A public database of workplace injuries

The U.S. Department of Labor has plans to create a publicly available database of workplace safety incidents for large corporations.  The idea is that adding this kind of public visibility for injuries and illnesses will encourage companies to develop more effective safety programs and will raise the public visibility for the 3 million workers a year who are hurt or sickened.  If companies do a really good job, they can use this as a benefit by promoting their safety record in a more publicly credible way. 

Companies have legitimate concerns because data is often misused by advocates and by people who don’t understand statistical validity and reliability.  They might compare the safety records of companies in different industries, with different workforces, and come to erroneous conclusions. There are a slew of information processing heuristics that I have blogged about at length that would lead company advocates to underappreciate the problems and safety advocates to overappreciate the problems.

The trick then comes down to how the database user interface is designed.  With data from 500,000 companies available, accurate and fair comparisons will be available in most cases.  The interface should be designed to highlight these rather than to allow full flexibility to surf and mine the data to get whatever the user wants.  This is different from the approach taken for many database projects, but may be a better way to go in this case, both politically and in recognition of human nature. 

One way of achieving this would be to use a derived industry-specific metric.  Weighted and adjusted rates are common in public health, economics, and many other domains.  For example, a company with a 1.3 for injuries could indicate that they have 30% more injuries or 0.3 standard deviations more or whatever, compared to a carefully matched set of other companies.  Once the political minefield of developing the weighting and adjusting statistics has been waded through, it might be reasonable to presume that the rates will be used accurately or if not could easily be disputed. 

Of course, organizations will still try to game the system.  That is also human nature.  There will be underreporting incentives targeted directly at employees.  There will be coding schemes that minimize the apparent severity of injuries that are reported.  But if this is done across the board, it could be that the weighting and adjusting process will cancel these all out and we can compare companies within an industry, industries across the economy, and individual companies over time.  This benefit is too much to give up on. 

How oversimplified, media-created memes become accepted facts.

The problems in the Middle East are not because of colonial border-drawing: How oversimplified, media-created memes become accepted facts.

This is one of my societal pet peeves.  It is a practical reality that most of us get our knowledge of the world from the media.   We were not all history majors in college.  We don’t read the original scholarly analyses that research world history.  And we have enough going on in our lives to become experts in various problems around the world, even the important ones.  So when a variety of news media all say the same thing, we tend to believe it.  This makes total sense and is a good thing as long as the media are responsible in wielding this great power. 

But they are not.

For my entire life, I have heard over and over again that one of the leading causes of sectarian violence in the Middle East is because the colonial powers drew these nice straight lines that cut through ethnic regions, putting some Kurds in Turkey, some in Syria, and some in Iraq.  Iraq was part Kurd, part Sunni, and part Shia.  If they had just drawn the borders to put each group into its own country, much of the violence over the past half-century could have been avoided.

But it turns out, this is such as oversimplification as to be virtually false.  Why are the Alawites (of Bashar al Assad) in power in Syria despite being from the minority?  It was a very smart way for the French colonialists to maintain power during the time they wanted to remain colonialists.  If the group you prop up is the smallest, they have a great motivation to keep you as colonial overlords.  If not, they lose all of that great power.  But because they are small, they need to use some harsh and often violent methods to keep order.  This sets the stage for sectarian violence once the colonialists leave.  France leaves the Alawites in power and the majority Sunnis are ready to take back power and in many cases exact revenge for genuine atrocities.  It is not because of the borders.  The same thing happened in Lebanon and elsewhere.  The Soviet Union did this when creating their Central Asian republics.  Africa too.

Second, the borders that the colonial powers drew were not completely arbitrary after all.  Many of them followed borders that had existed for centuries – through the Ottoman Empire and in some cases centuries early.  They didn’t just draw straight lines because they wanted to save ink on the maps.  Another argument is that they should have let the people who lived there set their own borders.  That is what they tried in the Balkans and we know how that turned out. 

Recently, the Atlantic had a good article on this.  I heard a story on NPR recently that made some of the same points.  These are what opened my eyes.  But since most of us get our news from Jon Stewart (who got it wrong), and the broadcast nightly news (who frequently get it wrong), most of us have it wrong too. 

Hacking an election through Twitter

You have probably seen with the new kind of ticker that the cable news networks have become fond of showing on the bottom of the screen.  It used to be news headlines.  Now it is their live Twitter feed.  I wonder what this says about the famous firewall between the content and the business sides of the news media.  The headlines had real value.  The Twitter feeds get better ratings.  They stimulate the social impulses that we all have in the depths of our primitive brains.  On the other hand, the news headlines feed the modern analytical, knowledge-seeking brain – which is quick to fatigue and get bored. Knowledge is hard.  Gossip is juicy. 

This has not only taken over the ticker.  Many news shows now have a segment dedicated to reporting on what is trending on Twitter.  Twitter trends, instead of just quantifying what is being talked about, have become the subject of the discussion itself.

So why am I bringing this up today?  I heard about a very subtle but insidious trick that has been used to rig elections in places such as Russia and Mexico.  These could easily be used here as well. 

Let’s say we are in the run-up to an important election.  At some point, the media is bound to report on which candidate is trending higher on Twitter.  There was even a recent study that found Twitter volume was a better predictor of the winner of elections than the official polls are.  In the cases that were uncovered (and probably many that went uncovered), Twitter bots were programmed to post random (but human-sounding) tweets about one candidate.  This is not too hard.  You could but random constructions together in the form of “I liked what Joe said about taxes/education/guns/healthcare/etc. today/last night/yesterday/etc.”   All kinds of combinations works and when you only have 140 characters you don’t expect much detail anyway.  Thousands of these get sent out from hundreds of fake accounts.  At first, just the people following the candidate’s hashtag see them.   But when the media picks up on it, it becomes a self-fulfilling prophecy.  The media tells the world (or at least the voting region) that Joe is skyrocketing on Twitter.  The subconscious desire to find agreement with our social group makes us all give Joe a second look.  “Hey, maybe he is better than I thought !!”  This isn’t going to get a crazy radical elected, but it could certainly change the election by 5-10%, which is the margin of most contested elections. 

The same tactic has been used to overload a Congressional office.  If your local Representative gets 10,000 tweets/emails/Facebook posts all in favor of one policy or another, he/she is going to think twice about voting against it.  They all seem to be coming from different accounts, even if they are all bots.  If he/she wants to get re-elected, well, you know the rest.  

This doesn’t just bother me in the way that user experience problems do. This scares me.  We already have enough losers in government.  This could make it much worse. 

Religious parties in a democracy

Ok, I am revealing a little bit of my inner geek here, but the way I learn about the world is not Time Magazine or the Washington Post like most people, but rather by reading the geopolitical intelligence reports from Stratfor (the same place many US Presidents turn) and to research journals like Democratization.  These sources tend to be a lot more comprehensive, objective, penetrating.  But a lot harder to read because they are long and use lots of scholarly language. 

The latest of Democratization was a special issue that brought together a series of papers from some of the leading experts around the world on religious political parties.  The researchers are all from different places and so they have very different ideas (hence the more objective perspective you get when you read them all). 

The impetus for the collection was the Arab Spring and how the different Muslim Brotherhood parties are handling their newfound power in Egypt, Tunisia, Syria, etc.  For example, why did the Muslim Brotherhood in Tunisia put together a broad, pluralistic coalition while the Egyptian Brotherhood decided to become the autocrat they replaced? Is it because of the way the party had been treated over the last few decades by the former leadership/society?  Or because of the values of the society itself?  Or socio-economic differences?  Or was it due to differences in the parties?  The extremity of their religious views?  Their conservative/progressive political stance independent of their religion?  How connected they are to the military? If they were a minority part of a governing coalition in the past (and got a taste of governance)?  If they have a nationalistic ideology? 

It is really hard to do valid research in this area because there are no control groups.  Do these studies used a case study method and looked all over the world (I was shocked by how many religious parties there are around the world), including India, Northern Ireland, Italy, Chile, Turkey, Israel as well as the Arab Spring examples.  It is really the questions that they raised that I found interesting:

·         What would you think of a party that has extreme religious views but believes in pluralistic governance (compared to the moderately religious party that wants total control)?

·         What about a party with a great deal of support from the general population but dominates the military and favors a militaristic foreign policy?

·         What about a party that wants to force the whole country to follow its religious laws and practices, but those laws and practices happen to be very open and tolerant (allows same sex marriage, feeds the hungry, etc)?

More questions than answers, but I found this to be really good food for thought.

The Zimmerman verdict is about the threshold of evidence, not racism in America

I was going to hold back because everyone seems to be so emotional about the Zimmerman verdict.  But I have to present this, because it is a completely different perspective on this case. 

This is what struck me from this whole fiasco.  Both groups, the passionate individuals demonstrating in support of Treyvon Martin and who were so sure that George Zimmerman should have been convicted of Murder One and the equally passionate individuals demonstrating that Zimmerman was the victim of overzealous prosecution, were making the SAME BIG MISTAKE.

We live in a legal system based on a standard of reasonable doubt based on the evidence presented in court.  How many of us were in that court?  Were you?  I wasn't.  Neither were any of the people demonstrating - from either side. 

What does this tell us?  It doesn't tell us if the verdict was correct or not.  But it does tell us that NO ONE - Not the people on one side or the people on the other side - NO ONE can say that the verdict should have gone one way or the other.  Not you and not me.  Because we didn't see the evidence that was presented in court.  And so we can't say if that evidence meets the standard of convincing a jury of 12 U.S. citizens beyond a reasonable doubt that he was guilty of a strict legal definition of murder or manslaughter as explained to them in court. 

We can read the judge's instructions to the jury, and we can even read transcripts of the testimony at some point.  But we can't put ourselves in the place of the jury who were required to look into the eyes of the few witnesses and determine whether they were convincing or not.  Beyond a reasonable doubt.

My heart tells me the George Zimmerman went way over what he should have.  He was a neighborhood watch member who was pretending to be a cop.  He was looking for trouble.  He should not have been there in the first place.  He should not have gone after Treyvon.  And he definitely should not have shot his weapon.  This is a great example of why we need stronger gun laws.  But I also think that there are a gazillion scenarios in which the testimony would not reach the standard of reasonable doubt, even if he really was guilty of murder in the absolute sense.  I can also imagine scenarios in which it would have.  But since I was not in the courtroom, I don't know either way.  And neither do you.  And neither do any of the people demonstrating.

So I think it is great if they are demonstrating for something productive.  Perhaps they could demonstrate for programs to get better race relations.  Or to mandate better training for neighborhood watch volunteers.  Or for more volunteering in one's community.  But death threats to Robert Zimmerman (George's brother)?  Or even the people who are so sure that Zimmerman should have been found guilty and the verdict is just another example of institutionalized racism in the US.   This is ridiculous.

Law enforcement and fishing expeditions

This is a big overreach by US law enforcement. Collecting an arrestee's DNA to confirm his or her identity makes sense. So does comparing it to a limited set of cold cases if there is evidence linking the arrestee to those crimes. But comparing it to a set of all DNA from all cold cases is a violation of the 4th Amendment if I ever saw one. Isn't that the point? Isn't that why we have a pejorative term called "fishing expedition"? What makes this scary is that you might look similar to a suspect, get picked up for questioning even when you have done nothing, and then your DNA gets taken, submitted to the database, and then linked to a crime from 20 years ago. And you might not even have done THAT crime either, your DNA might have coicidentally been in the room where the crime happened. A bank ATM where a hold up occurred, a hotel room where a rape occurred. And then you get dragged through the mud again - Your name tarnished for nothing. Your reputation possibly ruined forever because people will always wonder. This is just WRONG!!!

http://www.washingtonpost.com/politics/supreme-court-upholds-maryland-law-says-police-may-take-dna-samples-from-arrestees/2013/06/03/0b619ade-cc5a-11e2-8845-d970ccb04497_story.html

Corporate Research transparency

I have a new found respect for Mars, Inc (the chocolate company).  I was listening to an interview with Hagen Schroeter, the director of fundamental health and nutrition research, at Mars.  About 20 minutes of the interview covered the now famous anti-oxidants in chocolate.  This has led to a major increase in sales of chocolate.

Hagen's research discovered that this is BS (sorry to all you chocolate fans that have been using this as an excuse to eat more).  The processing of chocolate destroys most of the flavonoids and  flavonols.  And the digestion and metabolism of the chocolate destroys the rest.  So test tube studies find anti-oxidants, but in vitro blood tests find zippo make it through.

Some of Hagen's research focuses on new ways to process the chocolate so more of it gets through.

What makes me respect Mars is that the interview asked Hagen if Mars discourages him from publishing the results.  After all, this could really kill all of those extra sales they are getting from the rationalizations of chocolate loving dieters. 

But his response was just the opposite.  He said that Mars encourages him to publish it.  One self-serving (for Mars) hope is that independent scientists will read the research and do their own studies on new processing methods that maintain anti-oxidant levels. 

But he insisted that it is more of a fundamental openness of their research policy.  I can't attest as to whether this is true or not, but the fact that he was disclosing in a publicly available interview that the anti-oxidant meme is BS was some pretty good evidence.

Liability and Generic Drug Makers

The Supreme Court is hearing a case that has be really fascinated. 

As you might know, generic drug manufacturers are not allowed to make any changes to the drug composition or labeling of the brand names they are genericizing. 

So what happens (as in this case) if a patient gets severely hurt by the side effects of a generic drug? 

1.  Can they sue the generic company for damages?  The generic company sold them the drug, but was not allowed by law to do anything about it.

2.  Can they sue the original pharma company that designed, tested, and brought the drug through Phase I, II, and III trials?  It was their mistake, but the patient has no connection with this company so where is the standing?

3.  Can they sue the FDA?  The FDA should have made sure that the Phase I, II, and III testing was adequate to catch dangerous side effects.  But of course this is not feasible economically or scientifically. There are too many variations in patient behavior and genes, remote side effects that don't show up, etc.  The drug approval process would be paralyzed.  And the federal government has safe harbor from law suits anyway.

4.  Maybe the patient sues the generic and the original pharma.  The combination is sort of the entity that is responsible. 

5.  Maybe they just can't sue anyone.  If a company gets a drug through FDA trials and approval, perhaps they should be protected from lawsuits.

Then we can imagine some real obfuscation that should add some corollaries to the law.  What if the original pharma company knew about the side effect but suppressed it in their drug application?  What if the generic company knew about the side effects and the fact that they were suppressed, but couldn't do anything about it because of the law handcuffing generics from making changes?  What if these side effects became known after the FDA had granted approval? 

The Supreme Court only has to consider one case this time.  But another one of these variations could come up at any time.  I hope they are ready.

International Governance by Proxy

Despite the officious title, this is a really important issue.  Here is one of the major challenges we face to get anything done that requires international agreement and/or participation.

There are no sovereignty rights for any international organizations.  So unless every single country agrees, the agreement stalls.  There are some organizations that can pass policies with less than 100% of the vote, but the countries that vote no can just opt-out.  That is what the US does with many of the UN initiatives they don't agree with.

The problem is that when you have 150+ countries involved in the negotiation, there will never be an agreement.  There are too many divergent issues and priorities.  But if you exclude anyone, you lose the value of unanimous consensus, the validity of the agreement, and in practice the compliance of countries that don't agree.

So my thought is to leverage the brilliance of our Founding Fathers and use a federalist representative model.  Here is what I am thinking:

1.  Each country votes in a secret ballot open ended format for who they want to represent them in the negotiations.  They are free to vote for themselves.

2.  The top five vote-getters form the negotiating committee.  So the risk of voting for yourself is that unless other countries respect your position, your vote is "wasted."  In general, larger and more influential countries will get selected.  This is what we want if we hope to get consensus and compliance at the end of the process. 

3.  There is another worldwide vote for weighting. Every country votes for one of the five committee member countries.  The committee negotiates equally, but then when any official votes are taken they are weighted according to this score.  It can be some hybrid of GDP, population, or whatever summed for all of the countries that voted for them.  For example, if many countries respect Brazil's position on environmental policy, then Brazil will receive many votes (in both rounds) and have a large influence in the final international agreement of the environmental committee.  But these same countries can vote for a different country when setting up the international trade committee, the human trafficking committee, or whatever.  Moderate and reasonable positions are likely to get on the committee.  Extremists won't, but will feel included because they got to vote both for committee membership and for committee weighting.

4.   These five countries then negotiate a policy.  I used five just because it is enough to represent a variety of positions but not unwieldy. If four or six is better, then so be it.

5.  The final negotiated policy becomes the international treaty on the environment or on trade or on whatever.  Everyone has previously committed to abide by the result and has participated in the process, so there should be a decent amount of buy-in.  Admittedly, North Korea and some other real outliers may never agree, but I don't think any system could handle them.

Because every country gives their proxy to one of the final committee members in the second round, their vote is counted in the final treaty vote.  Just not separately.

And because a new committee is formed for each domain (environment, trade . . . ), the major positions will always be represented, extreme positions will be muted, and countries will have a strong incentive to build an international reputation for developing smart positions because this was they get to form the final policy language in domains they care about.