Welcome to the 16th volume of Metaculus Mondays, a weekly series at Global Guessing where we forecast questions from the community forecasting platform Metaculus.
After last week’s in-depth volume on the Israel-Palestine Crisis of 2021, our live update thread, and a vacation for one of us, we decided to take things a bit easier this week. Instead of forecasting new questions this week, we are reviewing a number of forecasts from previous volumes that we feel are in need of another look and likely an update.
We begin this week’s volume by reviewing an-oft updated forecast: Will the US Return to the JCPOA by January 1, 2022? First forecasted all the way back in volume 2 with a 20% likelihood, we have updated our prediction 3 times so far.
This question was incredibly relevant at the time of our initial forecast due to what its outcome signals about the future of Middle East power politics, newly-elected President Biden’s foreign policy aspirations, and potentially the legacy of former president Donald Trump. Now, this question has only grown in significance because of its importance to the ongoing Israel-Palestine conflict which has spiked in violence since the aforementioned forecast. All of these factors have led us to Recast this question.
Why are we updating this forecast? First, news coming out of negotiation talks in Vienna have provided a promising outlook on the potential of a JCPOA deal being reached soon. While the same reports also concede that there is much work to be done, and difficult issues yet to be ironed out, that level of confidence coming from a non-United States negotiating entity stood out to us.
Second, Iran’s upcoming presidential election in June will likely put pressure on getting a deal done for sitting president Hassan Rouhani. The problem in the forthcoming election isn’t that Rouhani needs to take votes away from a competitor, as he has reached his term limit according to Iranian law. But given that 32% of Iranians eligible to vote have voiced that they won’t this year, he will need a deal to reinvigorate the Iranian populace as well as to garner more weight for a potential endorsement from his party.
As a result, we update our forecast to a 33% likelihood that the United States rejoins the JCPOA by 2022.
In the coming weeks we will be closely watching this forecast as these factors are likely to cause some volatility in outlook in the near future. The outcome of next month’s election and the short-term progression of the Israel-Palestine conflict will either relatively dramatically increase or decrease our forecast. So stay tuned to see how we process the upcoming signals and alter our perspectives accordingly!
- Previous Forecast: 38%
- Community Median: 21%
For our second update (and spoiler alert: recast) this week, we return to the origins of SARS-CoV-2 and our forecast on whether the lab leak hypothesis is true. First forecasted in volume 6 with a 19% likelihood, we have updated this prediction 2 times so far and it has had one of the most consistent and aggressive changes of all our forecasts.
Why are we revisiting this forecast now? First, the Wall Street Journal recently reported that three researchers from the Wuhan Institute of Virology were hospitalized due to a sickness in November 2019. Second, Dr. Fauci recently said he was “not convinced” of a natural origin for COVID-19—a shift in his previous stance on the potential of a lab-based origin. His shift is one among many, and is a noticeable development. Furthermore, in light of recent acceptance and coverage of the lab leak theory there has been no substantial rebuttal. This is important because it shows a robustness to the theory.
Taken together, these developments have changed our forecast in two key ways.
- We have increased confidence in Rootclaim’s analysis on the origins of COVID; specifically, we increased our confidence on the evidence of chimeric structure and Furin cleavage from 80% to 85%.
- We decreased the penalty given by Rootclaim due to missing evidence from the WIV from 1/2x to 2/3x.
We also added Rootclaim’s additional analysis on COVID’s initial adaptability to our forecast.
At the same time, we are still unsure about the political and obstruction factors for positive resolution. First of all, China could continue to prevent any real investigation into the origins of COVID meaning scientists may be unable to gather enough evidence in the meantime to conclude with sufficient certainty in a lab origin. Moreover, there is still the risk that the public health agencies listed for positive resolution would refuse to endorse a lab-based origin due to geopolitical reasons. Therefore, we are still discounting our true forecast by 30%.
As a result we forecast a 58% likelihood that at least two major public health agencies claim it is more likely than not COVID-19 originated from a Chinese virology lab. (82% likelihood without discount factor.)
- Previous Forecast: 20%
- Community Median: 12%
First forecasted in volume 3 with a 21% likelihood, we have updated this prediction once so far.
There are a number of factors that we considered which resulted in our lowered forecast for this question. First, the ongoing fighting between Israel and Palestine, irrespective of a ceasefire, makes us less confident that Saudi Arabia would normalize relations with Israel. As the geographic and religious center of Islam, Saudi Arabia’s position on the Israeli-Palestine conflict is arguably the most consequential in all of the Middle East. As such, the Saudis will likely practice extreme discretion when choosing when to finally enter into an agreement. For Saudi Arabia there isn’t a rush to move closer to Israel and the United States so with the current violence and disturbances, normalization now is unlikely. The country has voiced as much in its support for the Palestinian cause.
In support of this, and a second factor in itself, Saudi Arabia has also been moving closer to Iran in recent weeks. This is relevant because it also signals a shift in Saudi Arabia’s geopolitics orientation from being a pole of power in the Middle East heavily influenced by the United States, to state with more autonomy and a foreign policy agenda of its own. This shift in outlook could definitely entrench Saudi Arabia in its position on Palestine, as well as potentially balance with more powers in the region against a powerful Israeli state. With the United States gradually pulling out of the region as well, Saudi sovereignty will likely spell a prolonged normalization process between it and Israel.
As a result, we lower our forecast by a quarter and predict a 15% likelihood that Saudi Arabia and Israel establish official diplomatic relations by 2022.
Questions on the 2021 Israel-Palestine Crisis
We finish today’s volume by continuing with questions relating to Israel by returning to the 6 forecasts made last week on the Israel-Palestine conflict. Since last week’s volume we have updated these forecasts 3 times since a lot has changed; namely, a ceasefire was put in place on May 20, 2021 and has remained in effect since.
- Previous: 38%
- New Forecast: 33%
- Community Median: 48%
We lowered our forecast for this question from 38% to 33% for three primary reasons. First, similar to some of the other forecasts in this set, if we apply the base-rate of deaths in the six months after the 2008-09 Gaza War, the 2012 Israeli operation in the Gaza Strip, and the 2014 Gaza War, the tally of deaths by the end of the year would not be sufficient for a positive resolution. Second, after fighting in early 2008 (an analog for nature of violence this year) major fighting broke out 5.5 months later. 5.5 months from now would be early November, 2021, making it difficult to reach the 400 death threshold. Finally, Israel is holding its second election of the year in October, and violence could not only disrupt the electoral process but also influence voter behavior in difficult-to-predict ways, so we see this election as more reason for dropping our forecast. We still sit lower than the Metaculus community median which is likely an indication of us using more discretion when distinguishing signals from noise.
Although the situation in Israel and Palestine appears to have settled in the short-term, this is a question we will have to keep a close eye on. On the community forecasting platform Good Judgement Open, there is a question posed by The Economist asking: “Before 22 July 2021, will Hamas or the Israeli government accuse the other of violating the Gaza ceasefire agreed to on 20 May 2021?”
The crowd forecast? 87%. We’re not sure that we agree with this forecast, but it is not a question we’ve spent a lot of time thinking about. As a result, we are taking their crowd forecast as a signal to be willing to change our position on this question (and the rest in the series) in potentially short order.
- Previous Forecast: 20%
- New Forecast: 18%
- Community Median: 19%
Our drop for this forecast from 20% to 18% is primarily a result of our forecast for 400 Palestinians, as our drop in likelihood is in line with that forecast. Given that we have been actively tracking this set of questions in particular, we updated our forecast for this question on May 20th, making the passage of time since our last forecast as a rationale for dropping our percentages somewhat negligible.
- Previous: 9%
- New Forecast: 10%
- Community Median: 10%
We ultimately arrived at our forecast for this question like a Sudoku question. We first solved for the question regarding 1,000 Palestinian deaths in 2021, and then used power-law distribution to arrive at our forecast for 10,000 Palestinian deaths in 2021. We then backtracked our calculus to work out this forecast for 2,500 Palestinian deaths which sits in-between the two other predictions. For context, 2,500 deaths would put this conflict at an equal level to the ongoing internal strife in Sudan which has persisted since 2011.
- Previous: 3%
- New Forecast: 3%
- Community Median: 4%
We decided to not change our forecast for 10,000 Palestinian deaths in 2021. This decision was informed by a combination of our updated forecast for 1,000 Palestinian deaths in 2021, alongside Richardson’s Law and the observation of power-law distribution in conflict deaths. For context, the War in the Donbass has been counted at just over 10,000 deaths since 2014, so 10,000 deaths in one year would require a significant violent uptick.
- Previous: 96%
- New Forecast: 94%
- Community Median: 91%
For this question we lowered our forecast from 96% to 94%, a relatively minor drop as we still remained very confident that this threshold for rockets fired at Israel will be met. Our primary detractors pulling down the forecast is that if the base-rate for rockets fired after the start of conflict was superimposed on this situation, the number of rockets fired would be insufficient for a positive resolution. With that being said, the nature of rockets firing is different in 2021 than it was in 2008 or 2014 for example, and the presence of Iran and other foreign powers in this conflict also raises the chances that a higher number of rockets will be fired this year.
- Previous: 30%
- New Forecast: 33%
- Community Median: 24%
For this question we raised our forecast from 30% to 33% due in large part to our base of prior data. If fighting resumed like it did in 2008, this question would resolve positively. If fighting followed the pattern of post-2012 and post-2014 conflict, then the question would resolve negatively. Also, given that Hamas and Palestine know how adept Israel is at destroying Hamas’ tunnels, they might impose more urgency and volume of their next attacks.