|
Public Superforecasts
|
| Featured forecasts from Good Judgment’s Professional Superforecasters. These forecasts are just a sample of the 63 currently available with FutureFirst™, our subscription forecast monitoring tool. For unlimited access to FutureFirst or to pose your own questions, please contact our sales department. Media and government organizations can request partial access here. Probability values are updated daily at 7:00AM US Eastern time. |
Get these forecasts delivered to your inbox each day: Subscribe |
| Will Fitch, Moody's, and/or S&P announce that the US has defaulted on its debt before 1 November 2024? | Today's Forecast | 1-week Change |
|||
| Today's Forecast: | |||||
| A | 0% | 0 | |||
| B | 100% | 0 | |||
Forecast History:![]() | |||||
BACKGROUND: Opened 17 March 2023 With a political fight looming over the US debt ceiling, there are concerns that the US could default on its trillions of dollars of debt. Fitch, Moody's, and S&P are the "Big Three" global credit rating agencies, which assign ratings on a borrower's likelihood of repaying its debts. Restricted and selective defaults would count, as would a technical default (see "RD"; see "SD"). Examples of Superforecaster commentary in italics This is a preview only.
FutureFirst subscribers enjoy full access to Good Judgment's standard reports containing the consensus forecast and Superforecasters' rationales since the launch of the question. Learn more about FutureFirst to follow questions that matter to your organization. [Start of Preview] 5 June 2023 - Good Judgment's professional Superforecasters see a 1% probability that the US will default on its debt before 1 November 2023. The base rate indicates this would be unprecedented, although there have been multiple close calls. SUPERFORECASTER COMMENTARY HIGHLIGHTS: 02 Jun 23 - Comment: We appear to have a two-year agreement in place, thus forecasting this at zero. 30 May 23 - Comment: The debt limit deal clears House rules committee. There had been rumors that McCarthy, in negotiating his speakership position, had given assurances that he wouldn't rely on Democrat votes to clear such a committee hearing, so I was watching. Don't know whether the rumors were true, and if so, what it bodes for McCarthy's longevity. 29 May 23 - Comment: I'm a bit surprised they settled on a deal rather than resorting to the 14th. Not outside the realm of possibility that we could see a last minute confusion along the lines of the 29 September 2008 TARP balk, but I think the votes are ultimately there. 27 May 23 - Comment: The deal, if it happens, is for two years - something that precludes the same situation in 2024, which was my main driver for being relatively high. 21 May 23 - Comment: Yellen says that invoking the 14th Amendment would cause a constitutional crisis and would be among the "not good options." Biden says there's no time to invoke the 14th Amendment. Republicans and Democrats are negotiating, but it's not clear to me that an agreement will be reached. While allowing the US Govt to default would be catastrophic for the US and the world, it's not clear to me that all of the parties involved in making decisions that could lead to a default understand that. 11 May 23 - Comment: The fact that the 14th amendment and/or platinum coin 'solution' is being kicked around this time as a potentially viable, if not ideal, option makes it far more likely (in my opinion) that it will be used, because it creates a dynamic in which McCarthy can refuse to budge--and in doing keep his position as speaker--while (probably) not triggering an actual default. Put simply: it's a lousy option (there's a reason it hasn't been used before), but political battle lines are so hardened that it's a better option for Biden than giving in to McCarthy, and a better option for McCarthy than giving in to Biden. Also, see Summers on chance of technical default (2-3% he says). 10 May 23 - Comment: If there's no deal, I think there are two possible outcomes: (1) Biden goes down in history as the first president who sides with the 14th Amendment. (2) Biden goes down in history as the first president who defaults, torches the US' credit rating, and causes a worldwide recession. I think if there's no deal, Biden will pick door number 1, which would cause disruption (and possibly a smaller credit rating downgrade) but nothing like door number 2. This would also permanently end the games of chicken over the debt limit. 09 May 23 - Comment: Warren Buffett says they will never allow that to happen. Others say that the threat to let it happen is larger next year, because it's an election year, so whatever happens in the next few weeks will not let all steam out of the kettle. 05 May 23 - Comment: My forecast was so high (15%-33%) for so long (3/18-5/4) because I thought that Biden wouldn't engage in negotiations over raising the debt ceiling (P=60%) but wasn't sure that Plan B would work (P=50%). So, .60*.50=30. After all, none of the Plan B options had ever been tried before. It's possible that choosing those options, which look feasible on paper, would still lead to default. It took some reading on the subject to convince me that two of the Plan B options (platinum coin, premium bonds) would work in practice. I'm still uncertain about the feasibility of the third option (14th Amend.). Revised forecast: Biden is likely (P=60%) to order the minting of the platinum coin or the issuing of premium bonds or simply decide to ignore the debt ceiling by invoking the 14th Amendment. The first two options would require the cooperation of Jay Powell, the third the tacit or explicit consent of the Supreme Court. If I had to bet, I think he would go with the premium bonds option. Still, there's a significant chance (P=30%) that Biden, like Obama in 2011, reaches a deal with the Republicans that would enable the US to avoid default but would also result in the US missing its climate targets by an even wider margin. There's a small chance (P=5%) that the Democrats can successfully pull off the "discharge petition" maneuver. Finally, there's a chance (P=5%) that Plan B, negotiations, and getting a few Republicans to cross the aisle to vote for a clean bill all fail and the US defaults. 04 May 23 - Comment: Even if party leaders can find some middle ground, not convinced either caucus is prepared to back down. Nonetheless, the consequences of a default are huge, so there's much reason to find a path forward. 03 May 23 - Comment: Washington Post article "Debt ceiling showdown: 5 possible outcomes" provides outcomes that could help frame default probabilities. Outcome 1, Biden-McCarthy deal. Difficult to achieve due to distance between two positions and House dynamics. May set up another showdown in 2024, still in our question time frame. Outcome 2, End run in Congress. Lawmakers use a "discharge petition" to call a bill to the House floor without speaker approval. Requires support of 218 members, so some Republicans. Also must pass the Senate. Outcome 3, Bill to suspend. Allows time to negotiate a deal. Just an interim path to one of the other outcomes? It would shift the potential default date backwards, but probably not enough to move it out of the question's time frame. Outcome 4, White House acts on its own. Two scenarios mentioned are: 1) mint new money to pay debt and 2) Invoke 14th amendment. Seems better than defaulting, but does the White House exercise this option before default or after? Outcome 5, US default. Both sides may decide the other will take a bigger political hit. Chris Rupkey, chief economist at research company FWDBonds, estimates "investment markets" would fall 5 to 7% within hours. Likely to result in a recession. 24 Apr 23 - Comment: We are in a brinkmanship/chicken scenario on the debt limit. Playing chicken is easier if (you think) you have a good plan B, or if the (perceived) personal consequences aren't high enough. Both factors may be in play. Plan B scenarios: The GOP has floated the idea that default wouldn't be that bad because spending could be prioritized. There have also been arguments the administration could just ignore the debt limit. If this were the Trump White House I'd expect that approach, but Biden could try it. Consequences: Since a major default hasn't happened before, actual consequences could be perceived as limited. More importantly, to the extent blame can be shifted to the other side, default "doesn't matter". Both sides might think they have the upper hand. Credit Default Swap Prices: Implied default risk is higher than earlier debt crises. The price of 1 year Credit Default Swaps is higher than at any point since the beginning of the available time series in 2008. (1 yr CDS closed at 90.05 on 04/20/2023.) Forecast: The current stalemate causes me to raise my chance of default (10%) approximately proportionally to the increase in CDS price, plus a bit for the time frame of the question. My prior forecast was base rate-driven at 5%. A final point is that there is about a year and a half left on the question, so we may have multiple bites at the apple. 22 Apr 23 - Comment: 17% based on the long time frame + shaky US financials + volatile US politics. 13 Apr 23 - Comment: Common sense says 'no way', but this is the highly divided US Congress we are talking about. From CNN: "A US default would have such devastating economic and financial consequences that many observers dismiss the possibility out of hand. But investors are not ruling out such a nightmare scenario." 08 Apr 23 - Comment: Sources: (1) "Republicans largely shrug off warnings on debt ceiling as they smell blood, saying Biden's 'poll numbers are in the tank and they're going to keep going down'" (Fortune); (2) "The House Republican Proposal to Avert a Debt Ceiling Crisis Is Untested and Unworkable: Proposals to prioritize certain payments in lieu of increasing the debt ceiling would increase, not decrease, the federal government's risk of default" (American Progress). 24 Mar 23 - Comment: The only conceivable way this happens is Congress refuses to raise the debt limit. Politically, the consequences would be disastrous. 21 Mar 23 - Comment: Nearly every congressional district has at least one small/regional bank, and they are often quite well connected politically. I don't think stress in financial system is done (despite recent assurances from FDIC), but even some of the more conservative GOP members will be getting calls from their districts to make sure there's no default/new banks don't come down. So I think recent stress in small/regional banks marginally decreases the likelihood of a long GOP holdout on debt ceiling. 19 Mar 23 - Comment: No cases so far, although it has been close very often. But this time it looks worse than ever, with the real danger of this happening as early as this summer. So will the House risk a recession and the loss of millions of jobs? Starting with 20%, ready to reduce to close to zero, if there is an agreement soon, or go further up on possible bad news. [End of Preview]
| |||||
| What will be the US real GDP growth rate in the fourth quarter of 2024 relative to the fourth quarter of 2023? | Today's Forecast | 1-week Change |
|||
| Today's Forecast: | |||||
| A | 6% | 0 | |||
| B | 30% | 0 | |||
| C | 49% | 0 | |||
| D | 13% | 0 | |||
| E | 2% | 0 | |||
Forecast History:![]() | |||||
BACKGROUND: Opened 21 January 2022 The question will be suspended on 31 December 2024 and the outcome determined using the Bureau of Economic Analysis' advance estimate of the real GDP growth rate in the fourth quarter of 2024 relative to the fourth quarter of 2023, typically released in late January. In January 2021, the advance estimate of the real GDP growth rate in the fourth quarter of 2020 relative to the fourth quarter of 2019 was -2.5% (see Table 6, line 1). The Federal Open Market Committee releases quarterly projections on GDP; see "Projection Materials" here. Examples of Superforecaster commentary in italics This is a preview only.
FutureFirst subscribers enjoy full access to Good Judgment's standard reports containing the consensus forecast and Superforecasters' rationales since the launch of the question. Learn more about FutureFirst to follow questions that matter to your organization. [Start of Preview]
AT A GLANCE: 6 June 2023 - According to Good Judgment's professional Superforecasters, there is a 48% probability that the US real GDP growth rate in the fourth quarter of 2024 relative to the fourth quarter of 2023 will be more than 1.5% but less than 3.0%. The Superforecasters also see a 31% probability that the growth rate will be between 0% and 1.5%, inclusive, in this period. Factors that would make the Superforecasters assign more probability to higher growth rates (currently 15% probability in total) include the effects of government spending and a slowdown in 2023 followed by a rebound in growth. The effects of a tight labor market and the possibility of a recession or stock market/real estate market correction, on the other hand, are the factors leading to lower growth projections in this forecast. SUPERFORECASTER COMMENTARY HIGHLIGHTS: 06 Jun 23 - Comment: The consensus logic seems to have been that there will be a recession in 2023 or early 2024 with a recovery later in the year. However, a recession in the next few quarters is looking more and more unlikely. Per WSJ, there are several indicators that any recession will be delayed. 30 May 23 - Comment: Looks like the much anticipated recession is continually being pushed back (if it happens at all). This should lead to growth being a little better than I thought in 2023 and a little softer than I expected in 2024. Fiscal policy will also provide less of a boost. 17 May 23 - Comment: The economy still seems to be doing reasonably well. For example, retail sales are up. However, there are growing signs of an economic slowdown, e.g., a growing number of small businesses have hit the brakes on hiring. All this suggests economic slowdown - maybe a recession is coming. If so, the growth in 2024 should be significantly affected. The average length of a recession is 10 months. Assume we don't see a recession until at least the 3rd quarter of 2023, then we should expect 2024 growth to be lower as well. 03 May 23 - Comment: The economy is showing surprising resilience. It seems unlikely that demand will slow dramatically until unemployment increases. But given the demographics, unemployment is likely to remain low by historical standards. 10 Apr 23 - Comment: Looks like we are going to get a gradual slowdown and soft landing by late 2023 into early 2024, given the strength of the US labor market and service sector plus the growth rebound in Asia. I don't see a rapid recovery by end 2024. More like a baseline growth scenario which is something like 1.5 pct or 2 pct. 31 Mar 23 - Comment: Uncertainty about banks after SVB, etc. It could mean the second half of 2023 will see weak growth, and in comparison, 2024 will look better, but recession could come later as well. 13 Mar 23 - Comment: The sudden prospect of Fed rates topping out this year below 5% could mean there won't be much of a recession, or that it does come in 2023. Until last week, it looked like the Fed would need to aggressively raise rates all this year as inflation persisted and the economy did not cool down. That left open the possibility of recession holding off until late this year or even next year. The new post-Silicon Valley Bank scenario looks more like a cautious Fed tolerating some inflation to avoid a financial crisis. That might allow the main street economy to go business as usual and avoid recession. The early recession this year scenario would be the result of a financial panic and dried-up liquidity. Either way, recession in late 2024 looks less likely now than it did a week ago. But maybe a booming economy with real GDP above 3% also looks unlikely. 04 Mar 23 - Comment: Unemployment rate and claims data in 2023 so far suggest that the US labour market continues to show resilience. A recession doesn't look particularly likely any time soon if the Fed stays on its current course. However, it could increase the pace of its rate rises (especially given the worrying PCE inflation data in January). Even if a recession is 'engineered' by the end of 2023, though, my central estimate is that it will be mild, which reduces the probability of Q4/Q4 real GDP growth in 2024 being greater than 4.5% (for that, you'd probably need the kind of aggressive hiking scenario we saw in the early 1980s or something that's of a similar magnitude to Covid). 14 Feb 23 - Comment: 2024 is likely to be a recovery year, but I don't expect a rapid recovery (more like 1.5-2pct), as unemployment will probably have peaked at high 4% or low 5%, there should be no extraordinary fiscal stimulus (other than shock absorbers), and the Fed will be very hesitant to cut rates too quickly. 07 Feb 23 - Comment: It seems any recession in 2023 is likely to be in the second half of the year and mild (rather than earlier in the year) and prospects for a raging, catch-up recovery in 2024 are lower in that scenario. With both fiscal and monetary policy acting as a drag in 2023, and probably only slight monetary easing in 2024, growth should be close to baseline: 1.5 to 2 pct. 06 Feb 23 - Comment: Here is another indication that 2023 GDP may be better than currently expected: Not only is employment robust, now housing is looking better. If housing can maintain or slight increase sales, it will also support broader spending on durable goods. Services are still growing fairly well. 28 Jan 23 - Comment: On balance, the case for a 'soft landing' continues to get stronger. The GDP figures from Q4 of 2022, the continued low level of new jobless claims, and the low unemployment rate can all be cited in favour of this judgment. And if the US can avoid a recession (or only endure a mild one) in 2023, then this lessens the case for a particularly large rebound. It could also be that a recession comes later than expected, perhaps in 2024 (which would favour growth being below 1.5%), but this is more speculative. 26 Jan 23 - Comment: Today's BEA advance estimate for Q4 2022 relative Q4 2021 was 1.0 after Q4 grew 2.9 vs the previous quarter, at an annual rate. Expectations continue to be that a recession is going to be here any minute now, even if a mild one. Those tech layoffs are about the strongest evidence of a slowing economy, but mostly due to a correction in that sector. While there's always chance of a recession, the timing late 2024 is not too likely. 27 Dec 22 - Comment: Sticking to a roughly base rate-informed forecast corrected for near-term projections of recession and recovery. There are a couple of events that could potentially justify a wider 80% or 85% confidence interval than we currently have. Firstly, the prospect of a war between China and Taiwan that disrupts trade. Secondly, the possibility of advanced artificial intelligence quickly generating higher economic growth (Microsoft's president has said that we'll see changes in 2023 that were expected in 2033; Bill Gates has said that AI will change the nature of white-collar work in the near future; and the OpenAI CEO Sam Altman has predicted that we'll see artificial general intelligence within the next few years and that there will be "significant disruption"). 11 Dec 22 - Comment: The OECD's economic outlook for the US sees growth at 0.5% in 2023 and 1.0% in 2024. "High inflation and tighter financial conditions will further crimp spending plans across the economy." 11 Dec 22 - Comment: With high interest rates attempting to slow the economy, the US is expected to enter a recession in 2023, thus the anemic GDP that is forecast. If inflation is defeated (not likely in one annual cycle), interest rates could slowly erode in 2024, improving GDP, perhaps to levels seen in 2022. I am assuming that FOMC rates will stay higher for longer and not decrease until inflation is seen to be on the run, maybe in 2024. 07 Dec 22 - Comment: Hard to see a greater than 3 percent growth - it does not appear we will have a deep recession in 2023 (so less chance of a sharp rebound) and Fed seems to shifting to a gradual approach which may keep rates and inflation higher for longer than the market is expecting (having negative impact on the 2024 growth). 06 Dec 22 - Comment: Upping answer option B to 24%. The Philadelphia Fed's quarterly survey of forecasters (which doesn't include us) has growth in 2024 at 1.8%, which is close to option B. There may be more an economic upswing towards the end of the year that the average may not take full account of. 20 Nov 22 - Comment: Most likely, we'll see a return to the slow growth of 2010-2019, 2%-3%. Recession is possible but more likely in 2023. Always some chance of an unexpected event like Covid or Ukraine war. 29 Oct 22 - Comment: Still too much uncertainty about inflation, recession, Covid mutations, Ukraine war, Taiwan (with Xi stronger and more determined than ever) for anything other than a broad distribution around 1-3%. 28 Oct 22 - Comment: The timing of a possible war in the Taiwan Strait could impact our forecasts for this question. If we think it's more likely to occur in 2024 than in 2023, it might cause us to update downward given the material risk of a global recession if such a war occurs. Other events such as the use of a nuclear weapon in Ukraine (perhaps more likely to occur in 2023 than in 2024) could push us in the opposite direction, though. 13 Oct 22 - Comment: Odds are in favor of a milder, shorter recession in 2023. 06 Oct 22 - Comment: This NPR story talks about the recent large drop in productivity. The causes it outlines could well have a long-term effect. If this happens, then we can expect a much deeper recession and a much slower recovery. 29 Sep 22 - Comment: Looking at early signals from the housing market, the US still appears to be on course for a recession in 2023. Noting another Superforecaster's thoughts (below) about whether higher interest rates will mean any 'recovery' is rather more modest than previous ones have been. This factor is why I think answer option C remains the likeliest option. 31 Aug 22 - Comment: Chairman Powell's Jackson Hole speech suggests that the Fed is comfortable with an economic downturn over the next 11 months or so, which in turn suggests that 2024, as most of us appear to be assuming, will be a recovery year. 18 Aug 22 - Comment: Support for the idea that a recession will hit the US in 2023 and we'll be in recovery mode by late 2024. 30 Jul 22 - Comment: There appears to be some debate about the definition of a "recession" in the United States, but the US is in an economic downturn on some metrics, and others (like unemployment, which is a lagging indicator) are likely to follow suit, especially as the Federal Reserve continues to tighten monetary policy. Therefore, as others have said, it seems increasingly likely that 2024 will be a recovery year. Historically, recovery years have been more likely to see real GDP growth above 4.5% than between 3.0% and 4.5%, but I recognise that there are some arguments in favour of the "natural" growth rate being lower these days. 15 Jun 22 - CBO boosts GDP growth projections for 2023 and 2024. Comment: Note the upbeat tone of the revised CBO report "included an implicit prediction that the Federal Reserve will be able to raise interest rates without tipping the U.S. economy into a recession." Seems like wishful thinking to me. 02 May 22 - Comment: Q1 2022 has reminded us that when forecasting real GDP growth, you must look at dollar GDP growth versus inflation. Q1 had positive dollar growth in GDP, consumer spending, and other indicators. But all that was offset by inflation. This problem may persist into 2023 with growth indicators being less than inflation rates. 12 Apr 22 - Comment: Recession fears jump around from week to week; at the moment, the increase in 10-year and 30-year rates relative to the short term (no inversion) implies less chance of recession. Still chances of a recession in either 2023 or 2024 are above the base rate. 13 Mar 22 - Comment: This article outlines some underlying drivers that might keep the US economic growth low for an extended period. Specifically, rising inflation driven by increasing commodity prices will absorb disposable income. 04 Mar 22 - Comment: Some pre-mortem thinking might be useful. Let's assume there was a recession in 2024. What might be the likely causes? One likely cause might be high interest rates. Historically, Fed interest rate cycles tend to last a few years. If the Fed interest rate increases don't slow inflation over the next year or so, then steady increases may lead to a tipping point where the economy goes into recession. Domestic Politics: 2024 is the presidential election year. There is currently lots of discord between parties and within parties. If we get political unrest after the primaries, would social unrest occur? This is doubly true for post election, but that will have less influence on the overall GDP unless it is severe. International Politics: We can only hope that Russia refrains from any further violence and Ukraine is resolved in the next few months. But there are other hot spots that might affect the overall economy. China and Taiwan would be the most disruptive. Housing Costs and Overall Inflation: Housing costs seem unsustainable. If we don't see at least a leveling, housing costs could eat up any discretionary income. Spending in other areas would drop. Currently, the economy is being buoyed by very high savings and record low debt due to the pandemic. But people are already eating into that money. When it runs out, demand will drop. If business doesn't plan properly, then over supply may cause a cut in production. Pandemic: It seems SAR-Cov2 is finally becoming "normalized." But the chance for variants still exists. Response: Another
pre-mortem possibility is a booming economy. We know that the Ukraine
war has affected food and key manufacturing components--not to mention
energy prices. Suppose those effects linger for another 1 to 1.5
years. If all this gets resolved, then the economy might boom in 2024
as buying picks up with increased supply and lower energy costs.
China's economy is currently having problems like too much consumer
saving and bloated real estate markets. If that all resolves in the
next two years, then increased demand in China might drive the world
economy. If Democrats take more political control, there could be
increased government spending. They were one or two votes off huge
spending increases in 2021. Broader conflict in Europe could cause a
pickup in the US economy to cover production losses overseas and
supply conflict torn regions.
28 Feb 22 - Comment: The average Q4 real GDP growth rate on an annualised basis, from 1948-2021, was 3.1%. Real GDP growth was below 0% in just under 1 in 10 of those years, between 0 and 1.5% in just over 1 in 10 of those years, between 1.5% and 3% in around 4 in 10 of those years, between 3% and 4.5% just under 1 in 5 of those years, and more than 4.5% in just over 1 in 5 of those years. Real GDP growth rates have therefore historically been more likely to land in Bin E than in Bin D, in part due to high values while the economy has bounced back from economic downturns. I've adjusted somewhat downward from these figures because real GDP growth rates more recently (say, in the 2010s have been consistently lower in normal times than those seen in the 1990s). 24 Feb 22 - Comment: Risk that the war in Ukraine causes a medium-term disruption that lasts as far as the end of 2024. 21 Feb 22 - Comment: Recognizing that inflation impacts GDP and that inflation may not quickly dissipate, and an over-reaction--whether by consumers, the government, or the Federal Reserve--could make things worse. 17 Feb 22 - Comment: 3-4.5% is the scenario where we have a recession in mid-2023 due to Fed tightening (possibly abetted by war), which makes 2024 an above-trend recovery year. 03 Feb 22 - Comment: Fortune argues that unemployment, inflation, and yield curves are similar those prior to other recessions. At the same time, looks like consumer spending patterns are already starting to return to normal, shifting toward services and reducing pressure on supply chain and inflation. 31 Jan 22 - Comment: In the base-rate sample back to 1990, a growth of 3.0% or higher hits 30% of the time. The Fed projects 2.0 for 2024. This far out, the Fed misses by 2.3% or more 30% of the time. 23 Jan 22 - Comment: The Bipartisan Infrastructure bill increases government spending on infrastructure over the next five years. If Build Back Better eventually passes (I give it a 55% chance), it will increase government safety net spending over the next 10 years. Because of this increased government investment, I'm anticipating better than average growth in GDP. 23 Jan 22 - Comment: In terms of labor supply, several things are holding us back. It appears many people have left the job market permanently. Add in Covid deaths. Then there is the anti-immigration sentiment which isn't going away. All this may make labor markets tight for the next few years. Also, it's unclear where trade policy is headed. Another factor will be the dollar's strength. There might also be a slightly higher chance of political/armed conflict as China and Russia continue pressure on border disputes. This certainly opens the door for economic disruptions. So, the distribution around the base rate might be a little broader than normal. 21 Jan 22 - Comment: My expectation is that the US will have returned to baseline growth (1-2%) by 2024, probably by Q2 2023, and probably at the lower end of the baseline range given slow labor force growth. Growth should be supported during H2 2022 into early 2023 by a global recovery. Even so, US growth may be more like 2-3% during that period rather than some sort of boom. Employment gains will be gradual, and consumers have already done a lot of durable goods spending. The fly in the ointment is the timing and potential for a recession due to Fed tightening and/or a stock market and real estate correction. A recession (or close to 0% growth scenario) for a couple of quarters seems more likely to be in early 2023 than 2024. [End of Preview]
| |||||
| Will the People's Republic of China (PRC) or Taiwan publicly accuse the other of using a weapon against its national military, militia, and/or law enforcement forces before 1 January 2025? | Today's Forecast | 1-week Change |
|||
| Today's Forecast: | |||||
| A | 26% | -1 | |||
| B | 74% | +1 | |||
Forecast History:![]() | |||||
BACKGROUND: Opened 20 September 2022 Tensions continue to increase between the PRC and Taiwan. For the purposes of this question, drones may be considered part of "forces" if so claimed by the accusing party's forces. Accusations of "warning shots" being fired would count. An accusation of a vessel or aircraft deliberately ramming another would count, but accidental collisions would not. Cyberattacks and the use of directed-energy weapons (e.g., microwaves, sonic weapons) would not count. Both the accused use and the accusation must occur during the question's open period to count. Examples of Superforecaster commentary in italics This is a preview only.
FutureFirst subscribers enjoy full access to Good Judgment's standard reports containing the consensus forecast and Superforecasters' rationales since the launch of the question. Learn more about FutureFirst to follow questions that matter to your organization. [Start of Preview] 8 June 2023 - According to Good Judgment's professional Superforecasters, there is a 27% probability that the People's Republic of China (PRC) or Taiwan will accuse the other of using a weapon against its forces before 1 January 2025. The Superforecasters point out the tensions remain high and are likely to continue to rise, but a full-scale invasion is unlikely in this time frame. However, an incident could take place in this period that could lead to an accusation without it being part of a full-scale conflict. SUPERFORECASTER COMMENTARY HIGHLIGHTS: 06 Jun 23 - Comment: I believe Xi will be relatively patient until at least after Taiwan's elections in January. There will likely be other provocations before strategically intended military engagement, but incidental weapons fire cannot be ruled out. Here is an article from the Institute for the Study of War on Taiwan's upcoming elections. 02 Jun 23 - Comment: China ruled Taiwan for 200 years and that is why they view Taiwan as part of China. It was only when the Qing dynasty became weak that Taiwan broke away in 1895. China wants it back to show that it is now back to the full strength. The public there supports that. With recent events like the buzzing of a US plane, an accident can happen. What if the aggressive maneuver were toward a Taiwanese military asset? 27 May 23 - Comment: I have seen the 2030 timeline mentioned by some Superforecasters in other articles. It fits with the narrative coming out of China that most mainland Chinese people view Taiwan as part of their country. They would not want to alienate the Taiwanese. And Putin has demonstrated how badly that can go. The argument that China may want to hurry up their timeline so Taiwan does not become further embraced by the West might have been true a couple of years ago, but I think that ship has sailed in the last year. However, I'm not going any lower than 25%. The rhetoric and military exercises are intense, and I don't think China is worried about Western sanctions anymore. 19 May 23 - Comment: China will probably be waiting for the results of the 2024 election. KMT is more likely to draw closer to China. If KMT were far ahead in the polls, I might decrease this forecast, but it is close. Considering the continuous military intimidation, an accident could happen. Western politicians' showing up in Taiwan is leading to ramped-up military exercises, thus increasing a chance of some exchange. China has been careful not to go too far, but the definition of "too far" can change. 17 May 23 - Comment: Rhetoric continues to be high, but I think China will wait and watch until after the Taiwan elections. And then there are lessons learned from Russia's invasion of Ukraine. 12 May 23 - Guardian: Liz Truss to visit Taiwan and give speech that could upset UK's China strategy. Comment: Every time a high-profile Western politician goes to Taiwan, China gets upset and ramps up military intimidation, which increases a chance of an 'accident'. Drones flying over becomes increasingly likely. Both sides will be careful not to wander into war at this time, but that is a lot of military hardware to control. 10 May 23 - Comment: An interesting article in Foreign Affairs about the dangers of groupthink in China: It begins with the now classical quote by William Burns about Xi ordering his military to be ready by 2027 to conduct a successful invasion of Taiwan if needed, then goes on to list the internal factors that might make for a very volatile outcome regarding Taiwan. 14 Apr 23 - Comment: China was practicing invasion but was careful in its three-day exercise. It did not board Taiwanese vessels as announced. On the other hand, with all that military hardware on both sides, an accident or misunderstanding could occur. 09 Apr 23 - Comment: China may do more patrols and inspections and that would lead to more contact. On the other hand, based on our Taiwan election question, the KMT has a good chance of taking power next January. 06 Apr 23 - Comment: New retaliatory measure by Beijing: "China's Fujian maritime safety administration launched a three-day special joint patrol and inspection operation in the central and northern parts of the Taiwan Strait that includes moves to board ships, it said on its WeChat account." If this was to become permanent, or if this were to be repeated from time to time, the possibilities for conflict would be endless. What happens, for instance, if besides inspection, there is also confiscation? What if Beijing decides to detain a container ship with chips for export to the US, for instance? 23 Mar 23 - Comment: Taiwan's President Tsai will visit the US next week. 17 Mar 23 - Comment: Taiwan has new types of drones. The Taiwan Strait may be clogged with military hardware. China continues to intimidate militarily. Accidents or misunderstandings do happen. If the independence forces in Taiwan win the next election, China is likely to make some sort of military move. Military practice that temporarily blockades Taiwan but do it over and over enough to be disruptive? It has been done. 03 Mar 23 - Comment: Back in September, Taiwan did shoot down a drone--it just was not identified as Chinese. This could happen again with it being attributed to the Chinese. Leaving my forecast at 35% for now. This is a long timeline. 02 Mar 23 - Comment: China seems content for now with its gray-zone aggression, but I'm recognizing things could quickly heat back up as they did last summer when Pelosi visited Taiwan. 24 Feb 23 - Comment: I will decrease my probability (35%) if we get through this time of increased tension between China and the US. China could increase its military preparation activities around Taiwan and that would increase a chance of a mistake or miscommunication. This is a long timeline, so trying to focus on actions rather than just rhetoric. 23 Feb 23 - Comment: The timeline is still long and things continue to heat up. If there is enough saber-rattling, someone will eventually make a public accusation. From katu.com: "The deployment is said to be the largest deployment of U.S. troops to Taiwan in decades, and reportedly aims to provide Taipei with abilities it can use to defend itself against Beijing." 16 Feb 23 - Comment: There's potential for quick escalation from Taiwan shooting down a spy balloon or drones to China using military force in some manner against Taiwan in retaliation. However, I still see any accusations in the near term to most likely be as a result of unintended interactions--like China and India along their disputed border. 07 Feb 23 - Comment: The US military does not believe China will have the capability to successfully invade until 2027 but, since this question is much broader than invading, I will remain at 35%. 06 Feb 23 - Comment: Xi appears to be preparing a soft sell of an invasion to Mainland China: "This month China's Eastern Theatre Command, which is responsible for Taiwan, released a music video called 'My war eagle flies around the precious island'. It depicts Chinese jets and warships surrounding Taiwan, while a singer softly calls for a lost love to return home." 04 Feb 23 - Comment: If 2025/2027 is the talk of the town of a potential direct military conflict between China, Taiwan, and/or the United States, it is understandable that a smaller-scale military confrontation may occur in the periphery of this timeframe, say, 2024, which is well-covered by this time frame. Then the bar is lowered a bit because: (1) drones may count; (2) warning shots would count (we saw a similar accusation/confrontation between Greece and Turkey last month or so); (3a) tensions are elevated given the Chinese spy balloon allegations and recent fighter intercepts of Chinese aircraft near American airspace (although this has been going on for years); (3b) I assess there to be misinformation/misinterpretation of the recent Chinese balloon sighting over the continental US, which may further inflame the political dynamics between China, Taiwan, and the United States. I know the question is about China and Taiwan, but it's hard to exclude the United States here. The US may be pivotal in how China reacts and stages its short-, medium-, and long-term plans. 23 Jan 23 - Comment: Moving up until after a McCarthy visit. A cat can't balance forever: "Ms. Hsiao acknowledged her difficult balancing act. In a play on the 'wolf warrior' label for China's new breed of aggressive diplomats, she called herself a 'cat warrior.' 'Cats can tread on tight ropes and, you know, balance themselves in very nimble and flexible ways,' she said." 15 Jan 23 - Comment: Warning shots/drones/other technical resolutions to the question are more likely than an isolated downing of an aircraft/vessel, which in turn is more likely than an attack. 13 Jan 23 - Comment: Long timeline. As long as status quo remains, this isn't likely anytime soon. The Taiwan election in 2024 will be telling. If independence leaders on Taiwan win, then problems are likely to result. More likely an embargo, cheap unarmed drone overflights, continual sabotage, and cyber-attacks to make life miserable in Taiwan rather than risking nuclear war with US. 10 Jan 23 - Comment: Decline in the number of US Navy freedom of navigation transits through the Taiwan Strait. 04 Jan 23 - Comment: Heightened tensions as others have pointed out, but I believe Xi is still trying for a velvet takeover. The pro-China faction in Taiwan did well in the last elections and the golden egg of Taiwan is the semiconductor industry. They can't risk that being damaged or sabotaged. 31 Dec 22 - Comment: Tensions continue. An accident is certainly possible. The US increasing arms to Taiwan is likely to increase tension rather than make China back off. On the other hand, one theory is that China cannot afford to attack Taiwan because it gets a lot of microchips from there. If the industry got ruined in war, where would it get them? Tracking how China is doing building that industry. This is a long time frame, and though experts have said 2027 should be the time frame for China to attack Taiwan, I think it could come sooner. China will wait for the 2024 elections in Taiwan to assess whether a negotiated path is possible. If KMT does not make inroads, then chances go way up. 18 Dec 22 - Comment: The Biden-Xi meeting went well, but this doesn't change the probability of accidents and warning shots. 17 Dec 22 - Comment: Lowering a tad though China continues to buzz Taiwan with bombers. Even nuclear bombers. Taiwan population wants status quo and last election voted more for the KMT, which favors negotiations with China over declaring independence. It is possible the KMT will win next presidential election and there will be an opening of more commerce and cross strait contact. Until then, China will keep intimidating. 14 Dec 22 - Comment: A Rand researcher credibly concludes China's Taiwan policy continues to be takeover without military invasion, essentially the same as it has been. He points out that Xi's autocracy is less complete than Putin's, Xi's behavior has been less warlike than Putin's, and that Xi has not tried to drum up pro-invasion sentiment on the part of the public, as would be expected. 03 Dec 22 - Comment: The KMT did well in local elections. Presidential election is in 2024. If the KMT candidate wins, China may think they will be able to influence the acceptance of the "one nation two systems" and then possibly apply the Hong Kong scenario. On the other hand, any move for total independence will bring on an immediate invasion. 30 Nov 22 - Comment: I don't see an invasion in the next two years, but continued harassment could well result in an accusation. Thinking back to the Azerbaijan/Armenia example where a flare-up at the local level took place even as talks were ongoing. 20 Nov 22 - Comment: Use of a weapon is a low bar, especially when warning shots count. So, although I agree China is playing it slow, it seems easy for this one to fall across the line. 19 Nov 22 - Comment: Tension lowered a bit after the Biden-Xi meeting. Xi will not wait forever, however, as the population grows ever more a sense of being Taiwanese rather than Chinese. Factors: (1) China building a better self-sustaining domestic economy. Not as export driven. (2) China having at least parity with US military. (3) Party in control of Taiwan is not willing to negotiate the "two systems one country" approach that China can later erode. Those factors will take time. If Taiwan declared independence, China would attack the next day, no matter when that is. China will continue to militarily harass Taiwan to remind the people there the devastation they will face if they go too far in breaking away. Due to all the military hardware, an 'accident' or minor incident like shooting over the bow or ramming is possible. 16 Nov 22 - Comment: Interesting long read from the New Yorker. Includes discussion of China's aggressive military flights, building up of atolls, prospects for blockades, past artillery duels, history, politics, etc. 12 Nov 22 - Comment: Local elections might also give a clue as to which way the wind is blowing. The KMT Party is much more China-friendly. If they are ascendant, then China may have more hopes of settling the issue through negotiation over time. 05 Nov 22 - Comment: Noting the reshuffle in Chinese military command to the army and weapons acquisition, so that seems pointed at invasion capability. In other words, Taiwan. 03 Nov 22 - Comment: China wants to unite as soon as possible. Taiwan wants to remain independent. The threats have stopped for the time being to Taiwan's air and water space. The CCP Conference has ended with Xi in charge solidly for at another 5 years. Xi is strongly in favor of unification. Will Xi attempt it within his third term at the height of his power? Has Ukraine's response to Russia put a pause in Xi's thinking? Humiliation on the world stage is usually not good for a leader's career. Something to ponder and perhaps push out any rash plans. 02 Nov 22 - Comment: Found this article informative for warning signs that an assault or invasion is imminent. Worth a read in its entirety. However, this question includes incidental interaction far short of a military attack. 17 Oct 22 - Comment: Low bar for resolution. Xi opened China's Communist Party Congress applauding his government's success in turning Hong Kong from "chaos to governance" and in countering "independence forces" in Taiwan, clearly signalling his intention to continue down the path of unification. I expect Beijing to not only maintain pressure on Taiwan but ramp it up, thereby increasing the risk of warning shots, deliberate ramming, or more significant military engagement in the next couple of years. 16 Oct 22 - Comment: An all-out attack on Taiwan in this timeframe seems unlikely. It would be a big risk for China with a lot of downside. There are other ways this could resolve though--a conflict in the air or at sea, or China taking Kinmen, or warning shots. That is where most of my probability is on this question. 16 Oct 22 - Comment: It may well depend on who wins the next election in Taiwan in 2024. This may be disinformation, but the mainland China-leaning KMT candidate is ahead at this point. If a separatist wins, then the chance of China attacking goes up. Until then, China will just practice and intimidate. They may wait until they have more of a military parity with the US. This is said to be 2027. If Taiwan declared independence, there would be an immediate attack. 09 Oct 22 - Comment: An attack by either side in this timeframe seems very unlikely because the PLA will not be ready yet to take Taiwan (the earliest estimate I have seen is 2027) and the PRC leadership will not want to launch an attack that spectacularly fails. If anything, Xi being confirmed for a third term makes an attack in the near-term less likely, in my opinion. That will more or less set him up as leader for life, so he can take his time to get it right and invade by his fourth term. 03 Oct 22 - Comment: Once the Party Congress ends and the new power structure is in place, anything can happen. One path could be younger, more nationalistic leaders for whom bravado would be a goal. Another path could be younger, enlightened leaders for whom soft world power would be the goal. There are also paths in between which may or may not focus on Taiwan. I'll start with optimism that the new leaders will take a world view recognizing that it is China's interest to nurture a peaceful planet. 02 Oct 22 - Comment: One of those things could happen if China took a Taiwanese-controlled island or blockaded Taiwan. The election in Taiwan in 2024 is likely to draw more Chinese military action around Taiwan as intimidation. Staying relatively high. On the other hand, China and Taiwan do a lot of trade. 27 Sep 22 - Comment: Firing warning shots against a drone in August would count, but I cannot find evidence of China accusing Taiwan. High probability of an event happening, low probability of accusation. 26 Sep 22 - Comment: Something is bound to flare up, or PRC may go ahead with an invasion after Xi's confirmation. I think Xi is planning it (as evidenced by prior rhetoric, success in Hong Kong, and recent scrimmages) to establish his legacy. Two years is not excessively long for an invasion, but long for the daily tensions. 24 Sep 22 - Comment: Long time left: 829 days, in which Xi might most likely decide that his role in history is to finally bring Taiwan "home." Also, the bar for a "yes" resolution is low. Quite a few possibilities: drones, warning shots, deliberate ramming, and, of course, an all-out war. How likely is a combination of drones being shot down, warning shots, and ramming? Adding to that a dictator with internal trouble possibly looking for an external enemy? Much too likely, in my opinion. Starting this with 40%. [End of Preview]
| |||||
| Will Vladimir Putin cease to be the president of the Russian Federation before 1 October 2023? | Today's Forecast | 1-week Change |
|||
| Today's Forecast: | |||||
| A | 2% | 0 | |||
| B | 98% | 0 | |||
Forecast History:![]() | |||||
BACKGROUND: Opened 14 October 2022 President Vladimir Putin has dominated Russian politics since he succeeded Boris Yeltsin as president in 1999, but has faced growing discontent in Russia as the war in Ukraine continues. Examples of Superforecaster commentary in italics This is a preview only.
FutureFirst subscribers enjoy full access to Good Judgment's standard reports containing the consensus forecast and Superforecasters' rationales since the launch of the question. Learn more about FutureFirst to follow questions that matter to your organization. [Start of Preview] 7 June 2023 - Good Judgment's professional Superforecasters see a 2% probability Vladimir Putin will cease to be the Russian president before 1 October 2023. The chief pathway to this outcome, they argue, is the possibility of deteriorating health. While the probability of a successful coup is low, and that of a successful mass uprising even lower, several Superforecasters suggest that the war in Ukraine, which is arguably going poorly for Russia, brings additional risks for Putin's tenure as president. This question was developed in collaboration with The Economist. SUPERFORECASTER COMMENTARY HIGHLIGHTS: 06 Jun 23 - Comment: Some opposition to Putin clearly remains, but not to the level to unseat him in the near future. 31 May 23 - Comment: Andrei Kolesnikov, a Moscow-based political analyst with the Carnegie Endowment for International Peace, describes Putin's regime as "increasingly authoritarian 'but with elements of totalitarianism,' and predicts difficult years ahead." Noting that Putin continues to control the media and the war messaging as Russia blames Kyiv terrorists rather than Russian dissidents for the recent attacks. 23 May 23 - Comment: Anti-Putin Russian nationals/saboteurs at work, but hardly a large-scale movement. 15 May 23 - Comment: Oil prices are down, so that should hurt Putin's effort to fund his war and also keep his domestic economy afloat. Putin likes to invoke WWII because the Russian people and troops were willing to suffer much in order to win. This is a different time and he has to be careful about widening conscription and one wonders how far standards of living fall before people are willing to push back despite his police state. This is in the long run of some years. I see no chance things will collapse for him by October. 05 May 23 - Comment: The next few weeks will be important. Putin's grip is tight for now, but a successful Ukraine offensive and continued infighting by elites could weaken his position. 18 Apr 23 - Comment: If he can visit the frontlines, it appears he is not that worried about being assassinated. War seems like a stalemate this year, but it is war, so who knows for sure? He can spin it as, "look at all this territory we clawed back for Russia," even if Ukraine takes back some of the Russian gains. He controls the media and police. Putin, again, has interlocking groups that check each other. Complain and you end up in jail for 25 years or thrown out a window. Chances are low that he will be out by October. 03 Apr 23 - Comment: One has to wonder if all the FSB people Putin purged and the way he keeps shuffling military leadership will come back to bite him. I see no sign he is losing grip on power, but Russian history is replete with leaders here today gone tomorrow with no forewarning. 21 Mar 23 - Comment: Passage of time and no changes to war, economy, or political situation. 10 Mar 23 - Comment: Like many authoritarian regimes before it, Putin seems to be strengthened by the squabbling and power plays within the wider security apparatus. There's no clear threat to him from the security elite and that's the only way short of death that he will be removed within this time frame. 06 Mar 23 - Comment: It may be worth reading this piece by Prof Michael McFaul former Ambassador to the Russian Federation. He goes out of his way to say that Putin will always have options, and because he is unchallenged at home, his definition of "always" is correct. 28 Feb 23 - AP: Drones fly deep inside Russia; Putin orders border tightened. Comment: Clearly, Putin has vulnerabilities and detractors, but autocrats are not easily deposed. 21 Feb 23 - Comment: There are two groups disaffected with Putin: the oligarch and educated elites who dislike the diminishing economy and the nationalist hawks angry at not winning the war so far. The elites are afraid of being arrested, so say nothing openly. The hawks do complain, but by shuffling generals, Putin demonstrates he is in charge. He did ramp up destroying the Ukrainian infrastructure when the hawks were complaining. 14 Feb 23 - Comment: There's a lot that we can't see; a mere 7 months to go for this resolution date still allows for compounded effects of the turbulent path he chose. From WP: "Russian President Vladimir Putin's war has set off a historic exodus of his own people. Initial data shows that at least 500,000, and perhaps nearly 1 million, have left in the year since the invasion began -- a tidal wave on scale with emigration following the 1917 Bolshevik Revolution and the Soviet Union's collapse in 1991. Now, as then, the departures stand to redefine the country for generations. And the flood may still be in its early stages. The war seems nowhere near finished. Any new conscription effort by the Kremlin will spark new departures, as will worsening economic conditions, which are expected as the conflict drags on." 04 Feb 23 - Comment: Putin maintains high approval and meaningful popular resistance during the question timeframe remains implausible. There have been rumors of health issues for a long time, but no reliable evidence has surfaced, to my knowledge. The base rate of assassination of world leaders is extremely low, less than 1% per year. I don't think his risk of exiting office is much higher than his risk of natural death, which is roughly 1-2% per year for someone his age. 31 Jan 23 - Comment: Base rate set to 1% from life insurance actuarial tables for a 70-year-old male's 2.2% chance of death over one year, cut in half because of the question's six-month duration. Taking that rate as a lower bound. There's no hard data about Putin having cancer that I've found reliable. But there's enough anecdotal evidence to consider him to be a 70-year-old with cancer risk. That in itself doesn't point to death in the next six months, but if I were an insurance underwriter, I'd add another one or two percent to the probability that he'd die in this period to cover it. That leaves the question of Putin being removed from office by constitutional means (0%) or non-constitutional (<1%) means over the next six months, which I'll just use for an upward rounding of any fractions. 31 Jan 23 - Comment: Russia's economy appears to be chugging along and Putin seems to be quite effective in his control of media messaging. Putin's September draft coincided with a dip in his approval from 83 to 77, which then slowly rebounded to December's 81. Another callup to fill military ranks will likely impact his approval rating once again, but I still don't see him being pushed from office when there are scapegoats in abundance to blame for the state of the war and no functioning opposition. 23 Jan 23 - Comment: It's likely that if Russia calls up more citizens for military service in Ukraine, and as the number of casualties filters through to the Russian public, Putin will become more unpopular with both the public and the army. It's unlikely that this will cause Putin any trouble, but in the past Russians have thrown out leaders who failed in wars. 15 Jan 23 - Comment: Putin has been good at separating power centers to play them off against each other. There are the technocrats, the siloviki, which belong to the spy agencies and the military, and the oligarchs. The siloviki are the ones who kill the oligarchs and technocrats if they disapprove of Putin. The siloviki though present the biggest challenge to Putin. Especially the hawk generals. That maybe why he shifts them around. Putin also has other means to deal with the siloviki. He has the Wagner Group, Kadyrov in Chechnia, and probably access to criminal groups as well. I think he can last to October. 13 Jan 23 - Comment: Given that the first 9 months of the war have been quite disastrous for Russia and Putin still seems to be clearly in power, I don't see why the next 9 months would change that. Further, Russia seems to be immune to some of the factors that normally cause leaders' downfall, such as a deterioration of the economic situation or an uprising of a rival faction. 08 Jan 23 - Comment: I do not see the general populace deposing him in a society that is used to dictatorship, corruption, and repression. If enough troops rebel or hawk generals are impatient for results, perhaps that would do it, but he has mechanisms to prevent that. Blocking troops that kill deserters or those surrendering are known to be part of his military. Generals have been replaced and people who oppose Putin keep dying. Here is a list from September. 06 Jan 23 - Comment: The only plausible way I see Putin going (apart from dying) is being deposed by the elite. I'm not confident of putting a low forecast in for this (<5%), we just don't know enough. It's really difficult to fathom what's going on in there, it's like Soviet Kremlin watching. The longer the war drags on, the more likely there will be factions within the armed forces and security services who may feel enabled to act. 17 Dec 22 - Comment: I think the war is going to grind on for a long time as the Russians lose very slowly. I still think Putin will be gone at the end of the war, but I now think that will take a lot longer. 16 Dec 22 - Comment: Putin still has the levers of propaganda and state power. The mothers' groups pushing back may be a real threat to him, but he handles that with propaganda by meeting with a group of hand-picked 'patriotic' mothers. 06 Dec 22 - Comment: No signs that I can see of this happening, but if it did happen, it could happen very quickly. 17 Nov 22 - Comment: The advantages of tight control over the media: "A latest poll showed on Friday that 75.8 percent of Russians approved of Vladimir Putin's work as the Russian president." 10 Nov 22 - Comment: Moving up slightly for now on possible repercussions from Kherson withdrawal. 06 Nov 22 - Comment: Putin's popularity has rebounded some; mobilization has ended (supposedly/for now). But the war and economic situation are likely to cause increasing criticism and increase his vulnerability over the next year. 04 Nov 22 - Comment: If Russia is forced to sue for peace, because its army is collapsing, I think it would be hard for him to survive. Yet there's no credible successor who can implement a peace deal: I think we can write off Medvedev by now and I don't see Prigozhin or Kadyrov doing it. 23 Oct 22 - Comment: Putin is a threat to elites. Elites might have to move on Putin to prevent themselves from being involved in a nuclear war. Current threat to Putin is posed by the hawk generals if they think he is not being aggressive enough. If India and China cut off support, his economy might be such that the public will rise up. Full mobilization might bring that on too, though he can pretend it is because of the West waging war on Russia. Propaganda works there. There are all kinds of possibilities, so his grip on power is not a lock. 19 Oct 22 - Comment: Outside of health reasons, it might take multiple republics/oblasts revolting and threatening to secede, combined with a scale of protests Russia has not seen in over 100 years. If the conflict in Ukraine ends really badly for Putin--and it's currently not going well for him--he could end up in a precarious political position. But if that led to him leaving office, I would not expect it to happen for a few more years. In 1980, Saddam Hussein started a disastrous war against Iran, which seemed to have no effect on his control. In what ways is Putin similar to and different from Hussein? I found this post intriguing. 18 Oct 22 - Comment: Autocrats don't go easily. Still, Putin's health seems questionable, his public support shaky, and the war in Ukraine is going badly. 18 Oct 22 - Comment: Mobilisation is unpopular and the war is now affecting the Russians. Too much uncertainty to assume that Putin will still be the president in a year. 17 Oct 22 - Comment: Deteriorating health is really the only plausible way he is no longer president in a year. The chance of a coup or assassination in the next year is near 0%, in my opinion--at least until we hear some evidence of a coordinated movement against him. 17 Oct 22 - Comment: Unlikely. In theory, Russia should be able to hold enough ground in eastern Ukraine for Putin to continue to frame the war as a victory of sorts. But it's not outside the realm of possibility that Western weaponry, intelligence, and sanctions (particularly those that ban imports needed for precision weapons)--paired with plummeting Russian morale and poor strategic decisions cooked up by Putin--will cause Russia to buckle faster and harder than anticipated. And if Russia is overrun in Ukraine, Putin will likely fall. 17 Oct 22 - Comment: Despite rumors to the contrary, Putin seems to be in reasonably good health. He is unlikely to step down voluntarily, and, not surprisingly, there is no heir-apparent. 16 Oct 22 - Comment: The anti-war leaders have left or are in jail. His mobilization has mostly been among the already more oppressed minority ethnic groups rather than in Moscow or St Petersburg. Connected people can get out of being drafted. The economy is kept afloat by selling energy to China and India, so while there are problems, it will not collapse. That leaves the hawkish generals. His recent attacks on Ukrainian cities and infrastructure have put them at bay for now. Russia is getting new Iranian weapons and if he can stop Ukrainian advances, he can probably keep the hawkish generals at bay. If not, he may be forced to use a tactical nuke to please them. The other group of all the dismissed FSB people and/or generals might move against him, but I note many potential rivals have fallen out of windows or had premature heart attacks. His public polling is still high enough for him to stay. 16 Oct 22 - Comment: I want to say 1%, but a lot could happen in a year. I keep thinking back to one of our forecasters who's shared the fate of leaders who start wars that don't end in their favour (they usually get sacked): "Using last two leadership changes [Gorbachev, Yeltsin], it's two years from crisis to replacement." 15 Oct 22 - Comment: Breaking the probability of this outcome into three buckets: (i) death by natural causes, (ii) removal by force, (iii) voluntary removal: (i.a) Actuarial table base rate of ~5%; (i.b) Adding another 1% for late-stage cancer rumors. (ii) 2% for all scenarios involving removal by force. (iii) I'd give this something like 0.4%, rounding to zero. (5 + 1) + 2 + 0 = 8% Response: Using actuary stats as base rates for heads of states is extremely misleading. E.g., even the chronic alcoholic Yeltsin lived to 76--when an average male in Russia at the time was expected to die at ~63. 15 Oct 22 - Comment: There is a very small chance of challengers organizing or of health problems or other outside factors, but I would say half of my assigned probability (7%) are unknown unknowns as war and its consequences are hard to predict. 15 Oct 22 - Comment: 1) Serious health issues or natural death. He's will never leave office voluntarily. At the age of only 70 and with the best medical care his country has available, he should stay healthy enough for 350 more days with a very high probability. The average 70-year-old Russian has 8.7 years left to live (link in German), and it should be longer for him. (3%) 2) There is no single person in sight to challenge or replace him. He has had more than 20 years to establish a huge security apparatus around himself, with all powerful people being handpicked old friends and KGB colleagues. Anyone with power and/or money in Russia has it because of him. Extremely unlikely for any of them killing him, unless the situation in Ukraine gets totally out of control (total humiliation or nuclear escalation). (2%) 3) External assassins are extremely unlikely to be able to get him (Close to 0%). 3+2+0.001=5% [End of Preview]
| |||||
| Will the World Health Organization (WHO) declare a Public Health Emergency of International Concern (PHEIC) regarding H5N1 avian influenza before 1 January 2025? | Today's Forecast | 1-week Change |
|||
| Today's Forecast: | |||||
| A | 3% | 0 | |||
| B | 97% | 0 | |||
Forecast History:![]() | |||||
BACKGROUND: Opened 7 February 2023 A PHEIC (Public Health Emergency of International Concern) is defined by WHO as "an extraordinary event which is determined to constitute a public health risk to other States through the international spread of disease and to potentially require a coordinated international response" under International Health Regulations, with recent examples including COVID-19, Monkeypox, and wild polio (apps.who.int/iris/rest/bitstreams/1031116/retrieve, linked here under publications). Examples of Superforecaster commentary in italics AT A GLANCE: 5 June 2023 - According to Good Judgment's professional Superforecasters, there is a 3% probability that the WHO will declare a Public Health Emergency of International Concern (PHEIC) regarding H5N1 avian influenza before 1 January 2025. Key drivers behind this forecast are: the low transmission to humans so far, even though HPAI H5N1 strains have been around for over two decades; greater surveillance; and the argument that more mutations of the virus would be needed to produce a pandemic among humans in the next two years. Risk factors include mink farm conditions that may be conducive to the spread between species and the WHO erring on the side of caution in declaring a PHEIC early. SUPERFORECASTER COMMENTARY HIGHLIGHTS: 23 May 23 - Comment: From Al Jazeera: "Over the weekend, the Health Ministry said samples of 33 suspected cases of avian influenza in humans in Espirito Santo, where Brazil confirmed the first cases in wild birds last week, came back negative for the H5N1 subtype." 18 May 23 - Comment: It's looking increasingly unlikely. The BBC reports that two workers at a farm with an outbreak among birds had tested positive, but they had not had any symptoms and had not spread human to human. This appears to be generally true. In which case there would seem to be little to be gained by declaring a PHEIC at this stage. 17 May 23 - Comment: 1 January 2025 is still a ways away, so I'm remaining at 10%. From the Guardian: "Bird flu could become the next human pandemic -- and politicians aren't paying attention." 12 May 23 - Comment: While HPAI H5N1 is absolutely a concern, I think we're far more likely to see a swine influenza strain cause the next flu PHEIC. I'm lowering my forecast from 2% to 1%. 04 May 23 - Comment: (1) More on the human infection caused by H5N1 in Chile (dated April 21): "A total of 12 contacts (close contacts and healthcare workers) were identified. All tested negative for influenza and have completed the monitoring period. No further cases in Chile have been identified so far." (2) The WHO's Western Pacific Region Avian Influenza Weekly Update Number 892 (dated April 21) once again includes the cases in Cambodia and China and the one death in Cambodia. 14 Apr 23 - Comment: Will a PHEIC for Marburg be issued before one for bird flu? 13 Apr 23 - Comment: (1) A recent H3N8 human fatality, but this virus is "unrelated to H5N1." (2) Another isolated human H5N1 case, in Chile. (3) Human-to-human H5N1 transmission would require the virus "to undergo a major transformation," according to this Science article. 05 Apr 23 - Comment: We've got some infected foxes in Germany, but, hey, they eat birds as do sea lions, minks, etc. Yes, H5N1 is more infectious to birds than any variant before, but the world has been living with that for months now. I still see no reason for a PHEIC in sight, if things don't get much worse (and why should they?). Even Marburg Virus might be a stronger candidate for a PHEIC as of today. 30 Mar 23 - Comment: We're asked about H5N1, but that seems like a narrow focus to me. Consider influenza D or there is H3N8. This look at how long it would take to detect spread of a novel flu virus in the UK, is a bit worrisome - could take 3-10 weeks, perhaps. 21 Mar 23 - Comment: Not seeing any signs of additional spreading to mammals or people, and the US discussing vaccinating birds to limit the spread. 17 Mar 23 - Comment: CDC technical report came out yesterday. Bottom line - concerning impact to birds and some animals, no signs of adaptation to humans. But need to remain watchful. 03 Mar 23 - Comment: CDC says watch your pets. The case in Cambodia and a new one in China appear to be a poultry-to-human transmission rather than a human-to-human transmission. The numbers so far are a long way from a PHEIC. 02 Mar 23 - Comment: Investigation into the Cambodia cases yielded a conclusion that the father and daughter who contracted H5N1 got it from birds, and it wasn't likely a human-to-human spread. However, I don't think these negative animal-to-human cases strongly correlate to whether the WHO will or will not declare a PHEIC. The bar for a PHEIC is lower, but we would need to see more such incidents for an H5N1 PHEIC to become a real possibility. 22 Feb 23 - Comment: Lowering forecast due to the WHO weekly update number 883. While there have been thousands of cases and deaths among birds, there have been no new human cases. However, the WHO considers zoonatic instances to be a viable consideration regarding public health. 21 Feb 23 - Comment: It's two years, and perhaps a PHEIC might be issued to increase monitoring efforts. But this is a little too much speculation about an individual virus when there is a whole wide world of candidates for the next big (human) pandemic. 21 Feb 23 - Comment: H5N1 clade 2.3.4.4b was first detected in January 2021. It appears to have increased transmissibility among birds. Seven human cases were detected since January 2022 with one death and four asymptomatic cases among them. Unlike humans, minks do possess weak receptors for H5N1. Which is why they are much more susceptible to the infections with H5N1 than humans. And even then, there is no solid evidence of mink to mink transmission of H5N1. 19 Feb 23 - Comment: The WHO won't declare a PHEIC without sustained community transmission internationally - so basically, person to person transmission that they can't anticipate through contact tracing, that is moving across national borders. With the amount of concern right now around just the poultry piece of this, for it to get to community human transmission, it would essentially need to be a different virus. They are working on vaccinating poultry. Public health researchers will certainly have vaccine candidates on the shelf should there be a turn toward human infection. 09 Feb 23 - Comment: Starting at 5%. Excellent comments, analyses of possible base rates, etc., below. I will, however, be on the high side owing to the fact that there is an important subjective component about when to declare a PHEIC. 08 Feb 23 - Comment: The "highly pathogenic" in highly pathogenic avian influenza (HPAI) refers to a strain's ability to cause severe disease in poultry, although in this case (and others), HPAI H5N1 also causes severe disease in humans. HPAI H5N1 strains have been around for over two decades. An HPAI H5N1 strain first caused a number of human cases in 1997 in Hong Kong. After Hong Kong's remarkable handling of the outbreak, further cases in humans weren't seen until 2003. Since then, fewer than 1000 cases in humans have been reported to the WHO. Other HPAI subtypes have caused zoonotic cases; perhaps the most notable among these is H7N9, which caused >1000 human cases, mostly in China, in 2013-2017, with very little human-to-human transmission. What is different now? In 2020 and 2021, HPAI H5N1 clade 2.3.4.4b viruses emerged and spread via wild birds in Asia, Europe, and Africa. By 2022, HPAI H5N1 had become the most commonly detected HPAI subtype in wild birds in Europe. In the US, HPAI H5N1 was detected in 2014-2015. With growing numbers of detections throughout North and South America, it appears that HPAI H5N1 strains are now circulating in the wild on every continent, with the possible current exception of Antarctica. With this enlarged territory has come increased numbers of detections of HPAI H5N1 in mammals. Detection is easier now than it has been in the past, and poorer countries often don't have enough resources for sufficient disease surveillance. HPAI H5N1 strains have caused infections in mammals in the past; without additional information, it would be hard to know whether strains and clades circulating now in various regions are more or less likely to cause infections in mammals. Obviously, the most important question to answer is whether currently circulating H5N1 strains are more readily transmissible to and among mammals. Clade 2.3.4.4b HPAI H5N1 virus was found to be circulating on a mink farm in the Netherlands in October, and this finding has caused many to worry that the virus is acquiring mutations that make mammalian infections and adaptation to mammals more likely. It is possible that three sea mammal die-offs in the past year could also have been caused by H5N1 transmission in these populations, although infections in those cases could have been caused by direct contact with infected birds or their feces. While the finding that this strain adapted to spread among densely housed mink is certainly concerning, it is less concerning to me than if an H5N1 strain had been found to circulate among a wild mammal population. I have not seen any evidence that any currently circulating HPAI H5N1 strains possess any mutations known to confer any increased ability to infect, replicate in, or transmit from mammals. Moreover, I have to wonder whether incidents like that observed at the mink farm in the Netherlands reflect greater surveillance and detection in Europe vs. that present in poorer countries in which HPAI H5N1 has been seen in the past. It's important to keep in mind that many flu strains to which humans have no immunity currently circulate in birds and non-human mammals, and many more that never jumped to humans have circulated in the past as well. HPAI H5N1 strains are very worrying because of their extreme potential to cause severe disease in humans, of course. But it's a big jump from birds to mammals, and especially from birds to humans. One potential scenario that concerns me about HPAI H5N1 is that a recombinant virus with the HA and NA genes from an H5N1 strain, which confer its H5 and N1 subtypes, respectively, could emerge to circulate in pigs, and from there could jump more easily to humans. To date, no such strain has been detected in pigs. Another thing to keep in mind is that virtually no human H5N1 cases have been detected in recent years. A total of 868 human H5N1 cases have been reported to WHO from January 2003 through 25 November 2022, but, "From 2020 to date [21 Dec 22], [only] six human cases of influenza A(H5N1) belonging to the 2.3.4.4b clade were reported to WHO." None of those 6 people has died. Thus, it is possible that the dominant H5N1 clade currently circulating may have a lower case fatality rate than has been observed with human H5N1 infections that occurred before 2020. Thus, while the chance that an HPAI H5N1 strain could cause a pandemic is and has been a serious cause for concern, H5N1 is not unique in this regard, with respect to either pandemic potential or disease severity. I am at least as concerned about H7N9 as H5N1 over the long term, for instance. And I do not see any special evidence that the risk of an H5N1 pandemic is much higher than it was, say, a year ago. 08 Feb 23 - Comment: The risk to humans is low, but the conditions in mink farms and other types of commercial farming create conditions favorable to mutations that could make infecting humans easier. An additional concern is an increase in other mammals with H5N1 from around the world, but no mammal-to-mammal transmission has been detected other than in mink. The hope is surveillance increases and governments take steps to prepare for human transmission that would make that scenario less deadly. At this point, I do not see reasons for alarm. This article uses the term "yellow caution light" and that seems appropriate. 08 Feb 23 - Comment: There is a PHEIC declared approximately once per two year period. After Covid-19 I would expect the WHO to be keen to ensure they err on the side of a prompt warning. Poultry farms and mink farms are perfect places for a virus to mutate - many animals crowded to together. We know other bird flu viruses have spread to humans. 08 Feb 23 - Comment: Applying Laplace's rule of succession gives us a probability of around 4% per year of a human outbreak of H5N1. In many of the years since its discovery in 1996-1997, there has been some spillover from birds to mammals, which isn't that concerning. However, the discovery of mink-to-mink transmission causes me to update upward in favour of there being human-to-human transmission, to around 8%. If there was human-to-human transmission, I expect that the World Health Organization would be very likely to declare a PHEIC especially if the case fatality rate is extremely high (40-60%). 07 Feb 23 - Comment: This question has a resolution risk with WHO designating a new strain to be H5N1 when it actually is not. Another risk is WHO lowering the standards and announcing PHEIC over something rather trivial. I am going to ignore all that and forecast an actual H5N1 producing pandemic potential in the next two years. Which is < 0.5%, in my opinion. The reason is highly technical and I can only summarize it as "too many successful mutations would be needed" (for those willing to delve deeper, see this) I would be a lot higher for "will a high-mortality flu appear within the next 50 years?" because that, when it happens (and it almost certainly will), will not be H5N1 for as long as we are sticking to the conventional nomenclature. 07 Feb 23 - Comment: The current concern is H5N1 clade 2.3.4.4b, which has been found in many different mammals. The largest reason for the current concern, beyond the devastation of bird populations, is probably the spread between mink that was seen in Spain. Of note, WHO is following many types of bird flu and there has been only 1 human case documented in the last year of H5N1 in that region. On the other hand, if human-human transmission is found, WHO would probably declare a PHEIC at a low number of cases. Currently, this looks like something to watch but not be overly concerned about. WHO doesn't generally rush to declare PHEICs, so seems like a relatively low probability of a declaration for H5N1. But this question is for 2 years, so a lot could change. 07 Feb 23 - Comment: H5N1 is a recurring virus. The fact that this is a bad year does not mean that the WHO will likely jump to declare a PHEIC. For it to be a human transmissible, we'd need some more mutations. Over the last two decades, there have been ~450 human deaths. The key wildcard here is whether we will see mammal mixing. While more mammal cases are being observed, I expect this to be due to increased surveillance. Key risk is whether the WHO wants to flex its new muscles. I do not think this is likely, however. | |||||
| How many deaths attributed to H5N1 avian influenza will the World Health Organization (WHO) report between 7 February 2023 and 31 December 2024? | Today's Forecast | 1-week Change |
|||
| Today's Forecast: | |||||
| A | 95% | +1 | |||
| B | 4% | 0 | |||
| C | 1% | 0 | |||
| D | 0% | -1 | |||
| E | 0% | 0 | |||
Forecast History:![]() | |||||
BACKGROUND: Opened 7 February 2023 There have been relatively few reported human deaths attributed to H5N1 over the past two decades, but there are fears that it could mutate to more readily infect humans and bring rise to a new pandemic. The question will be suspended on 31 December 2024 and the outcome determined using data including the entire pertinent time period once published by WHO (see also this). Examples of Superforecaster commentary in italics AT A GLANCE: 5 June 2023 - According to Good Judgment's professional Superforecasters, there is a 94% probability that fewer than 100 deaths will be officially attributed to H5N1 avian influenza between 7 February 2023 and 31 December 2024. They also see a 4% probability this number will be 100-1,000, inclusive. Key drivers behind this forecast are the historically low mortality rate (about 23 per year over the course of two decades) and low probability that H5N1 can infect many humans. SUPERFORECASTER COMMENTARY HIGHLIGHTS: 31 May 23 - Comment: 580 days left in this question. Avian flu is not adapted to infecting people, making human cases rare and person-to-person transmission even more difficult. While this was first detected in 1997, the number of human cases has remained low. Looking at a NIH Report covering 2003-2007, the number of deaths was under 100 each year. 23 May 23 - Comment: Lots of bird cases, practically no humans ones, and a paucity in animals eating birds. 14 May 23 - Comment: While I understand the reasons for concern, I don't believe the conditions will exist for a transfer of the disease to humans other than isolated contacts during this time frame. 10 May 23 - Comment: CDC update as of 3 May: Person to Person Infection None. Person in the US got it directly from a bird. 04 May 23 - Comment: (1) More on the human infection caused by H5N1 in Chile (dated April 21): "The genomic sequencing had 99.9% identity with H5 hemagglutinin sequences from Chilean birds, and complete neuraminidase (NA) had 100% identity with N1 sequences from Chilean birds. A total of 12 contacts (close contacts and healthcare workers) were identified. All tested negative for influenza and have completed the monitoring period. No further cases in Chile have been identified so far." (2) The WHO's Western Pacific Region Avian Influenza Weekly Update Number 892 (dated April 21) once again includes the cases in Cambodia and China and the one death in Cambodia. 13 Apr 23 - Comment: (1) There was a recent new human H5N1 case. (2) Influenza at the human-animal interface summary and risk assessment, from 27 January to 3 March 2023: "Since the last risk assessment on 26 January 2023, three human cases of infection with influenza A(H5N1) viruses were reported to WHO." 11 Apr 23 - Comment: Not H5N1 but avian flu - death from H3N8 in China, third known case ever. 10 Apr 23 - Comment: Three spots of good news: (1) Bird flu outbreak has slowed: "Luckily, the outbreaks so far have been less widespread than at this time last year, state and federal officials say." (2) France (and I assume others will follow) are preparing to mass vaccinations of farm birds. (3) It sounds like a human vaccine is close. 05 Apr 23 - Comment: No surge in sight, so we're down to the base rate. 02 Apr 23 - Comment: As it moves into more mammals, I believe there is increased chance that humans could either eat an infected animal, or come in contact with infected corpses. So even if it doesn't spread human to human, it could possibly creep above 100 by the end of 2024. 21 Mar 23 - Comment: No sign of spreading in humans, and hopefully there will be a vaccine available soon, at least for people working closely with birds. 20 Mar 23 - Comment: WHO reports no 2023 cases as of March 10. I'm not seeing a massive uptick generated by the spread in animals, and the probability of a mutation to human-to-human spread is very low on priors. 02 Mar 23 - Comment: Work has begun on a vaccine, just in case: "Although most experts say bird flu is not an immediate threat to humans, efforts are underway to produce vaccines for H5N1 or another potential pandemic virus." 02 Mar 23 - Comment: First human death from H5N1 recorded for 2023 (Cambodian girl), along with 5 cases thus far this year. UN: "From 2003 to 25 February 2023, a total of 873 human cases of H5N1 and 458 deaths have been reported globally in 21 countries." 24 Feb 23 - Comment: I think this one's a bimodal distribution. Either there is a pandemic or there isn't. If there is, then unless it begins near the end date of this question then there will almost certainly be more than 100,000 deaths by the end date, and probably much more. 17 Feb 23 - Comment: As of now, with the assumption of a single-digit probability of a serious outbreak, the distribution drops off quickly from bin A. But putting bin A only at 60% given the long timeframe and the unpredictability of the virus. 12 Feb 23 - Comment: This one has been a potential threat since 1997. I think it's more likely that another 'new' virus appears. 10 Feb 23 - Comment: Black Swan: bioterrorism. H5N1 could be weaponized and, at that, obscurely. What are the possibilities that H5N1 will be utilized in bioterrorism? There are a lot of ways germs, viruses, etc., can be weaponized. The idea that the Bird Flu can be weaponized is not an obscure thought. It can be made into an aerosol. Terrorists have plotted the use of crop dusters to spread pathogens (anthrax, etc). Some articles: Atlantic - The Next Pandemic Could Start with a Terrorist Attack MIT - Expert Knowledge in Intelligence Assessments: Bird Flu and Bioterrorism FAS - Science and Security: The Moratorium on H5N1 "Gain-of-Function" Experiments Journal of Terrorism Research - The Creation of a Contagious H5N1 Influenza Virus: Implications for the Education of Life Scientists NPR - Bird Flu Research Rattles Bioterrorism Field I believe this is the reason the question is focused on the H5N1 strain, and not the others. Geopolitical tensions are rising. If individuals not connected to the poultry industry become ill with H5N1, then this will alert us to a possible bioterrorism incident (and perhaps more occurrences of the same). 09 Feb 23 - Comment: During these last 20 years, we have two pairs of years with more than 100 deaths and another pair really close: 2005-2006: 43+79 = 122 2006-2007: 79+59 = 138 2007-2008: 59+33 = 92 That is, the total time with more than 100 deaths in a two-year period is more than 3 and less than 4 years. We can average to 3.5 or we might reduce this figure a bit because the time period in the question is about 23 months (instead of 24). This could be interpreted as having a base rate of 3.5 over 20, or a 17.5% probability for options different than A. Other factors: 1) The recent outbreak in mink farms, which suggests sustained mink-to-mink infection. However, apparently mink outbreaks are not particularly uncommon but have not up to this point ended up in a virus ready for human-to-human transmission. 2) The observation that it would be more difficult for H5N1 to adapt to humans than for other flu strains: this might be used to reduce the probability of the higher bins, which probably depend on sustained human-to-human transmission. However, this being a biological inside-view factor, reality might not comply to expectations as we have seen with Covid. 3) For the larger bins, the probability could be of the order of the base rate for a pandemic: maybe 1% per year? Well, it depends, because viruses generating 100k deaths should appear more often than viruses generating 10m deaths. But, independently of this factor, it cannot be too high: this 1% (or maybe more) per year has to be divided by the myriad of pathogens that could cause 100k deaths. Thus, a (1-2)% might already be an upper bound. 4) Also, for the H5N1 to cause a pandemic, besides being able to adapt to humans, it should also evolve as to be much, much less mortal. If we suddenly have a pandemic with a (25-50)% fatality rate, this is going to be on a different league than Covid-19: we will stay at home and self-confine voluntarily. 09 Feb 23 - Comment: The most important factor even in this case would be the virus's ability to mutate. While there have been 465 deaths over the past 19 years, most of them are before 2016. This would skew my estimate toward the lower bins. 08 Feb 23 - Comment: Noting that the base rate for ~2-year periods in which deaths exceeded 100 (2003-2022) would be 2/19, translating to 89.5% for answer option A bin and 10.5% for option B. Noting too that only 1 death was recorded in 2022 (despite the avian flu surge in the bird population over the course of 2022), perhaps making it more likely that the 2022-23 period will be an A bin resolution as well (given that the highest single-year deaths recorded in the 2005-6 and 2006-7 spikes was 79). 08 Feb 23 - Comment: This is deaths during almost 2 years, not a single year. In the WHO historical data of cumulative cases & deaths, each of the year-pairs 2005-2006 and 2006-2007 would have had more than 100 deaths. That's my reason for not putting even more in option A. 08 Feb 23 - Comment: Here is one thing the published papers didn't tell and the public (and ever the larger research community) does not appreciate about the scandalous gain-of-function experiments in 2011 that designed ferret-infecting H5N1: It was really hard to get that ferret-to-ferret transmission! There was a fan strategically positioned and working 24/7 to blow everything from a cage with infected ferrets directly into the cage of ferrets that were to be recipients. Even then, not all animals got infected. The point is: H5N1 is very well adapted to birds. To the extent that when we think we know almost everything about receptors and hemagglutinin's (the letter "H" in the virus name) specificity and are willing to introduce targeted mutations, it still does not work very well. H5 differs in many ways from H1 and H3 - to the point that there haven't been any "reassortment" recombinants observed between them (all such mutants have very low fitness in any host). The one-off infection is always possible, but the barriers for a sustained transmission seem to be high. Sure thing, it's possible that one day such a mutant appears - just very low probability. 08 Feb 23 - Comment: This isn't the first time that mink farming has been alleged to be a potential threat/source of animal to human virus transmission. The bigger question for H5N1 is whether it can spread in mink in such a way as to make it more likely to spread to and among humans. The potential for outbreaks and mutations seems to be well monitored and planned for. Without any evidence yet that it has mutated in a way to spread easily from mammal to human, let alone from human to human, favoring the lower numbers for this question. 08 Feb 23 - Comment: No matter how many sea lions, minks, or birds die from H5N1, the probability to deviate from the base rate looks very low. If there should be a mutation that allows humans to get infected much more easily, of course, all answer options are possible, but that looks very unlikely. 07 Feb 23 - Comment: H5N1 is not human-adapted and will not infect a lot of humans. Whatever does will likely not be H5. The base rate is ~23/year. 07 Feb 23 - Comment: WHO is following many types of bird flu. Total of 4 human cases of H5N1 last year, 1 death. This is quite a bit lower than the number of cases/deaths during 2004-2009. Here's probably the most useful data so far - historical cases and deaths. CDC officials see the current risk as low and say that the current viruses don't appear to infect people easily. The Avian Flu Diary blog discusses the case fatality rate - just over 50% of the H5N1 cases since 2003 have been fatal, but this varies a lot by location. And it probably doesn't account for mild or asymptomatic cases. Looking at the historical data from WHO, there don't appear to have been 100 deaths in any single year since tracking starts 2003. Obviously, that would change if a mutation in the virus made it more transmissible to and between humans. | |||||
| Copyright © 2023, by Good Judgment Inc. Good Judgment®, Superforecaster®, and Superforecasting® are registered trademarks of Good Judgment Inc. This presentation is solely for informational purposes. The information contained herein are not to be construed as legal, business, investment or tax advice. No representation or warranty (express or implied) is provided with respect to the accuracy, completeness, or reliability of this presentation. The future is inherently uncertain, and investors should exercise prudence and their own judgment in making investment decisions. Neither Good Judgment nor any of its directors, employees, or agents accept any liability for any loss (direct or indirect), including investment loss, or damage arising out of the use of the information herein. Any opinions expressed in this document may change without prior notice. |
.png)
