for democracy. Fundraising demands favor candidates who can self-fund or raise large sums — potentially, for House candidates, on the order of $1,800 every day for two years — and those candidates are often mismatches for the constituencies they are trying to represent. Some have more money than political savvy.

In 2016, for instance, Democrats recruited Randy Perkins, a wealthy disaster-recovery company owner and political neophyte, to run in Florida’s 18th Congressional District because he could fund his own TV ads. The results were disastrous: Republicans were able to charge that Perkins’s company overbilled a Florida school district for hurricane cleanup. Perkins was attacked for connections to the locally hated sugar industry; for being accused a decade earlier of overbilling the federal government for Hurricane Katrina cleanup work; and for not registering to vote as a Democrat before the run-up to his campaign.

Other Democrats with less baggage lost the primary — one context in which television ads may still have considerable power — in part because they lacked the fundraising capacity to attract party support. Perkins, fatally flawed, was soundly beaten in the general election.

As TV audiences continue to shrink, we cannot assume institutionalized television spending will naturally phase itself out. Right now, media consultants make more money when candidates air more ads, regardless of whether the candidate wins. If they were instead compensated for the amount of work they do, and also, perhaps, awarded bonuses for winning, they could become more receptive to different campaign strategies. Campaigns could then wean themselves from overreliance on TV. A change as simple as a flat-fee structure — increasingly popular at nonpolitical marketing agencies for its tendency to encourage media-neutral spending recommendations — would help.

It is true that pollsters in a particular campaign may have race-specific data that suggest maximizing TV spending remains essential. (This issue generates a lot of heat when I raise it with consultants.) But the publicly available evidence is sufficiently equivocal that the burden of proof lies with those who believe a payment model that pushes up the cost of campaigns is worth the accompanying distortions of money in our politics.