Chapter 5
What’s Going On? Polls, Debates, Money, and Ballot Measures
When it comes to the election process in the United States, one thing everyone can agree about is that the whole experience goes on a bit.
For one thing, it takes much more time to run for president here than it does for similar jobs in most of the world’s democracies. The 2015 Canadian election campaign took seventy-eight days, longer than that country’s two previous election cycles combined. In the United Kingdom, Prime Minister Theresa May called an election on April 19, 2017, and voters went to the polls on June 8. On February 1 the same year, New Zealand’s prime minister announced an election with a September 23 date.
By American standards, that barely qualifies as a warm-up.
In the United States, the process has extended to the point that the first debate of the 2016 election was held on August 6, 2015—fifteen months before the November election (and a full six months before the Iowa caucuses kicked off the primary schedule). The first debate of the 2020 cycle was held even earlier, on June 26, 2019. Several candidates saw their presidential runs end months before any voters had a chance to decide their fate.
We’ve already talked a lot about the election process itself, but there are many aspects of elections that we’ve only touched on briefly that can play significant roles in how things play out. Over a long election cycle, topics such as public opinion and fund-raising are ongoing processes, and they can change who winds up winning once the cycle finally wraps up. In this chapter, we’ll talk about those and other parts of the election process in a bit more depth.
Section I
Polling
Polls have come up a few times in this book already. Basically, they’re surveys that measure public opinion. That said, there are a variety of polls, with different methods and purposes.
Polls aren’t specific to politics and government. Businesses use polls for market research, and polling firms also interview people about things like pop culture or purchasing habits. To keep things simple, however, this book is just going to discuss polls in terms of how they relate to politics. You’ll hear a lot about them during any election, especially at the federal and state levels, so it’s a good idea to understand how they work and what they measure.
Polls might look at a particular political race, asking potential voters which candidate they plan to support, how they rank the different candidates on specific issues, or who they think has the best chance of winning. That might be done specifically to predict a winner or to gauge voters’ feelings along the way. In other cases, they might look at the performance of politicians currently in office, such as whether people approve of the job a president or governor is doing (that’s where their approval rating comes from), how they view his or her performance on particular issues, and how they feel about ideas that person has proposed. Polls might also measure what Americans think about a range of political issues, such as gun laws or healthcare or the environment, without even mentioning any candidate or official. Or they might simply survey Americans’ knowledge about the country, which is how we got the statistics in the introduction to this book.
When it comes to government and elections, the most useful polls are the ones that use a scientific method to try to obtain unbiased data about what the American people believe. That’s the goal of the more well-known polling firms, and some of them have been in the game awhile. The Gallup Organization is considered the first modern polling firm and has conducted scientific public-opinion polls since the 1930s—it correctly predicted Franklin Roosevelt’s landslide reelection victory in 1936 (admittedly, not the hardest prediction). Zogby International started surveying voters in the 1980s, and the Pew Research Center began in 1990. Those are just a few of the professional polling organizations trying to get unbiased data, and news organizations often conduct their own similar polls (sometimes partnering with other news organizations or with polling firms).
Interviewing every American about an issue is obviously an impossible logistical nightmare, so pollsters instead try to interview a representative sample of the overall population they’re trying to measure. Depending on the poll, that might mean trying to get a sense of all Americans or of a particular subset. When it comes to political issues, polls might measure everyone, or they might focus only on registered voters or likely voters (voters with a history of turning out on Election Day). Some might specifically interview Democrats or Republicans or independents. If the poll is about a race for governor or senator, it’s only going to look at the state where that race is happening.
How each of the dozens of active polling firms creates its sample, including how it finds the people to contact, is generally a trade secret. After all, how close their predictions come to the actual outcome of an election is a big part of how they stand out from their competitors. But in all cases, they’re trying to use probability to create the most representative sample possible. The industry standard for national polls is that about fifteen hundred respondents or more should be interviewed (though that number’s not set in stone, and samples are usually much larger than that).
Polls are usually conducted by phone. A number of interviewers working for a pollster will call numbers from the list put together for the sample and ask the same series of questions to anyone who answers the call and agrees to take the survey. Once the poll questions are complete, the interviewer asks the respondents for demographic information, such as their age, race, gender, or party affiliation.
The data from all interviewees are then collected and analyzed, and the polling firm will sometimes use demographic data to weigh responses and make the sample as representative of the population as possible. For example, if the people who agree to take a certain poll are mostly male, but the overall population being surveyed is mostly female, some pollsters might count the responses from women more heavily in the final results. During any poll, a certain percentage of calls will go unanswered or will be answered by people who don’t want to take the survey, and that affects how representative the sample can be.
In the process, pollsters calculate a margin of error, which uses probability to gauge the accuracy of the numbers found through the poll. The more representative a sample the pollster feels his or her survey actually reached, the smaller the margin of error will be, and vice versa. For example, if a poll shows a candidate has support from 65 percent of voters, and the margin of error is listed as 7 percent, that means the pollster conducting the survey believes the actual number is somewhere between 58 and 72 percent. If a poll shows a split of 51 to 49 percent, with a 2 percent margin of error, that means it actually shows that the race is a toss-up.
It became a popular (but misleading) talking point after the 2016 election that all polls showed Hillary Clinton was going to win. Even leaving her popular-vote victory aside, what the polls actually showed was that her odds of winning were fairly high, which is similar, but the difference is important. Most polls gave her a one-in-five chance of losing. That’s a greater likelihood than rolling a one when you roll a six-sided die. It’s more likely you’ll get a different number, but that doesn’t mean you won’t roll a one.
Telephone polling is the most common technique, but pollsters also conduct in-person, face-to-face interviews with samples of voters. Sometimes, those involve one-on-one meetings for longer, in-depth interviews. In other cases, pollsters will put together focus groups, which bring several voters together in one room, and ask them questions.
All of that describes pollsters who are trying to put together unbiased data. In addition to those, there are professional polling firms that are allied (either intellectually or officially) with one political party or another. National parties conduct their own polls, as do presidential campaigns, the White House, and campaigns for any number of other offices. Those kinds of polls usually have an agenda beyond just straight opinion polling, whether it’s testing support for a particular policy or introducing voters to an idea before a candidate starts talking about it.
Why Polls Matter
Public-opinion polling informs government policy in numerous ways. Like any tool, it can have good or bad impacts, depending on how the polling data are used.
On the good side of the ledger, polling is the best way to understand how the American people in general feel about what’s going on in the country (or at least the best way anybody’s developed yet). It lets officials and candidates know whether their ideas are popular or unpopular, as well as the reasons why. It drives a lot of news coverage, because in a country where the will of the people is supposed to influence how government does its job, it’s a way to explain what that will supports at that moment.
If the public seems skeptical about an idea, that might inspire a politician to do a better job explaining the details. If the public overwhelmingly supports a bill that’s facing opposition in Congress, it might make some members of the House or Senate more willing to back its passage. Especially at the local level, where it’s easier to get a representative sample of voters and a smaller number of voters changing sides can swing an election, polling can really influence what elected officials try to get done.
Generally, public opinion can help officials prioritize among the many issues on their plates. During an election, polls focused on who voters support can also be a huge help to campaigns, because they let candidates know where they stand and can help them plan accordingly. Politicians often claim that they don’t read polls. Maybe they don’t do it personally, but somebody on their staff definitely does, and it would be foolish if they didn’t. Nobody needs to take polls as gospel or change anything based on them, but they’re still useful data to consider. Plus, it must be said that, when it comes to predicting a winner, polls close to an election usually do a fairly good job.
On the other hand, polls can add to the horse-race nature of some political analysis, where who’s winning or losing in the polls at a particular moment can become a bigger news story than what those candidates want to do. Also, while politicians responding to the public is sometimes a positive thing, it should also be important for leaders to do what’s right instead of simply what’s popular and to think about the long term. The public doesn’t necessarily know or understand all of the options. As Henry Ford put it when describing the first car he built, if he’d asked people what they wanted, they would have asked for faster horses.
Polls can, unfortunately, influence voting behavior. In some cases, people don’t show up to vote for a candidate they think is going to lose anyway, and in the process they help ensure that they’re right about that. The reverse boomerang effect can also be true, meaning people are so sure their candidate is going to win that they don’t feel a need to vote . . . and enough people feeling that way can flip the outcome.
Another unfortunate tendency is that when polls ask voters their opinions on things that are matters of fact, the results are sometimes treated (by politicians and even some media outlets) as if people being wrong makes the facts less true. The previous chapter includes an example of a poll that found a significant number of Americans who believed provably wrong information about climate change or the September 11 attacks. That poll demonstrated ignorance on the part of those respondents, but it also gave people a safety-in-numbers feeling for believing things that were complete BS, and some politicians have benefited by reinforcing wrong information as a way to win votes.
When evaluating poll data, it’s also important to look at how the question was worded. For example, if a question just asks whether voters think a certain policy is working, saying no doesn’t mean they oppose that policy in principle. They might support the idea but not the execution. Or they might think the policy goes too far, or not far enough, but both of those sides could be treated the same way in the data. To take a different example, a poll might ask voters to select what they consider a first choice among many options. They might like three other choices almost as much, or they might like just their first option. People who only read headlines about poll results can miss nuances like that, or the full data might not be widely reported.
Of course, who answers the polls also makes a huge difference in their outcome. Pollsters try to find as representative a sample as they can, but they can only do so much.
One problem that pollsters have really struggled with is the switch from landlines to mobile phones. As recently as 2003, more than 40 percent of American homes only used a landline for their phone service; by 2013, that number had fallen to below 10 percent. That’s tough on pollsters for a few reasons. For one, federal law means automated dialing machines can’t be used to call mobile phones, so anyone without a landline only counts in polls that involve interviewers manually dialing the numbers. For another thing, landlines always use a local area code, while people can keep the same mobile number no matter how often or where they move. A 310 area code on a landline means that number is in the Los Angeles area, but a 310 mobile phone could be located anywhere in the country, which can be a problem if someone’s specifically trying to poll California voters.
Many pollsters believe that online surveys will eventually need to replace phone surveys as the main source of public-opinion information. Polling firms are working to develop ways of conducting online surveys that will be more representative, but as of this writing, that’s still a work in progress.
The difficulty of reaching mobile phones has also made it harder to reach younger voters, because it’s become normal for people to grow up without ever having a landline in their home. Because of that, polls often overweigh the responses of the young people who do respond, and the lower numbers are by definition less representative than the numbers for other age groups. Instead, some pollsters simply include fewer young voters in the data.
So if polls include more voices from older voters (and that is often what happens), the easiest solution is for more young people to answer their phones when a pollster calls and spend the few minutes it takes to answer the survey. Love them or hate them, as long as polls are the most popular tool for gauging citizens’ sentiments, those who take part will always have a better chance of having their views count as part of public opinion.
Types of Polls
The standard, one-time phone poll we just discussed is still the most common form, but not the only type of poll used during an election campaign.
Tracking polls look at the same issues over a specific period of time, usually a few consecutive days, for the purpose of tracking how day-to-day events change public perception. They’re particularly useful around something like a major policy speech or a debate, because they can show how the public reacted to the same question before, during, and after the event. Tracking polls might survey the same people over time or involve different people in similar samples.
Rapid-response surveys are conducted during or immediately after an event to measure the first impressions of its impact. For example, during televised presidential debates, networks often host focus groups that include undecided voters, supporters of each of the candidates, or both, and ask them all to turn dials to indicate when speakers say something the participants like or don’t like. Another example is flash polls, which are phone polls conducted within the first few hours after an event. These kinds of polls are good at getting people’s knee-jerk reactions, though it’s important to remember—as you’ve probably learned in your own life—that first instincts often change once people have time to think about things a little more.
Exit polls are a little different, and they can be extremely important. Their specific purpose is to predict who will win an election, and they are conducted among voters leaving polling places, usually by news organizations. (CBS conducted the first poll of this type, during Kentucky’s 1967 race for governor.) Pollsters collect data by asking a sampling of people at polling places which candidates they chose, and then they calculate early results based on those numbers. They also measure things like what voters considered the most important issues, and they include demographic information about voters.
If you’ve watched TV or followed online news on election night and seen networks or sites “call” races as soon as polls close or really soon afterward, that’s usually because exit poll data in those races predicted that one candidate won easily. Because they go on throughout the day, exit polls collect data more quickly than actual vote counting can be done, so news organizations tend to use the polls when reporting early returns. They’re not always perfectly accurate—voters can lie, or numbers can be skewed if an uneven number of one side’s voters is ignored or refuses to answer—but they’re generally pretty close, and responsible news organizations hold off on calling results until the exit poll data are clear.
The polls this section has covered so far are all designed to get a realistic picture of the group being sampled. Not every survey works that way. Dozens of organizations also conduct opt-in polls. Your local TV news station might ask viewers to call in and vote yes or no on a particular question. Websites have online polls that any visitor can fill in and see their results. During presidential elections, some fast-food chains even ask customers to “vote” by selecting one of two products (for example, choosing a red or blue coffee cup), and they keep track of the total as a fun promotion.
These kinds of polls are fine, and the results might be interesting, but they’re not at all scientific. The people being polled are seeking to participate, whether they’re choosing to click a link or to shop at a certain store. That automatically skews the makeup of the sample. Plus, people making a special effort to call into a news show and vote probably hold stronger opinions, pro or con, than the general public might.
No matter which approach they use, legitimate pollsters will identify their organization upfront when talking with voters and will never ask them for money. One of the struggles pollsters sometimes face is that voters have had bad experiences with telemarketers or other unexpected phone calls, which makes them suspicious or unwilling to participate. Another problem is that campaigns or supporters sometimes disguise themselves as pollsters to spread negative or flat-out false information about another candidate (see sidebar on page 182), which can unfairly reflect badly on real polls.
Section II
Debates
Debates are among the highest-profile parts of any major campaign. At the presidential level, they are watched by millions of people. The first Democratic debate of the 2016 cycle drew more than 15 million live viewers. The first Republican debate of the cycle was seen by 24 million. If you don’t count sporting events, that debate would have broken a record at the time for the most-watched program on American cable TV. During the general election, the three debates between Hillary Clinton and Donald Trump drew a total of 259 million viewers, breaking 1992’s record of 250 million. For voters who don’t follow politics closely, debates (along with convention speeches) are the kinds of special events that make them tune in and help them make their choice.
Presidential debates understandably get the most attention, both in the primaries and in the general election, but debates are also common in races for the Senate and House, state offices, and local elected offices. In any election, debates are a rare opportunity to see the candidates in the same place at the same time and for them to answer questions they don’t necessarily expect. The candidates aren’t going to agree on “right” answers, but undecided voters can use the candidates’ disagreements and how they present their ideas as the basis for choosing which of them to support.
They’re such a big part of elections now that it’s worth mentioning that debates are not mandatory and never have been. They are just an entrenched tradition, and not even a particularly old one.
The first fifteen presidents never had face-to-face presidential debates as candidates do now. Abraham Lincoln did have a series of famous debates with Stephen Douglas, but those took place when they faced each other in an 1858 race to represent Illinois in the Senate. (See sidebar on page 188.) The two of them ran against each other again in the presidential election two years later, but they never debated in that race.
Debates made a small comeback in the middle of the twentieth century, but they still didn’t quite catch on. The 1948, 1952, and 1956 elections each included at least one debate during the primaries with candidates or their representatives, but they weren’t major events like the debates we see today.
It wasn’t until the 1960 race, featuring John F. Kennedy against Richard Nixon, that presidential candidates debated during the general election. They faced off four times between late September and late October, with debates held in four cities—Chicago, Washington, Los Angeles, and New York—and broadcast nationwide on all three television networks (ABC, NBC, and CBS) and on the radio. (One of the debates actually featured the candidates being filmed in different places but both responding live to the moderator’s questions.)
The 1960 debates were a huge hit with Americans. The first one was watched by more than sixty-six million people. It’s now historical lore that people listening on the radio thought Nixon won, but those watching on TV thought Kennedy did (partly because Nixon looked unhealthy after dealing with an illness, and he struggled with talking to the camera instead of directly to the moderators). In reality, the radio audience was a small portion; nearly 90 percent of American homes had televisions by that point, and the candidates’ four debates were each watched by between 57.8 and 61 percent of households with TV. In the past, most voters never actually saw a presidential candidate campaign live unless they attended a speech; they read about them, heard them on the radio, or watched prerecorded news footage. These debates were a pretty big deal.
You might think that the success of those debates would have made future debates a no-brainer, but they didn’t happen again until 1976. There were a few reasons for that, but mostly it was because the 1960 election was a special case. That race was always going to be close—only 0.2 percent of the popular vote and eighty-four electoral votes ultimately divided the two candidates—so both Nixon and Kennedy saw value in speaking before a huge national audience, and the public was extremely interested in the election. The networks also wanted to show they were civic-minded by broadcasting the debates, though the huge viewing numbers certainly didn’t hurt.
Also, Congress actually suspended a law to allow the debate. The Communications Act of 1934 required that broadcasters give equal time to all candidates, not just those from the Republican and Democratic Parties. To make a debate between just Kennedy and Nixon legal, Congress had to temporarily suspend that law, and in 1975, the Federal Communications Commission (FCC) formally changed the rules so that debates would never need to include every party.
Presidential debates didn’t happen in the next three elections, simply because one candidate refused to take part. Lyndon Johnson in 1964 was so consistently ahead in the polls that his campaign saw no reason to risk a debate, and Richard Nixon chose to avoid a repeat of his poor 1960 performance by avoiding debates in his successful 1968 and 1972 runs. In 1976 Gerald Ford and Jimmy Carter finally brought back general election debates, and they’re probably not going away anytime soon.
From Then to Now
Voters in the United States have had a chance to watch major candidates debate in every presidential race since the Ford-Carter showdown. That’s just the general election. Republicans held 85 primary debates between the 1948 and 2016 cycles, and the Democrats held 104. (See sidebar on page 193.) Most of those happened from 1980 onward. As noted above, the earliest modern debates were actually between members of the same party, including a famously dull 1956 primary match (the first-ever televised debate) in which Democrats Adlai Stevenson and Estes Kefauver agreed on almost everything. Since 1976, the Democratic and Republican candidates for vice president have also held one debate in all but one election year.
While debates are now a tradition, that doesn’t mean anyone has to participate. For one thing, incumbent presidents have always refused to take part in primary debates rather than make challengers seem more legitimate. In 1980 the Democratic and Republican tickets couldn’t agree to terms for the vice-presidential debate, so there wasn’t one. Also in 1980, Republican representative John Anderson of Illinois was running as a third-party candidate, and he was polling well enough that the League of Women Voters (LWV) invited him to the first general election debate. President Jimmy Carter refused to join, so Anderson and Republican nominee Ronald Reagan debated each other. The second debate was canceled when Carter wouldn’t join, but the third, between only Carter and Reagan, set a viewership record—80.6 million Americans tuned in, seeing as it was the only opportunity to watch the two of them debate.
Over the years, the ways in which presidential debates work has changed, and that’s given the parties a lot more control over things.
The League of Women Voters, a genuinely nonpartisan organization, used to moderate all of the presidential debates. It hosted the early primary versions in the 1950s and ran the general election debates in 1976, 1980, and 1984. There was always some back-and-forth involved in getting candidates to agree to the rules, but the LWV decided which of their demands were reasonable.
However, in 1988 representatives from the two major parties secretly agreed on a range of demands—about everything from who could serve as moderator for the debate to how tall the candidates’ podiums could be. Once the LWV found out, it withdrew as a sponsor, saying the demands “would perpetrate a fraud on the American voter.”
That gave the Democratic Party and Republican Party an excuse to take over the process, forming a bipartisan (not nonpartisan) nonprofit organization called the Commission on Presidential Debates (CPD). The CPD formally selects the venues, the moderators, and the format for presidential debates, but all those decisions really come from agreements between the parties and candidates. On one hand, the new system has meant more primary debates, so more chances for voters to see the candidates face questions. On the other, the fact that the parties run the show means they’re able to avoid any surprises and the debates are quite a bit more predictable. The debates might change format slightly—a standard moderator format with the candidates at podiums, a “town hall” where they take questions (usually prescreened by the moderator) from the audience, a panel where multiple moderators ask questions, or a forum where candidates are seated at a table—but those differences are still approved by both major parties.
Also, until fairly recently, third-party candidates had a realistic chance of sharing the debate stage if they received enough support. In 1992 businessman Ross Perot was performing so well in early polls that the CPD included the third-party candidate in the three presidential debates (and his running mate, James Stockdale, in the vice-presidential one). Perot didn’t win any states in November, but he won almost 19 percent of the popular vote. Because his ideas appealed to factions of both major parties, each had a plausible argument that Perot’s debate performance hurt its candidate.
After that Perot experience, the two major parties decided to make it harder for anyone else to get on the stage. Rather than ban third parties, which would rightly be seen as unfair, they set the rules so that any candidate needed to average at least 15 percent in a series of polls before getting an invite. (Remember, the FCC already got rid of the rule requiring all candidates to receive equal time.) In 1996 even Perot wasn’t able to participate in any debates under the new rules. A few third-party candidates have sued the CPD over its control of the debates since then, but so far nothing has changed.
Third parties can host their own debates, too, of course. It’s just hard for those debates to get on TV for a wide audience to check out, although a channel like C-SPAN might choose to air them. In 2016 Libertarian Gary Johnson and Green Party candidate Jill Stein had two televised debates, one on PBS and one on Russian news channel RT.
The two main parties also set the rules for which candidates appear on the stage during their primary debates and determine the debate schedule.
Early in the 2016 election cycle, the Republican Party had enough contenders for the nomination that it took the unusual step of hosting two back-to-back debates. The top contenders based on poll data debated during prime time, while those who missed the cut faced one another in a smaller, earlier, and less-watched debate (which earned nicknames like the kids’ table or junior varsity), until enough candidates dropped out for everyone to fit on one stage. The number of primary debates has gone up since 2000, and the debates have started earlier in the election cycle. It’s become normal for candidates who don’t gain much support after the debates to leave the race months before anybody votes in an actual primary or caucus.
The debate formats used at the presidential level have become so popular that state and local races usually use similar ones. In races where one candidate has a big lead, it can sometimes be hard for challengers to get the front-runner to agree to a debate—someone sure they’re going to win might think an unscripted debate can only hurt them. Still, most governor, Senate, and House races will have at least one debate, usually broadcast by a local news station.
Smaller races, such as for a seat in the state legislature or a mayoral race in a small town, might or might not have debates. It all comes down to whether the candidates or their parties can come to an agreement on the rules. In a city such as New York or Chicago, local news usually shows the debates for mayor. Public-access television and radio should broadcast local debates, and spectators can always watch them in person, depending on the event’s ticket policies.
Section III
Fund-Raising
Say what you will about the problems with how debates work, but at least they give the candidates who qualify a more or less equal chance to get their message to voters. Campaign fund-raising is a very different case, as supporters of different candidates can raise wildly different amounts of money, which can mean more ads, more staff, and generally more chances to convince voters that they’re the right choice.
When people complain about democracy being broken or special interests having too much power, this is one area they usually mention, and it’s hard to argue they don’t have a point. In this section, we’ll focus on how fund-raising works in federal elections.
One thing members of Congress (of both parties) complain about is the amount of time they need to spend calling donors to ask for money and attending fund-raisers. How much time that takes depends on the person. A senator or representative in a close race in a big state probably needs to raise more than someone expected to win easily in a smaller territory without a pricey TV market.
Candidates for federal offices raise money from a combination of sources—individual donors, their party’s national and state (and, sometimes, district and local) committees, political action committees (better known as PACs), and their own personal money.
An individual donor is any person who donates to a candidate on his or her own. As of 2020, the individual donation limit is $2,800. That number is adjusted for inflation during every election cycle. So as an individual contributor, you can only give $2,800 to any one candidate, but you could give that same amount to two or three different candidates (or more, if you’re rolling in money). Also, that’s the limit for the entire cycle, but it can be broken up into a few pieces. If you donated $1,800 to a candidate in the Democratic or Republican primary, you’d be able to donate another $1,000 to them during the general election.
When campaigns call to raise money, they’ll say that even small donations help, and they’re right, because of volume. Twenty-eight individuals donating $100 each are as valuable as one person donating the limit, and potentially more valuable, because the campaign can ask them for more money later in the cycle, while the donor who gave $2,800 is already done. Presidential candidates routinely raise millions of dollars from individual donors, so combined they’re a major source of campaign cash.
Individuals can also donate: up to $5,000 per year to a PAC; $10,000 per year to party committees at the state, district, or local level ($10,000 total, not to each); and up to $35,000 per year to the national party committee. Obviously, that can add up to a level of money very few Americans can afford to give.
If you donate to a PAC or party committee, you don’t get to decide which candidates get the money. For example, both the Democrats and Republicans have campaign committees that focus on House races (the Democratic Congressional Campaign Committee and the National Republican Congressional Committee, respectively). They send money to the House races where the committee leaders think their candidates need the most help, or where they think extra money can make the biggest difference. There are also limits on how much any PAC or party committee can contribute to any one candidate (with $5,000 per election at the high end).
As you might guess, those numbers are still small compared to the kind of big money that’s sometimes spent on elections. That’s where outside funding groups come in.
Groups from the Outside
Other groups that spend significantly on elections include 527 groups, Super PACs, and 501(c)(4) organizations. Both numbered terms come from the section of the federal tax code that applies to them. The money these groups raise and spend can play a big role, so it’s worth understanding how they work.
A 527 group is defined as a tax-exempt political organization. That includes party groups such as the Republican National Committee and Democratic National Committee. It also includes committees that handle election activity for well-known advocacy organizations such as the Sierra Club, EMILY’S List, and the Club for Growth. Those kinds of advocacy organizations run campaign ads about particular issues relevant to what they do, but they don’t specifically tell voters which candidate the organization wants them to choose (“Vote for lower taxes,” not “Vote for this guy”). Those kinds of 527 groups have been around for ages. As long as they play by the rules, and their ads never expressly say how to vote, they can raise unlimited money and can take contributions from anyone.
In 2002 Congress passed the Bipartisan Campaign Reform Act (also called McCain-Feingold, after its cocreators). It introduced many reforms, most notably banning corporations or unions from broadcasting campaign ads in the last sixty days before an election (more on that shortly). But it didn’t change how 527s work, so a new kind of 527 organization started popping up during the 2004 election cycle, formed specifically to influence that election. One famous example was the group called Swift Boat Veterans and POWs for Truth, which ran misleading attack ads about Democratic nominee John Kerry and the medals he won in Vietnam, but never specifically told voters not to vote for him or to choose his opponent. A minor technicality, but one that made the ads legal. Other new 527 groups like America Coming Together and the Media Fund didn’t get as much attention, but similarly spent millions on that election.
Any 527s still need to disclose who donated to them, register with the Internal Revenue Service (IRS), and follow rules about their advocacy. Many reformers called for changes to the 527 rules after the 2004 election, but those organizations were soon replaced as campaign finance reformers’ biggest challenge.
One of the most important recent Supreme Court cases, the 2010 decision in Citizens United vs. Federal Election Commission (often referred to as simply Citizens United) fundamentally altered how campaign finance in the United States works. The ruling eliminated the ban on corporations and unions paying for electioneering communications and independent expenditures. Those organizations still can’t give money directly to a candidate for federal office, but they can now spend unlimited money pushing for any candidate’s victory or defeat. Another 2010 case that used Citizens United as a legal precedent, SpeechNow.org v. FEC, ruled that there’s no limit on how much anyone can donate to independent-expenditure groups. Combined, those two cases meant any entity—a wealthy person, a corporation, you name it—can donate as much money as they want to groups working outside the campaigns to influence the outcome.
Those two rulings led to the 2012 rise of expenditure-only committees, better known as Super PACs. Unlike regular PACs, Super PACs can’t donate directly to candidates, since that would violate the ban on corporations and unions donating directly. Unlike 527s, however, they can support specific candidates. Many of the largest Super PACs exist for only one election cycle and only to support or oppose one candidate (there are a few exceptions, such as FreedomWorks for America and House Majority PAC). They’re not allowed to coordinate their efforts with campaigns or parties, but the rules around that come with a lot of loopholes. After all, it’s easy enough for a Super PAC to figure out what a candidate wants from public speeches and interviews with the candidate or his or her campaign staff. If someone lets it be known publicly that they liked or didn’t like an ad the Super PAC ran, it’s easy for the Super PAC to react to that, without the two sides ever officially interacting.
Even Super PACs need to disclose where they get their money, though, so people looking to hide their influence can instead donate to 501(c)(4) groups.
On the one hand, 501(c)(4) groups are defined as nonprofit social welfare organizations. They can get involved with politics, as long as that isn’t their primary purpose (for example, they can’t spend more than half of their money on political activities). Historically, that applied to groups like local civics leagues, as well as organizations such as the American Civil Liberties Union (ACLU) and the American Association of Retired Persons (AARP). But after the 2010 court rulings, the IRS was flooded with applications from new groups wanting to be defined as 501(c)(4)s. Many of these groups are essentially nonprofit PACs that operate independently of the parties and campaigns, but can raise and spend unlimited money without having to disclose their donors. According to the Center for Responsive Politics, these nonprofits spent even more than Super PACs in the 2010 election cycle, by about a three-to-two ratio.
The Impact
These recent changes to the campaign finance system have been controversial for several reasons. For one, they make the process less democratic, as one rich individual or corporation can donate unlimited money and can make candidates at least appear to answer to those individuals. They’ve made the cost of elections, which already grew every cycle, grow even faster. Advertising is one of any campaign’s biggest expenses, and outside groups picking up the tab for much of it (even unofficially) has an impact on how candidates and parties raise and spend money.
Ads from outside groups can also be particularly nasty, since the candidate they support doesn’t have to take responsibility for them. And when ads come from groups that don’t need to disclose their donors at all, voters never find out who’s behind the messages they’re seeing. If, for example, a company that pollutes near a town is running ads trying to take out the member of Congress attempting to make it pay for the cleanup, voters should at least have the ability to know where those ads are coming from and consider that when voting. These rulings mean they don’t always get to find out.
By the 2012 election, the impact of these changes was already pretty obvious in terms of the big money in the race. The total spending on the presidential race from outside groups more than doubled between 2008 and 2012, making it the first cycle measured in billions instead of millions, and that number increased in 2016 (see sidebar on page 203). Even that understates the situation, because the 2008 election had two primary processes with several viable candidates, and in 2012 only the Republican primaries were contested.
It’s not just the presidential race. In 2014 the most expensive Senate race was in North Carolina, where Republican Thom Tillis defeated Democratic incumbent Kay Hagan. More than $118 million was spent on that seat in the general election, but only about $35 million of that was spent by the two candidates’ campaigns, with the rest coming from outside groups. On the House side, 2014’s priciest race was in California’s seventh district, where Democratic incumbent Ami Bera spent about $4.4 million in victory, Republican challenger Doug Ose spent about $5.1 million in the loss, and outside groups spent about $13.7 million.
In total, the 2018 midterms cost about 35 percent more than the 2014 edition, and outside spending was up 61 percent. The 2018 race for Georgia’s sixth district between Karen Handel and Lucy McBath cost more than twice as much as that California race between Bera and Ose.
If you watch broadcast television in the last few weeks before a general election, you can’t help seeing dozens and dozens of political ads. Thanks to the Bipartisan Campaign Reform Act, since 2002, politicians appear in any ad that comes directly from the campaign, to give it an official endorsement (as ordered in part of the law known as the “stand by your ad” provision). If the ad features audio or video, the candidate will say something along the lines of, “I’m Joe Candidate, and I approve this message.” Ads from outside groups might or might not feel like they were approved by the candidate, but that distinction can be lost on voters who don’t know about or don’t fully understand how campaign-finance laws have changed in a short amount of time.
Section IV
Ballot Measures
Most laws are passed by Congress or by its state and local doppelgangers. However, some states allow a certain amount of direct democracy in the form of ballot measures, meaning voters on Election Day also need to choose whether the proposals on the ballot become law.
These measures only apply to state and local, not federal, law, and each state has its own rules for whether ballot measures are allowed and how they work. The idea of ballot measures in the United States goes back to just after the founding, when Georgia included the concept in its state constitution. However, the modern process really began during the Progressive Era of the late nineteenth century and early twentieth century as one of many reforms aimed at helping the people serve as a check on the government. Today, some states heavily rely on them, while others use ballot measures only in special circumstances.
There are three main types of ballot measures in the United States.
Legislative referendum, also called legislative referral, means the state legislature makes the decision to put a proposed law up to a public vote. Sometimes it’s mandatory, as when an amendment to the state constitution is being proposed—every state except Delaware requires that voters get to weigh in on any constitutional amendment, as a reasonable check on the state legislature’s power. It’s also common with laws that involve bond issues, because citizens will need to pay taxes to fund a project, and this way they have a say in whether they’ll spend that money.
A second type of ballot measure, popular referendum, means voters are trying to repeal a specific law passed by the state legislature. This is sometimes called a citizens’ veto, and that’s a pretty good description for it. More than twenty states allow this kind of measure, and petitioners need to get a minimum number of valid signatures before the referendum qualifies for the ballot. That number is different from state to state.
The third, and most popular, variety is the ballot initiative. In the states that allow initiatives—as of 2018, twenty-four plus the District of Columbia—any citizen or organization can write a measure and have it appear on the ballot. The initiative still needs to meet certain state-specific requirements, and it always needs a minimum number of signatures before appearing on the ballot—a way to demonstrate the measure has enough support to justify having voters weigh in on it.
What They’ve Done
Well more than half of the ballot initiatives in the United States have come from only six states—Arizona, California, Colorado, North Dakota, Oregon, and Washington—so they’re clearly more popular in the West.
California is particularly fond of ballot initiatives, so we’ll use it as an example. Since 2000, every even-year election has seen between six and eighteen statewide initiatives for voters to consider, plus any number of local measures. For many years, voters even received booklets not much smaller than an old-fashioned phone book that included the wording of every initiative and arguments for and against each. In 2011 California governor Jerry Brown signed a law passed by the state legislature that made an important change to the process. Rather than appear on a ballot in any election (including off-year primaries), initiatives would only appear during general elections in November, when turnout is higher and more voters would decide every ballot entry’s fate.
To continue with the California example, here’s just some of what voters there did via statewide ballot initiatives during the 2012 election alone. They agreed to increase sales and use taxes by one quarter of one cent for a four-year period and increase income taxes for seven years on anybody making more than a quarter of a million dollars per year, with the money going to education. They voted in favor of congressional redistricting being handled by a nonpartisan commission. They supported a change to term-limit laws for the state legislature, limiting legislators to twelve years combined in either body (instead of the old system, with separate limits for the state’s two houses). They voted to increase prison sentences for anyone convicted of human trafficking and to make convicted traffickers register as sex offenders.
They even voted for an initiative that changed an earlier voter-passed initiative. In 1994 California voters had passed a “three strikes” law, which won with 72 percent of the vote and meant that anyone convicted of three felonies was subject to life in prison. That law had its share of unintended consequences, as many nonviolent criminals were subject to the same life sentences as murderers and rapists. In 2012 voters passed a modification to the three strikes rule, so that only violent offenders would receive life sentences, and people already serving life for their personal drug use or other less-serious crimes were able to appeal.
Overall, that year the same voters passed six initiatives and rejected seven, so it’s not as if getting a measure on the ballot guarantees success. Among the things California voters rejected that year: a statewide ban on the death penalty; a ban on corporate and union contributions to state candidates; and the mandatory labeling of genetically modified food. In the latter case, agribusiness company Monsanto (for which genetically modified food is a big part of business) and other companies spent an estimated $46 million campaigning against the ballot initiative.
Even though initiatives only directly affect the state where they’re passed, they can impact other states too. If a measure passes and it winds up working well for one state, voters and legislators elsewhere are more likely to try the same idea in their state. The federal government has also been known to adopt ideas that have already been proven to work at the state level. As the Monsanto example shows, state initiatives can have huge political stakes, with one side trying to stop an idea from spreading, and the other trying to establish a model that others can copy. (For more examples, see sidebar on page 208.) Because of that, campaigns for and against some ballot measures can feel a bit like campaigns for and against candidates, with everything from negative advertising to huge get-out-the-vote efforts.
Ballot measures seem like a true example of democracy, and they can be. Or they can be ways for special interests to get around the legislative process. Once again, they’re a tool that can be used for good or bad, depending on your perspective.
Getting a measure on the ballot is also one of the easiest ways for young people to get more involved in the political process. In the next (and final) chapter, we’ll talk about some of the other ways, beyond simply voting, that you can play a part in politics, during an election or any other time.
