Trump could win even if he stands in Fifth Avenue and shoots somebody
Biden needs voters to feel better economic vibes
Welcome to the 2nd post to my EconomyStupid.substack.com newsletter. In the table above, I have updated my EconomyStupid Model election projection with the University of Michigan’s final Index of Consumer Sentiment reading for April 2024 and RealClear Polling’s averages of credible surveys as of 2024-5-1 9:30 am EDT.
Today’s hot takes:
Donald Trump should be leading Joe Biden by 8 points (54%-46%), IF Trump were taking full advantage of voters’ negative economic vibes.
Trump may lose support as his trials proceed. But, unless consumer sentiment improves, Trump could win even if convicted of standing in the middle of Fifth Avenue and shooting somebody.
With 6 months to go, all is not lost for Joe Biden. Obama was in a similar hole in 2012. He won as the Index of Consumer Sentiment (ICS) rose before the election.
Biden’s best hope is a fall in annual inflation. Not much chance of help coming this month with CPI annual inflation forecast to stay at 3.5%.
The race is close only because Trump’s behavior costs him 2.6 points.
I wrote this post before the news broke about one of Trump’s potential Vice-Presidents, Governor Kristi Noem of South Dakota. Governor Noem writes in her autobiography that she shot a 14-month-old dog who was difficult to train.
Trump could NOT be elected with a dog-murderer as his running mate.
Reconciling Nate Silver and James Carville
In my first post, I challenged Nate Silver’s recent ribbing of James Carville: “It's not just the economy anymore, stupid.”
Silver sets fire to a straw man by debunking Carville’s “economy, stupid” mantra. Like Silver, Carville recognized the importance of other issues.
Even in 1992, Carville supplemented “economy, stupid” with two more points on his sign:
Change vs. more of the same
Don’t forget health care
The list of both non-economic and economic issues has changed since 1992. Health care and unemployment have slipped down the list of voter priorities. Silver is correct that the state of the economy does not dominate the 2024 presidential campaign the way it did in past elections such as 1992 and 2012. But, today’s top economic issue, inflation, still ranks ahead of immigration or any other public concern .
Are economic voting models bad science?
Beyond pointing out that non-economic issues also affect voters, Silver has a broader case against the “economy, stupid” approach. He also condemns economic voting models as “bad science [marred by] methods that tortured the data and overfit models on small sample sizes”.
All of Silver’s criticisms of past economic voting models apply to my EconomyStupid Model (ESM). I plead guilty on all counts. To take just one point, ESM cannot generate statistically significant results from 18 presidential elections from 1952 to 2020. Not to mention that the population of voters is constantly changing. Almost all 1952 voters were dead and gone by 2020.
How to use the EconomyStupid Model (ESM)
I admit that my ESM does not pass statistical tests. But, ESM can still provide a useful framework for understanding a presidential election. The trick is to know what ESM can and cannot do.
First, I have to acknowledge Silver’s point (and Carville’s) that the economy is usually the most important issue, but never the only issue. My ESM backcast predicted the popular vote winner in 14 elections back to 1952. But, ESM was wrong in 4 elections.
For example, with consumer sentiment relatively high, ESM projected a win for Hubert Humphrey in 1968. Instead, Richard Nixon was elected. The unpopularity of the Vietnam War outweighed good economic vibes.
However, I argue that even when ESM is wrong, we learn something. Comparing the ESM projection to the actual 1968 result, I estimate that the backlash against the war and social turmoil of the 1960s cost Humphrey almost 3 percentage points – enough to tip the election to Nixon. (More on the 1968 election in a post this summer.)
What about the Electoral College?
Another shortcoming: I use ESM to project the popular vote, not who will win the state-by-state Electoral College battle that decides the presidency. In 2 of the last 6 presidential elections, the winner prevailed in the Electoral College despite receiving fewer total votes from all Americans than the loser.
If you could wait until election day on November 5th to find out who won, you wouldn’t be reading this. ESM can’t help you predict the future. But, ESM can give us an idea of how the race might unfold between now and election day.
Here’s how I look at the election. I think of my ESM projection as Biden’s floor. With Americans so gloomy about the economy, Biden would still get about 46% of the 2-party vote if his opponent were a pre-MAGA Republican focusing on the economy.
If Trump were as smart as he says, he would stick to Ronald Reagan’s ballot question: “Are you better off today than you were 4 years ago?” Trump would be leading Biden by 8 points in the national polls, enough for an Electoral College landslide.
Biden faces a small disadvantage with RFK Jr. and other minor-party candidates attracting more Biden-leaning voters than Trump-leaners. Nevertheless, Biden is polling well above his floor level of 46% – voters loyal to the Democratic Party even during this spring of economic discontent.
Biden is polling about 49% (as a share of the 2-party vote) because some voters ask another ballot question: Is Trump fit to be president? 2.6% are answering NO despite concerns about the economy and other issues such as Biden’s age.
Biden can’t win the Electoral College with 49% of the 2-party national vote. Hillary Clinton won the popular vote by 2 points in 2016, but Trump still won the Electoral College 306-232 thanks to the small-state tilt favoring any Republican candidate.
New York Times political analyst Nate Cohn has estimated that Trump’s Electoral College advantage may be narrower in 2024. According to Cohn, if Trump loses the national popular vote by 1 point this year, he could also lose the Electoral College.
Biden needs to get above 50% of the 2-party vote to have any chance of winning the Electoral College.
What could shift the race?
One possibility: Trump’s behavior gets worse as he shuttles between court trials and campaigning. More swing voters switch to Biden. Trump’s behavior cuts his support even further below where he should be, if the election were decided solely by voters’ economic sentiment.
But, part of Trump’s political magic has been his ability to convince so many Americans to ignore (or, in many cases, revel in) his boorishness. Trump’s antics were not enough to elect Hillary Clinton in 2016, when fellow Democrat Obama’s administration left Americans feeling better about the economy than they do today.
Trump may reach a limit of how many voters he can drive away with bad behavior alone. To borrow one of Donald J. Trump’s infamous boasts, it may well be that he would regain the presidency even if he stands in the middle of Fifth Avenue and shoots somebody. The Index of Consumer Sentiment (ICS) has been below 80 for almost 3 years. No president has ever been re-elected with the ICS below 80 on election day. If the ICS stays low, Trump may be elected president no matter what.
What could move the needle for Biden? 12 Aprils ago, the ICS was even lower at 76.4 before falling further that summer of 2012. But, the ICS rebounded to 82.6 by October. President Barack Obama was re-elected along with Vice President Joe Biden. (2012 election post coming in September.)
Americans’ economic mood improved in the autumn of 2012. Biden needs a similar shift in sentiment to consolidate his base. If, and it’s a big IF, we see the Index of Consumer Sentiment (ICS) rise over the rest of the year, my ESM projection of Biden’s floor vote will rise. Combined with swing voters repelled by Trump, Biden’s share of the 2-party vote could reach the 50%+ range needed to win the Electoral College.