New US election polls are constantly appearing in the news world. How meaningful are they actually? A short guide through the demoscopy jungle.
Americans call the honeymoon the first big hello, the weeks after saying “I do”. The period when the candidates for the presidential election have been decided – and their hearts are first fluttering. Just like with Kamala Harris. But that’s over now: For the first time since Vice President Harris replaced US President Joe Biden as the US Democratic candidate, her poll numbers are falling. Just a little, but enough to make you sit up and take notice before the upcoming TV duel.
Left-liberal America’s new hope for its competitor Donald Trump has lost two and a half percentage points. On the other hand, only half a percentage point in the hotly contested states – i.e. where the elections are very likely to be decided. According to the New York Times newspaper, it is now 47 to 48 percent for Trump in the election duel. However, these numbers are only the result of a survey of dozens: On average, Harris is still slightly ahead of the Republican.
So what exactly do these numbers mean? Do they say anything about the November 5 election results? And what about the simultaneous elections for the Senate and the House of Representatives? An overview.
The catch with the election polls
Eight weeks before the vote, the outcome is still completely open. Although Kamala Harris has continued to gain in popularity since Biden’s withdrawal and has even overtaken the Republican Trump, the lead is too small to derive any kind of forecast. The survey results, which are generally good for Harris, have a few catches:
The margin of error: Surveys, like all measurements, contain a percentage of inaccuracy and/or sampling error. Also, people don’t always answer truthfully. This margin of error is approximately between three and five percent, depending on the number of respondents and the survey method.
This means that if a poll comes to the conclusion that Kamala Harris is at 47 percent, then you would actually have to say she is at around 46 to 48 percent. We don’t know for sure. And if Donald Trump gets 48 percent, then his value is roughly between around 47 and 49 percent. Only if the difference between Kamala Harris and Donald Trump were around three percentage points would that make a truly reliable statement. Currently, however, the two are closer together in the majority of surveys.
Surveys are not forecasts: There were two big election surprises in 2016: when the British voted for Brexit in the summer and the Americans elected Donald Trump as US President in the fall – the polls painted a different picture in both cases. The problem back then was not just the quality of the numbers, but even more so the interpretation by journalists and pollsters. In fact, both votes were so close and within the margin of error that any prediction based on polls was essentially impossible.
Another example: In the last US presidential election four years ago, Joe Biden was around eight percentage points ahead of Trump in the polls a few days before the vote. In the end he won by just five percentage points. The difference of around three percentage points corresponds to the most comfortable cushion that Harris currently has over Trump.
Not all surveys are the same: There are countless ways to gather people’s opinions: by telephone or in person, through online voting or through questionnaires. You can also ask non-voters or party members, women, men, old people, young people, only black academics or only supermarket employees of Asian origin. The results will always be different.
Sometimes it depends on the question: Is it open-ended like, “Who would you rather see in the White House?” Or closed like: “Would you vote for Robert F. Kennedy Jr. for president?”. Pollsters often also have political favorites. At the Rasmussen Reports Institute from New Jersey, for example, Donald Trump is always well above the survey average. Other pollsters, however, prefer the Democrats. The site 538.com by demoscopy star Nate Silver has therefore created a ranking of survey institutes according to forecast quality and transparency. And of course a lot can happen before the election.
US election is decided in individual states
A large part of the survey is a beauty contest: Who do Americans like best? This reflects the mood in the country, but not the way the country votes.
The US electoral system: What matters is how many electoral votes the candidates get in each individual state. Most have split into a two-party system: California, for example, has voted Democratic since 1988, while southern states like Alabama have always voted conservative. Their electoral votes have basically already been taken: If Harris in California and Trump in Alabama get more votes than their own candidates in previous years, that will have no influence on the outcome of the election. It’s different in the competitive swing states. They do not traditionally align themselves with one of the two parties, so their voting behavior can determine who wins the White House.
How many electoral votes the states have available depends on how many people live there. In almost all states, the electors are allocated according to the “winner takes it all” principle. This means that even if a candidate only receives 51 percent of the vote, all the voters are assigned to him and 49 percent of the vote is practically lost. That’s why you can get the most votes in the entire country – and still lose the election.
This year, Pennsylvania, Wisconsin, Michigan, Georgia, Nevada and Arizona are likely to be decisive. Whoever wants to win the “most important election in the history of the United States” (Donald Trump), the “most existential and important election of our lifetime” (Kamala Harris), has to get the most votes there. In two of these six states, Harris is ahead of Trump, in Arizona the Republican is leading and in Georgia, Nevada and Pennsylvania both are tied.
Two months before the vote, it is completely unclear who will win the race for the White House.
Will government be governed with or against the Congress?
The question of whether the election winner will then have to govern with or against the US Congress is just as open. Representatives for both chambers of the US political system are elected at the same time.
House of Representatives: This choice often seems unspectacular, but it has serious consequences. The Congress Chamber itself is comparable to the German Bundestag, although less powerful. Nevertheless, a US president can only govern to a limited extent without a majority. The representatives, who are elected every two years, can at least significantly delay government projects.
This is also happening at the moment, because the Republicans are the largest of the two factions with 220 of the 435 seats. As in Germany, parliamentarians are elected in constituencies. According to current projections, the conservatives have a good chance of defending their majority.
Senate: Nothing works in US politics without the second, upper chamber of Congress. Even more than the House of Representatives, the senators control the head of state in the Oval Office. All members of the government, the highest judges and international treaties require the approval of this chamber.
Around a third of the 100 seats are elected every two years. Each state always sends two senators. It is currently a tie between Democrats and Republicans. In the event of a deadlock in voting, the US Vice President has the deciding vote – currently Kamala Harris. 34 seats will be re-elected this year; according to the polls, it remains a stalemate, although with a slight tendency towards a Republican majority.
Sources:RealClearPolitics, 538.com, AP, “New York Times”, CNN, Rasmussen Reports, Axios, DPA, AFP, Reuters, 270towin
Not a survey but a questionnaire is the best election oracle
Swing states and electoral college – this is how the US election works
Will this withdrawal help Donald Trump win the election?
