The last ten days have been mind-boggling, with the emergence of new political facts at light speed. Consider that we went from vague references, to a whistleblower complaint, to the launch of an impeachment inquiry, to the release of a “transcript” of the President’s conversation with the Ukrainian president, to the current conflict of claim and counterclaim.
As an example, in the early stages (Monday, Sept 23rd), we conducted a poll with our partner Reuters to gauge the public’s understanding of the scandal and support for impeachment (Reuters/Ipsos). I had two basic takeaways at the time. First, while a large chunk of the population was aware of the issue (73%), only a few felt they were well-informed (17%) about it. Second, support for impeachment remained low and materially unchanged from earlier this year.
However, our poll was “old news” the day it was released. After we concluded data collection but before we published, Speaker Pelosi initiated an impeachment inquiry, substantially changing the nature of the debate. As I discussed our findings for Fortune, the call “transcript,” followed by the whistle blower complaint, was released.
Here, it is important to stress that presidential impeachment is an extremely rare event in American history. Only two U.S. presidents have ever been impeached, and in the end neither was “convicted” or removed from office. This small sample size, however, has not stopped media pundits from providing their thoughts on all things impeachment!
At heart, I’m driven by data, and this imprecision (a sample size of 2) kept my mind racing late into the night this entire week. What do we really know about impeachment from an empirical perspective? Not much.
To overcome our sample size problem in U.S., we looked through a global lens and built a database of impeachments over the past 30 years, both inside and outside of the U.S. This database includes:
- Democratic countries with presidential systems where the chief executive was removed or an attempt was made to remove them through a legal or constitutional process as opposed to extra institutional means, like a coup.
- Additional variables such as an incumbent’s approval rating, economic growth, party control of the legislative body in charge of deciding impeachment, and what happened in the following election.
So what did we find?
- Even globally, impeachment proceedings are a very rare event. We identified only 21 cases that meet the above criteria since the end of World War II.
- An impeachment inquiry does not guarantee the removal of the president. Just over half (57%) of the presidents exposed to an impeachment process ended up leaving office, either through impeachment or resignation—leaving 43% who remained in office.
- However, not all impeachment processes are equal. Presidents who retained significant popularity going into impeachment proceedings usually, but not always, came out the other side still in power. Indeed, only 19% of presidents who left office had an approval rating of 40% or more.
- Control of the deciding legislative body is important. In only 14% of the cases – just three times – did a president leave office when his or her party controlled the deciding legislative body (e.g. the Senate in the case of the United States).
- Impeachment effects last beyond the proceedings, negatively impacting the party in power. We only found 3 cases, or 14% of those included in the database, where the party of a president who left office won the next election).
So what are the implications for Trump?
- Trump appears to be at the lower risk end of impeachment proceeding indicators. His approval rating is at 43%, and at that level only 19% of presidents have left office.
- Trump’s party (Republicans) controls the deciding body on removal (the Senate). Taking only our data into consideration, he has a 14% chance of leaving office.
- Putting aside ethical or legal sensibilities, impeachment might be a smart-odds play for Democrats. Indeed, there is very little evidence that the opposition party incurs a cost for initiating an impeachment process. If anything, the data shows that the parties out of power perform well in the general election immediately following an impeachment effort.
- Finally, approval ratings can change. I would pay special attention to the impact of any impeachment process on Trump’s approval numbers—even a few points consistently up or down could drastically change his odds of victory in the next election (Daily Beast). Additionally, changes in approval ratings also serve as signals to the political market about relative strength—a negative trend could carve away political support for Trump at this critical juncture.
Republican control of the Senate is the President’s Trump card. He is unlikely to be removed from office while his party controls the “jury”. However, impeachment is also a smart move for Democrats with few incumbent parties retaining power at the next election. Simply put, Trump might win the battle but lose the war.
See chart below displaying the probability of a president leaving office given certain conditions. This chart is based upon those presidents from our database who left office.
For more information, please contact:
+1 202 420-2016
Ipsos is now the third largest market research company in the world, present in 90 markets and employing more than 18,000 people.
Our research professionals, analysts and scientists have built unique multi-specialist capabilities that provide powerful insights into the actions, opinions and motivations of citizens, consumers, patients, customers or employees. Our 75 business solutions are based on primary data coming from our surveys, social media monitoring, and qualitative or observational techniques.
“Game Changers” — our tagline — summarizes our ambition to help our 5,000 clients to navigate more easily our deeply changing world.
Founded in France in 1975, Ipsos is listed on the Euronext Paris since July 1st, 1999. The company is part of the SBF 120 and the Mid-60 index and is eligible for the Deferred Settlement Service (SRD).
ISIN code FR0000073298, Reuters ISOS.PA, Bloomberg IPS:FP