Month: January 2019

Ethics, Model, Philosophy

Signing Up For Different Moralities

Posted on by Zach
When it comes to day to day living, many people are in agreement on what is right and what is wrong. Giving change to people who ask for it, shoveling your elderly neighbour’s driveway, and turning off the lights when you’re not in the room: good. Killing, robbing, and drug trafficking: bad. Helping the police to convict mobsters who kill, steal, and traffic drugs: good.

While many moral debates can get complicated, this one rarely does. Even when helping the police involves turning on your compatriots – “snitching” – many people (although notably not the President of the United States of America) think the practice is a net good. There’s a recent case in Australia where opinion has been rather more split. Why? Well, the informant was a lawyer – specifically, a lawyer who had worked with the accused parties. Here’s a sampling of commenters on both sides:

In this case I feel it is for the greater good that human garbage like Mokbel are convicted even if the system has to be bent to do so. [1]
The job requires strict adherence to the ethical rules. If you let your dog run the house, the house gets torn apart.
The brave lady in question went above and beyond to keep Victorians safer. If these thugs are released or sentences reduced there will be uproar.
The right to an open and fair trial is a hallmark of a democratic country even if sometimes a defendant who is in fact guilty gets acquitted.
While I’m normally happy to see violent mobsters go to jail, here I must disagree with everyone who offered support for the lawyer. I think it was wrong of her to inform on her clients and correct for the high court to rebuke the police in the strongest possible terms. I certainly don’t want any of those mobsters back on the street and I hope there’s enough other evidence that none of them have to be released.

But even if some of them do end up winning their appeals, I believe we are better off in a society where lawyers cannot inform on their clients. This, I think, is one of the ethical cases where precedent utilitarianism is particularly useful in analysis and one that demonstrates its strengths as a moral philosophy.

(To briefly recap: precedent utilitarianism is the strain of utilitarian thought that emphasizes the moral weight of precedents. Precedent utilitarians don’t just consider the first-order effects of their actions on global wellbeing. They also consider what precedents their actions create and how those precedents can be later used for others for good or ill.)

The common law legal system is premised on the belief that the burden of proof of crime rests upon the state. If the state wishes to take away someone’s liberty, it must prove to a jury that the person committed the crime. The accused is supposed to be vigourously defended by an advocate – a lawyer or barrister – who has a legal and professional duty to defend their client to the best of their abilities.

We place the burden of proof on the government because we acknowledge that the government can be flawed. To give into every demand it makes leads to tyranny. Only by forcing it to justify all of its actions can we ensure freedom for anyone.

(This sounds very pretty when laid out like this. In practice, we are rather less good at holding the government to account than many, including myself, would like. Especially when the defendant isn’t white. I believe part of why society fails to live up to its duty to hold the government to account is sympathies that commonly lie with police and against defendants, the very sympathies I’m arguing against holding too strongly.)

But it’s not just upon the government that we place a burden to avoid pre-judging. We require advocates to defend their clients to the best of their abilities because we are skeptical of them as well. If we let attorneys decide who deserves defending, then we have just shifted the tyranny. Attorneys can make snap judgements that aren’t borne out by the facts. They can be racist. They can be sexist. They can make mistakes. It’s only by forcing them to defend everyone, regardless of perceived innocence or guilt, that we can truly make the state do its duty.

This doesn’t mean that lawyers always have to go to trial and defend their clients in front of a judge and a jury. It could be that the best thing for a client is a guilty plea (ideally if they are actually guilty, although that’s also not how things currently work, especially when the accused isn’t white). If a lawyer truly believes in a legal strategy (like a guilty plea) and the client refuses to listen, the attorney always can walk away and leave the trial defense to another lawyer. The important thing is that someone must defend the accused and that that someone will be ethically bound to give it their best damn shot.

Many people don’t like this. It is obviously best if every guilty person is punished in accordance with their crime. Some people trust the government to the point where they view every accused as essential guilty. To them, lawyers are scum who defend criminals and prevent them from being justly punished.

I view things differently. I view lawyers as people who have signed up for an alternative morality. While conventional morality holds that we should punish criminals, lawyers have signed up to defend all of their clients, even criminals, and to do their best to prevent that punishment. This is very different from the rest of us!

But it’s complimentary to my (our?) morality. It is not only best if we appropriately punish those who break the law. I believe is also best if we do it without punishing anyone who is innocent.

We cannot ask lawyers to talk to their clients, figure out if they’re innocent or guilty, and then inform the judge or dump as clients all of the truly guilty. This will only work for a short while. Then everyone will figure out that you have to lie to your attorney (or tell the truth if you’re innocent) if you want to avoid jail. We’re now stuck trusting the judgement of attorneys as to who is lying and who is telling the truth – judgement that could be tainted by any number of mistakes or prejudices.

In the Australian case, the attorney made a decision she wasn’t qualified to make. She, not a jury, decided her client was guilty. She doesn’t appear to be wrong here (although really, how can we tell, given that a lot of the information used in the convictions came from her and her erstwhile clients weren’t able to cross-examine her testimony) but if we don’t want a system where a random lawyer gets to decide who is guilty or not, the important thing isn’t that her testimony is true. The important thing is that she arrogated power that wasn’t hers and thereby undermined the justice system. If we let things like this stand, we enable tyranny.

The next lawyer might not be telling the truth. He may just be biased against black clients and want to feel like a hero. Or she might be locked in a payment dispute and angry with her client. We don’t know. And that should scare us away from allowing this precedent to stand. A harsh rebuke here means that the police will be unable to use any future testimony from lawyers and protects everyone in Australia from arbitrary imprisonment based on the decisions of their lawyer.

Focusing on the precedents that actions set is important. If you don’t and instead focus solely on each issue in isolation, you can miss the slow erosion of the rights and freedoms that we all rely on (or desire). Its suitability for this sort of analysis is what makes precedent utilitarianism so appealing to me. It urges us to dig deeper and try to understand why society is set up the way it is.

I think alternative moralities, actively different moral systems that people sign up for as part of their professions are an important model to hold for precedent utilitarians. Alternative moralities encode good precedents, even if they stand in opposition to commonly held values.

We don’t just see this among lawyers. CEOs sign up for the alternative morality of fiduciary duty, which requires them to put the interests of their investors above everything but the law. Complaints about the downsides of this ignore the fact we need companies to grow and profit if we ever want to retire [2]. Engineers sign up for an alternative, stricter morality, which holds them personally and professionally responsible for the failures of any device or structure they sign off on.

Having alternative moralities around makes public morality more complicated. It becomes harder to agree on what is right or wrong; it might be right for a lawyer to help a criminal in a way that it would be wrong for anyone else, or wrong for an engineer to make a mistake in a way that would carry no moral blame for anyone outside of the profession. These alternative moralities require us to do a deeper analysis before judging and reward us with a stronger, more resilient society when we do.

Footnotes

[1] Even though I disagree strenuously with this poster, I have a bit of fondness for their comment. My very first serious essay – and my interest in moral philosophy – was inspired by a similar comment. ^

[2] This isn’t just a capitalism thing. Retirement really just looks like delay some consumption now in order to be able to consume more in retirement. Consumption, time value of [goods and services, money], and growth follow the same math whether you have central planning or free markets. Communists have to figure out how to do retirement as well and they’re faced with the prospect of either providing less for retired people, or using tactics that would make American CEOs blush in order to drive the sort of growth necessary to support an aging retired population. ^
1
Link Post

Link Post – January 2019

Posted on by Zach

Power generation and distribution is used in economics textbooks as the classic example of a natural monopoly. Lebanon proves this doesn’t have to be the case, with a mostly functional (and mostly illegal) parallel power distribution system that many people rely on to supplement the official system. I don’t think this should be used more generally – having a bunch of small generators is definitely more polluting – but it’s interesting to see that markets can work even in cases where people normally expect them to fail.

On the advice of a high school teacher, I once read most of the epic fantasy series “The Malazan Book of the Fallen”. One thing it made clear to me was just how dangerous and difficult it is to be an absolute ruler. Anyone who controls the information flow to you can essentially control you. By this token, it should be worrying that Xi Jinping is centralizing China around himself – leaving the government vulnerable to serious missteps if he’s given inaccurate information.

Speaking of China, here’s a story about two journalists accidentally winning a car chase in Xinjiang. Xinjiang is one of the most surveilled places in the world and up to one million Uighurs are currently held in concentration camps, which explains why the Chinese government is so on edge when journalists poke around.

Jackie Cochran had a life that reads like a novel; she founded the Women’s Airforce Service Pilots in World War II (which handled routine flying tasks, like airplane delivery, so more men could be fighter pilots), was instrumental in convincing Eisenhower to run for president, and served as a test pilot after the war. She was the first woman to break the sound barrier and holds more aviation height and distance records than anyone else who has ever lived. She also ran her own cosmetics company, charmed one of the ten richest men in the world, and helped kill the female astronaut program in the ’60s, probably “out of concern that she would no longer be the most prominent female aviator”. Fittingly, there’s a play about her unbelievable life.

Strangely touching recollections of working on Black Friday.

Tighter pollution laws can make cities much more livable – and be a boon to anyone with respiratory disease. But maybe they’re going too far when they threaten to destroy Montreal’s legendary bagel bakeries.

It’s almost a cliché to talk about how divided America is. This NYMag article takes the cliché to its (il-)logical conclusion and imagines an America divided between two interstate compacts (and a handful of neutrals). I don’t think it’s quite plausible, but there are many ways it seems better than the status quo?

I was surprised to find out that chicken pox is now considered a vaccine preventable disease and outbreaks of it are notable. I vaguely knew there was a vaccine now, but I didn’t realize that there’s now a whole generation of kids who never got chicken pox. Unfortunately, some parents dislike vaccines, so their kids are at risk of outbreaks (as well as shingles later in life).

Falsifiable

2019 Predictions

Posted on by Zach

Here are my 95 predictions about the next year. If you’d like to make you own predictions, you can download my blank template. Like previous years (2017, 2017 results, 2018, 2018 results), I’ve broken them down by region.

Canada

  1. Liberals win the next Canadian election – 60%
  2. Majority government in the next Canadian election – 60%
  3. NDP loses seats in the next Canadian election – 80%
  4. Conservatives gain seats in the next Canadian election – 70%
  5. Jagmeet Singh is not in parliament at the end of 2019 – 51%
  6. No court finds the assisted dying bill isn’t in line with Carter v Canada in 2019 – 80%
  7. No court rules on carbon tax constitutionality in 2019 – 90%
  8. No terrorist attack in Canada that kills > 10 Canadians in 2019 – 90%
  9. Opioid poisoning deaths decline from 2018 levels in Canada – 80%
  10. Construction on the Trans Mountain pipeline does not begin again in 2019 – 90%
  11. Canadian unemployment rate at or below 6% at the end of 2019 – 60%
  12. Health Canada still doesn’t allow Soylent to enter the country by the end of 2019 – 70%

America

  1. Less than 100km of concrete wall (or steel slat wall) on the border with Mexico will be constructed – 80%
  2. No registry of Muslims created – 90%
  3. No department of the Federal Government is eliminated – 90%
  4. Congress and Trump do not sign into law an extension of DACA – 60%
  5. Mueller’s investigation finishes in 2019 – 70%
  6. Impeachment proceedings aimed at Trump are not started in 2019 – 60%
  7.  More than a dozen credible (having held past elected office representing at least 100,000 people) candidates declared for the Democratic primary by the end of 2019 – 80%
  8. Less than two dozen credible candidates declared for the democratic primary by the end of 2019 – 70%
  9. Trump is still president at the end of 2019 – 90%
  10. No terrorist attack in America that kills > 10 Americans – 70%
  11. No terrorist attack in America that kills > 100 Americans – 90%
  12. Opioid poisoning deaths decline from 2018 levels in America – 70%
  13. NGDP growth below 5% for 2019 – 60%
  14. US unemployment rate at or below 4% at the end of 2019 – 60%
  15. Some form of North American free trade deal still in effect at the end of 2019 – 80%
  16. At least two metros beyond Phoenix see open, commercial self-driving car services by the end of 2019 – 51%
  17. More than one month of cumulative government shut downs in 2019 – 51%

South America

  1. FARC peace deal remains in place on January 1, 2020 – 60%
  2. Inflation in Venezuela is above 10,000% for the year of 2019 (as measured by DolarToday) – 80%
  3. Venezuelan GDP continues to contract in 2019 – 90%
  4. United Socialist party retains control of the Venezuelan presidency in 2019 – 80%
  5. Protests (and the official response to those protests) result in more than 100 fatalities in Venezuela in 2019 – 51%
  6. Major Venezuelan opposition groups do not enter any sort of power sharing agreement with the Venezuelan regime in 2019 – 90%
  7. No successful coup in Venezuela in 2019 – 90%
  8. Venezuelan oil production below 750,000 bpd by the end of 2019 – 60%
  9. Brazilian Real down against the US dollar at the end of 2019 – 51%
  10. Argentina meets its 17% inflation target by ±2% – 70%
  11. Argentina meets its IMF target of primary budget balance by the start of 2020 – 51%

Middle East

  1. No Israeli politician is indicted by the ICC over settlement activity in 2019 – 90%
  2. Netanyahu is no longer prime minister of Israel at the end of 2019 – 51%
  3. No Palestinian led Intifada in Israel that results in the deaths of >1000 combined attackers, security forces, and civilians (this is a conflict characterized by suicide bombing and police responses) – 90%
  4. No Israeli led operation in the West Bank or Gaza that results in the deaths of >1000 combined soldiers, civilians, and militants (this is a conflict characterized by rocket fire and military strikes) – 80%
  5. Fatah and Hamas do not meaningfully reconcile in 2019 (e.g. Fatah still doesn’t control Gaza by January 1, 2020) – 70%
  6. No significant resurgence in ISIL in 2019 (e.g. it does not gain territory over the next year) – 90%
  7. Fewer casualties in the Syrian Civil War in 2019 than in 2018 – 70%
  8. No power sharing agreement or durable ceasefire (typified by the three months following the agreement each having less than 500 fatalities) in Syria in 2019 – 80%
  9. Bashar Al Assad is still President of Syria on January 1, 2020 – 90%
  10. Protests in Iran do not result in more than 100 fatalities by the end of 2019 – 70%
  11. Hassan Rouhani is still President of Iran on January 1, 2020 – 90%
  12. No new international sanctions against Iran (does not include adding new organizations or individuals to old categories and requires coordinated participation of at least two countries) – 80%
  13. America does not rejoin the JCPA – 90%
  14. No attack on the Iranian nuclear program by Israel – 90%
  15. Iran does not withdraw from the deal limiting its nuclear program – 80%
  16. Conditional on Iran remaining in the nuclear deal, inspectors find no evidence of violations after the deal began – 90%
  17. Yemen Civil War continues – 90%
  18. The Hodeidah ceasefire in Yemen is maintained for 2019 – 51%
  19. America withdraws support for Saudi Arabia in the Yemeni civil war – 51%
  20. Mohammed bin Salman either remains as crown prince of Saudi Arabia, or becomes king (i.e. no coup or succession shake-up) – 80%
  21. No resolution or lifting of embargo in the Qatar crisis – 80%
  22. OPEC production cuts continue through to the end of 2019 – 70%

Africa

  1. South Sudan peace deal holds – 60%
  2. Libya still has two rival governments on January 1, 2020 – 60%
  3. No protests, riots, or rebellion in Egypt that kills >100 people in a one week period – 80%
  4. No protests, riots, or rebellion in Tunisian kills >50 people in a one week period – 90%
  5. No terrorist attack in Tunisia kills >20 people – 80%
  6. ANC wins South African elections – 80%
  7. Botswana’s Transparency International rating closest to January 1, 2020 shows improvement under president Masisi – 60%

Asia

  1.  Inflation rate in Japan remains below 2% in 2019 – 90%
  2. Japanese constitutional reform (removing pacifism) does not occur in 2019 – 80%
  3. At least two more countries ratify the CPTPP in 2019 – 51%
  4. All signatories have not ratified the CPTPP by the end of 2019 – 90%
  5. North Korean détente with Trump last through 2019 (typified by no new sanctions or missile tests) – 51%
  6. Xi Jinping visits North Korea in 2019 – 80%
  7. US-China trade war ends – 51%
  8. Modi wins elections in India – 70%
  9. No return to democracy in Thailand in 2019 – 90%
  10. Legal deadlock on same-sex marriage in Taiwan not resolved this year – 51%

Europe

  1. No resolution to the crisis in Ukraine – 80%
  2. Ukrainian elections are held – 70%
  3. Poroshenko no longer president of Ukraine by end of 2019 – 60%
  4. Russian GDP growth is less than 3% – 80%
  5. No gain of greater than 20% in the value of the ruble vs. the dollar – 90%
  6. Fall of more than 10% in the value of the ruble vs. the dollar – 51%
  7. Sanctions against Russia are not significantly rolled back (e.g. sanctions remain in place against Rosneft, Novate, Gazprombank and Vnesheconombank by all members of the G7 remain in place at the end of 2019) – 90%
  8. Teresa May remains prime minister of the United Kingdom – 70%
  9. Brexit does not occur on or by March 29, 2019 – 60%
  10. No snap election/vote of no-confidence in the UK in 2019 – 60%
  11. No resolution to the Julian Assange situation – 60%
  12. Poland’s EU voting rights aren’t suspended – 70%
  13. Euroskeptics make gains in 2019 EU elections – 80%
  14. The average of the last 5 opinion polls of Macron collected on Wikipedia have his net favourability below -30% – 80%
  15. Italy keeps its budget within the agreed upon EU boundaries (triggering no fines or other disciplinary action) – 80%
  16. DeepMind releases an algorithm that can beat top humans in StarCraft II – 51%
Falsifiable

Grading My 2018 Predictions

Posted on by Zach

In what is becoming a New Year’s tradition, let’s look at how accurately I predicted the future a year ago.

Canada

  1. Liberals remain ahead in the CBC Poll Tracker seat projection – 70%
  2. Trudeau has a higher net favorability rating than Andrew Scheer according to the CBC Leader Meter on January 1, 2019 – 80%
  3. Marijuana is legalized in time for Canada Day – 60%
  4. Marijuana is legalized in 2018 – 90%
  5. At least one court finds the assisted dying bill isn’t in line with Carter v Canada – 70%
  6. Ontario PC party wins the election – 60%
  7. The Ontario election results in a minority government – 80%
  8. The Quebec election results in a minority government – 80%
  9. No BC snap election in 2018 – 90%
  10. No terrorist attack in Canada that kills > 10 Canadians in 2018 – 90%
  11. More Canadian opioid poisoning deaths in 2018 than in 2017 – 60%
  12. Canada does better at the 2018 Winter Olympics (in both gold medals and total medals) than in 2014 – 90%
  13. Canada does not win a gold medal in men’s hockey at the 2018 Olympics – 70%
  14. Canada does win a gold medal in women’s hockey at the 2018 Olympics – 51%

America

  1. Trump announces that the US is pulling out of NAFTA and begins the process of putting the US withdrawal into motion – 51%
  2.  Less than 100km of concrete wall on the border with Mexico will be constructed – 90%
  3. No registry of Muslims created – 90%
  4. Congress doesn’t take action to extend DACA – 80%
  5. No department of the Federal Government is eliminated – 90%
  6. There isn’t a government shutdown before the midterm elections – 60%
  7. Democrats take back the house in the 2018 midterm elections – 80%
  8. Democrats take back the senate in the 2018 midterm elections – 60%
  9. Mueller’s investigation finishes in 2018 – 60%
  10. Impeachment proceedings aimed at Trump are not started in 2018 – 80%
  11. Trump is still president at the end of 2018 – 90%
  12. No terrorist attack in America that kills > 10 Americans – 70%
  13. No terrorist attack in America that kills > 100 Americans – 90%
  14. Susan Collins doesn’t get the Obamacare stabilization measures she was promised – 70%
  15. More US opioid poisoning deaths in 2018 than in 2017 – 80%

South America

  1. FARC peace deal remains in place on January 1, 2019 – 80%
  2. The black market exchange rate for Venezuelan Bolivars is above 110,000 to the US dollar on January 1, 2019 (as measured by DolarToday) – 80%
  3. Inflation in Venezuela is above 100% for the year of 2018 (as measured by DolarToday) – 90%
  4. United Socialist party retains control of the Venezuelan presidency in 2018 – 90%
  5. Protests (and the official response to those protests) result in more than 100 fatalities in Venezuela in 2018 – 60%
  6. Protests (and the official response to those protests) do not result in more than 1000 fatalities in Venezuela in 2018 – 70%
  7. Major Venezuelan opposition groups do not enter any sort of power sharing agreement with the Venezuelan regime in 2018 – 80%

Middle East

  1. No Israeli politician is indicted by the ICC over settlement activity in 2018 – 90%
  2. There isn’t an election in Israel in 2018 – 80%
  3. US does not physically relocate its embassy to Jerusalem in 2018 – 90%
  4. No Palestinian led Intifada in Israel that results in the deaths of >1000 combined attackers, security forces, and civilians (this is a conflict characterized by suicide bombing and police responses) – 70%
  5. No Israeli led operation in the West Bank or Gaza that results in the deaths of >1000 combined soldiers, civilians, and militants (this is a conflict characterized by rocket fire and military strikes) – 70%
  6. Fatah and Hamas do not meaningfully reconcile in 2018 (e.g. Fatah still doesn’t control Gaza by January 1, 2019) – 51%
  7. No significant resurgence in ISIL in 2018 (e.g. it does not gain territory over the next year) – 80%
  8. Fewer casualties in the Syrian Civil War in 2018 than in 2017 – 70%
  9. No power sharing agreement or durable ceasefire (typified by the three months following the agreement each having less than 500 fatalities) in Syria in 2018 – 80%
  10. Bashar Al Assad is still President of Syria on January 1, 2019 – 90%
  11. Protests in Iran do not result in more than 1000 fatalities by the end of 2018 – 70%
  12. Protests in Iran do not result in more than 100 fatalities by the end of 2018 – 51%
  13. Hassan Rouhani is still President of Iran on January 1, 2019 – 90%
  14. No new international sanctions against Iran (does not include adding new organizations or individuals to old categories and requires coordinated participation of at least two countries) – 80%
  15. No new US sanctions against Iran (does not include adding new organizations or individuals to old categories) – 51%
  16. No attack on the Iranian nuclear program by Israel – 90%
  17. Iran does not withdraw from the deal limiting its nuclear program – 90%
  18. Conditional on Iran remaining in the nuclear deal, inspectors find no evidence of violations after the deal began – 90%
  19. Yemen Civil War continues – 60%
  20. Saudi Arabia pulls troops out of Yemen – 51%
  21. Mohammed bin Salman either remains as crown prince of Saudi Arabia, or becomes king (i.e. no coup or succession shake-up) – 80%
  22. Rockets fired from Yemen cause casualties in another country – 51%
  23. No resolution or lifting of embargo in the Qatar crisis – 80%
  24. OPEC production cuts continue through to the end of 2018 – 60%

Africa

  1. No power sharing between ZANU-PF and the opposition will happen in Zimbabwe before the elections (if they occur) in 2018 – 80%
  2. Zimbabwe will hold election in 2018 – 70%
  3. No peace deal ends South Sudan fighting – 70%
  4. Libya still has two rival governments on January 1, 2019 – 70%
  5. No protests, riots, or rebellion in Egypt that kills >100 people in a one week period – 80%
  6. No protests, riots, or rebellion in Tunisian kills >50 people in a one week period – 90%
  7. No terrorist attack in Tunisia kills >20 people – 80%
  8. Zuma is not impeached in 2018 – 51%

Asia

  1.  Inflation rate in Japan still remains below 1% in 2018 – 70%
  2. Japanese constitutional reform (removing pacifism) does not occur in 2018 – 51%
  3. China will not deploy its military against Taiwan or Hong Kong in 2018 – 90%
  4. North Korea will test a submarine launched ballistic missile in 2018 – 70%
  5. North Korea will not test nuclear weapons or launch any missiles during the 2018 Olympics – 80%
  6. North Korea will test a nuclear weapon in 2018 – 51%
  7. No country will attempt to shoot down a North Korean missile test in 2018 – 80%
  8. If there is an attempt, it will succeed – 51%
  9. North Korea tests a missile that is judged by experts at 38 North as likely able to carry a plausible North Korean nuclear weapon to the United States – 60%
  10. No current member of China’s Politburo Standing Committee visits North Korea in 2018 – 70%
  11. No meeting between Kim Jung-un and Moon Jae-in in 2018 – 90%

Europe

  1. No resolution to the crisis in Ukraine – 80%
  2. Russian GDP growth is less than 3% – 80%
  3. No gain of greater than 20% in the value of the ruble vs. the dollar – 70%
  4. Sanctions against Russia are not significantly rolled back (e.g. sanctions remain in place against Rosneft, Novate, Gazprombank and Vnesheconombank by all members of the G7 remain in place at the end of 2018) – 90%
  5. Angela Merkel remains chancellor of Germany – 60%
  6. Germany holds another election before a government can be formed – 51%
  7. No date set for another Scottish referendum in 2018 – 80%
  8. Teresa May remains prime minister of the United Kingdom – 70%
  9. The UK does not terminate the process of Brexit in 2018 – 90%
  10. There is no final Brexit withdrawal deal reached in 2019 (Donald Tusk wishes to have one by October) – 51%
  11. No snap election/vote of no-confidence in the UK in 2018 – 80%
  12. Poland’s EU voting rights aren’t suspended – 90%
  13. Poland and Hungary continue to refuse to accept migrant quotas – 90%

Notes

  • I originally worried that I’d missed the mark with my Venezuelan devaluation predictions (South America #2), as DolarToday gave an exchange rate of 730.29 BsS. to the dollar. But it turns out the BsS. is a new currency, with each one equivalent to 100,000 of the old bolivars. This means the current exchange rate is 73,029,000 old bolivars to the USD. Inflation was 4305%. It is now an unfortunate fact that some online video game currencies are worth more than the old Venezuelan bolivar.
  • For Middle East #3, I got conflicting information from Wikipedia as to whether the embassy was officially moved, or just moved in name pending the ambassador finding a place to live in Jerusalem. I decided that it should count as moved for all practical purposes.
  • With respect to Middle East #15, I believe that the US withdrawing from the P5+1 deal and restarting sanctions should probably count as new sanctions.
  • I think Africa #8 is technically correct, because Zuma wasn’t impeached, but forced out by his party. Still, I think I will aim to make broader predictions about embattled leaders leaving in the future.
  • I calculate Japan’s inflation rate at around 1.05% based off the most recent statistics I could find, although this may need to be adjusted once December’s numbers are out and further adjusted once we have the year-end numbers. I got Asia #1 wrong, but not by much!
  • I was worried that no member of the standing committee had visited North Korea, potentially making Asia #10 a technical success, despite it obviously being falsified in spirit (by the three meetings in China by Xi Jinping and Kim Jong-un). Luckily, Li Zhanshu, a Standing Committee member, visited to commemorate the founding of North Korea.
  • I’m a bit unsure how I should treat Europe #10; the UK negotiated a Brexit deal, but no one actually believes in can pass parliament. I’m going to tentatively mark this as correct, but will revise if parliament passes this deal unaltered.
  • EDITED TO ADD: I originally didn’t mark America #6 as failed because I totally forgot about the one day government shutdown in January. I’ve since updated everything to add this in.

Calibration

The whole point of having predictions with binned probabilities (here, 51%, 60%, 70%, 80%, 90%) is that you can then check your accuracy; an individual prediction can be right or wrong, but when you pool several predictions you were equally sure of, you can tell if you were right or wrong to be that sure. Here’s how I did:

  • Of my predictions at a 51%confidence level, I got 6 right and 6 wrong (50%).
  • Of my predictions at a 60%confidence level, I got 4 right and 7 wrong (36%).
  • Of my predictions at a 70%confidence level, I got 13 right and 5 wrong (72%).
  • Of my predictions at an 80%confidence level, I got 19 right and 5 wrong (79%).
  • Of my predictions at a 90%confidence level, I got 23 right and 2 wrong (92%).
  • Two predictions were invalid, because they were predicated on a North Korean missile test that never occurred.

Some people prefer to look at this graphically. Here I’ve plotted my answers versus what you’d expect if I was perfectly calibrated (something I say is 70% likely to happen happens 70% of the time). Whenever the blue line labelled “Mine” is below the red line labelled “Perfect”, I was overconfident. Whenever it’s above the red line, I was under-confident.

Like last year, I’m pleased with these results. Most of my probability bins were very well calibrated, although I seem to have become rather overconfident at the 60% level. I was under-confident at that level last year by a significant amount, so this might just be normal variance. I’ll keep an eye on this for next year.

In terms of specific classes of failed predictions, I can identify two.

First, I tend to overestimate how quickly bureaucratic processes will finish (see: 3, 5, 23). Second, I tend to assume the future will look much like the present (see: 11, 29, 72, 74-79). This might be because of the availability heuristic; the present is more available than alternative futures, so I might be viewing it as more likely than I should.

I hope to account for these biases in my soon to be released predictions for 2019. Stay tuned!