Forensic Delusions

By Robin Hanson · March 20, 2010 6:00 am · Comments (0)

A groundbreaking report by the National Academy of Sciences (NAS) last year recognized that much of forensic science is not rooted in solid science. Many forensic disciplines — such as hair microscopy, bite-mark comparisons, fingerprint analysis, firearm testing and tool-mark analysis — were developed solely to solve crimes.  They evolved mainly in the context of individual cases, which often had significant variation in resources and expertise. They have not been subjected to rigorous experimental scrutiny, and there are no standards or oversight in the United States or elsewhere to ensure that validated, reliable forensic methods are used consistently. With the exception of DNA analysis, no forensic method has been proved to reliably and accurately demonstrate a connection between evidence and a specific source. …

We advocate the creation of an office of forensic science improvement and support (OFSIS) within the US Department of Commerce to spur independent research, develop standards and ensure compliance. … [Some] within the forensic-science and law-enforcement fields … argue that an OFSIS is not necessary and that laboratory accreditation is sufficient. … They argue that an OFSIS would cost too much … and that it could create chaos in the US justice system by reopening countless old cases. … Political and criminal-justice ends — rather than research imperatives — have taken forensic science off course. … See www.just-science.org.

More here.  The primary social pressure on law court practices is for courts to give the appearance of punishing guilty folks.  Observers have much less info on who is actually guilty.  So the main pressure on legal standards is that officially-accepted evidence seem to the public, juries, and judges to indicate guilt, not that it actually indicate guilt.  We expect the law to be overconfident about its evidence.

Requiring that legal evidence standards stand up to independent experimental scrutiny would create a more accurate legal system, but at the expense of reducing the apparent rate at which the guilty are punished.  So I expect, sadly, this proposal to be rejected.

Tagged as: ,

Future Discounts

By Robin Hanson · March 19, 2010 6:00 am · Comments (16)

In few billion years our descendants may spread across billions of galaxies. Even so, if they do not drastically change the structure of space-time, then within a trillion years they will fragment into billions of isolated galaxy-sized “universes”.  Standard physics, you see, says that in a trillion or two years all the galaxies near the Milky Way will merge into one big galaxy, and other galaxies will be too distant to see in any way.  For all practical purposes, that merged galaxy will be a separate universe.

If we do nothing to change the situation, then within ten or so trillion years, all current stars will be dead (degenerate), and no more stars will form.  Over the next billion trillion years, stars will occasionally smash in a flash, or pass close enough to each other to throw one out of the galaxy; in the end 1-10% remain in a central black hole.

What if we change the situation?  Most useful resources, such as hydrogen to turn into lead, or mass not yet dropped into the central black hole, will likely be identified and claimed within a few million years.  How fast will folks use up these resources?

In principle, most everything might be burned quickly in a few million years of party-hardy gluttony, or most might be saved to use steadily over the billion trillion trillion years or more before protons decay.  How fast resources are actually used would be determined by the discount factors of the creatures who control resources.  But what would those be?

If unused resources were completely stable and if property rights in resources were completely secure, then we’d mainly have a selection effect in discount rates.  Agents who discount fast would dominate early activity, while those who discount slowly would dominate late activity.  Even if initially only a tiny fraction of agents cared about activity in a billion trillion trillion years, those agents would dominate such late activity.

Any natural rate at which resources decay would set an upper limit on discounting.  There is no point in planning to use resources long after you expect them to decay.  Similarly, insecure property rights would increase discount rates. If you expect a 1% chance that your property will be stolen every million years, you won’t expect to still have much after a billion years, so you might as well plan to use most of it before then.   The same holds if your property is never stolen, but you have to spend 1% of your resources every million years to ensure that fact.

“Switzerlands,” from which theft is naturally harder, might be the last locations of activity in each galaxy.  These might be matter sent on very long secret orbits, to return back to galaxy central after a very long time.  Similarly, resources which simply could not be physically used until a long delay might ensure some late universe activity.

The inhabitants of a galaxy-universe could have different degrees of central coordination; some might have a strong central government, while others lived in anarchy.  With a strong central government, long term activity seems strongly influenced by the discount rate of that government. If this government taxed 1% of resources every million years, and didn’t invest those resources for the long run, then there would be little point in planning to use your resources after a billion years.  No obvious selection effect ensures that galaxy governments take a long view.

Physics may set the ultimate limits on how long resources, and life, can last, but governments and property rights will determine when they are actually used. Resources, and life, will likely die long before their physical expiration dates.

Tagged as: , ,

Twin Conspiracies

By Robin Hanson · March 18, 2010 1:50 pm · Comments (29)

In a twin conspiracy, a pair of identical twins would pretend to be only one person. For example, in college each twin could specialize in, and then ace, half of the classes; their GPA would soar.  They might together make partner in a law firm by handling a lot more work than other lawyers.  They could cheat on their spouse and while offering that spouse a near-constant video of “their” activities.  In fact, they could always have an alibi for anything they did.

This strategy seems tempting in “winner take all” areas of life where small productive gains are given huge rewards, or where secretly having more time can make you seem a lot more productive.  For example, high level managers attend a great many meetings to connect different parts of their organization.  A secretly-twin-CEO could attend twice as many meetings, and make twice the connections.

Of course if this actually happened often our institutions could easily adapt to check for secret twin conspiracies.  They don’t now look because they don’t expect them.  It would be interesting to search for such secret twins.  For example, one might take a list of top CEOs and compare the ratio of non-identical to identical twins in this group. If that ratio was substantially larger than in the larger population, that might suggests many secret twins hiding among CEOs.

One twin told me the loss of autonomy in this secret twin scenario would make it unacceptable to her, no matter what worldly success it produced.   Do people really care that much more about autonomy than success?

Tagged as:

Stop Stale Eggs, Jobs?

By Robin Hanson · March 17, 2010 9:15 am · Comments (37)

Some men see things as they are and ask why. Others dream things that never were and ask why not. Shaw

The average woman is born with around 300,000 eggs … 12 percent of those eggs remaining at the age of 30, and only 3 percent left by 40. … From the mid-30s on, the decline in fertility is much steeper with each passing year. … Female undergraduates significantly overestimated their fertility prospects at all ages. … The biological reality that female fertility peaks in the teens and early 20s can be difficult for many American women to swallow, as they delay childbirth further every year. … The older you get, the more difficult it is to get pregnant and the higher the chance of miscarriage, pregnancy problems such as gestational diabetes and hypertension, and chromosomal abnormalities such as Down syndrome. … The risk of autism increases with a mother’s age.

More here.  Also, Andrew Leigh:

We estimate the relationship between maternal age and child … learning outcomes and social outcomes. … Children of older mothers have better outcomes. … When we control for other socioeconomic characteristics, such as family income, parental education and single parenthood, the coefficients on maternal age become small and statistically insignificant.

Today high status women stay long in school, start careers, and take long to match up with a man before having kids.  They are often too late, their kids have more defects, and the interruption hurts their career.  Low status women more often have an accidental early kid out of wedlock.

Imagine a different equilibrium, where females pick a male at 15, then school more slowly to have kids till some standard age (20? 25? 30?), when females return to full-time school and uninterrupted careers.

While it is not entirely clear if this new equilibrium would be better or worse, it certainly has some positive features.  Kids and moms would be healthier, kids more numerous and less accidental, moms more energetic, older folk would enjoy more grand kids etc., and career interruptions wouldn’t make female employees suspect.

Early parenting would have to be paid for by grandparents or via loans (or perhaps income shares), presumably in trade for some loss of autonomy.  While childhood does seem to be lengthening, it is not clear if this autonomy loss could be accepted.

For the male pattern, there are two obvious variations: males switch life-plans along with females, or males stay on the current plan.  Having males also switch would keep mates at similar ages, promote healthier kids and more energetic dads, and reduce opportunities for gender discrimination.

Randomness in kid timing and number would make it a bit harder to estimate student quality based on student performance – could we find ways to correct for this?  And the fact that low status moms now have kids early makes it harder to coordinate a switch to this new equilibrium.  But still, it seems an interesting thing that never was, about which to ask: why not?

From a conversation with Rob Wiblin, Katja Grace.

Tagged as: , ,

Efficiency Disclaimers

By Robin Hanson · March 16, 2010 2:00 pm · Comments (4)

Though I generally avoid disclaimers, since Bryan Caplan calls my latest claim that econ efficiency is a good tool for finding win-win deals “compete nonsense,” let me try to clarify:

  • We have many purposes when we talk about “what to do”, and making deals is only one of our purposes.
  • Getting what we want is only one of many reasons we try to achieve deals; we also want to signal our features, for example.
  • Analysis aids are only one of many sorts of aids that can help us to make deals; aids can also to organize negotiations, enforce deals, etc.
  • Most useful deal aids are relatively specific to a particular context, such as real estate sales, or marriages; when available, more specific tools tend to be more useful.
  • Deal aids can specialize in what groups that they best assist.  A particular aid might be best suited for couples, club, firms, or nations.  Wider aids specialize in assisting larger groups.

Economic efficiency is our best wide general analysis tool for finding win-win deals that get people what they want.  That isn’t everything, but it is a lot.  I’m glad I mastered this tool and am eager to apply it.  Efficiency can:

  • Suggest Deals – Efficiency analysis suggests policies to make “the biggest pie.” A deal also needs folks to agree on a way to divide the pie, such as via cash transfers between the parties.  Even so, knowing better ways to make bigger pies should make parties more eager to agree to deals to lock in such gains.
  • Be Part Of A Deal -  Groups can make explicit deals to adopt the results of efficiency analysis.  For example, legal systems can adopt a general accident rule that puts responsibility on the least cost accident avoider, and government agencies can be instructed to make policy decisions on a cost-benefit basis.  Most can reasonably expect to gain from such policies, even if they do not expect to gain from each particular application.

Admittedly:

  • Efficiency analysis is a rough guide, and does not determine exact implications with certainty for each possible situation.
  • The policies efficiency recommends depend on particular modeling assumptions and parameter estimates, for example, and those depend on particular analysts and sources used.
  • Even when negotiators have access to solid analyses, deals can fail for many other reasons; good analysis doesn’t ensure good deals.
  • Most deals are not between all possible parties, and each deal may well disadvantage those not included in the deal.
  • People may expect to gain from a deal, but end up not actually benefiting.
  • As a wide general tool, efficiency is less useful for small deals or for contexts where specialized tools are available.
  • Efficient deals may well be immoral, or unattractive for other purposes of deal-making, or of “what to do” talking.

Few deals can guarantee to get everyone more of what they want, but by encouraging and enabling more better wider deals, the use of efficiency analysis sure seems to me to tend to get most everyone more of what they want.  Isn’t that good enough?

OK, now that I’ve tried this exercise of explicitly listing many possible disclaimers, when is this sort of exercise actually worth the effort?

Tagged as: ,

Uninsured ER Fallacy

By Robin Hanson · March 16, 2010 9:20 am · Comments (12)

Robert Samuelson:

The uninsured, it’s said, use emergency rooms for primary care. That’s expensive and ineffective. Once they’re insured, they’ll have regular doctors. Care will improve; costs will decline. Everyone wins. Great argument. Unfortunately, it’s untrue.  A study by the Robert Wood Johnson Foundation found that the insured accounted for 83 percent of emergency-room visits, reflecting their share of the population. After Massachusetts adopted universal insurance, emergency-room use remained higher than the national average, an Urban Institute study found. More than two-fifths of visits represented non-emergencies. Of those, a majority of adult respondents to a survey said it was “more convenient” to go to the emergency room or they couldn’t “get [a doctor's] appointment as soon as needed.” … Medicare’s introduction in 1966 produced no reduction in mortality; some studies of extensions of Medicaid for children didn’t find gains.

HT Tim Starr.

Tagged as:

Efficient Isn’t Moral

By Robin Hanson · March 15, 2010 12:00 pm · Comments (23)

Efficiency isn’t morality, and it is a serious confusion to think it should be. Let me try again to explain.  I said:

Economic welfare cares not about giving people experiences but about satisfying their preferences. … If we do something a dead person would have wanted, that counts as a benefit.

Adam Ozimek responded:

But we care about satisfying people’s preferences because, unlike the dead, they can know that those preferences being satisfied. … If were going to count the preferences of the non-existent, then it would seem that the number one priority of all society would be to bring as many of them as possible from non-existence into existence. The easiest way to do this is to mandate pregnancy. … If we care about satisfying the preferences of the dead even though they won’t know their preferences are satisfied, does that mean we should not be concerned with whether or not living people know when their preferences are satisfied?

Adam reminds us of Tyler’s position:

Dead people don’t count in the social welfare function. (If they did, how many of them would prefer non-democratic or racist outcomes?  And would we count that?  We shouldn’t and we don’t.)

When our distant ancestors sat around debating if to change locations, expel a troublemaker, or attack neighbors, they were often ambiguous about whether they were choosing what they wanted or what was moral; they preferred to pretend these were the same.  We similarly prefer ambiguity when we argue policy today.

So it is important to clarify: As an analysis tool, economic efficiency is designed and well-suited to finding win-win deals that [added: tend to] get us all more of what we want. It is not well-suited to achieving moral outcomes, except when morality happens to coincide with getting people what they want.  Otherwise, win-win deals will predictably not achieve morality when many involved do not want to be moral.

Many of us want things we will never experience directly; we want our children to prosper after we are gone, for example. This is especially true of our moral wants; we want our donations to Africa to actually help real Africans. So we are understandably wary of deal-making frameworks which explicitly suggest that they seek only to achieve the appearance, not the substance, of our wants.  So yes, a deal-finding analysis tool should definitely count unseen wants!  Furthermore, observers concerned that deals might neglect morals should be especially eager for our deals to achieve unseen wants.

Frameworks for finding win-win deals should also try to include as many things as possible that can have wants and participate in deals.  This includes racists, pedophiles, slaves-owners, robots, animals, distant past and future folk, and future folk who may or may not end up existing.  Yes many may be morally offended if racists get what they want, but that offense counts in what other folks want, and therefore enough offense will ensure that win-win deals will not give racists much of what they want.

Limits on contract may distort prices and interfere with the ability of efficiency analysis to help us find useful win-win deals.  But that is a good reason to enforce more kinds of deals, not to try to distort efficiency for a task to which it is poorly suited: choosing moral acts.

Added: Bryan Caplan responds.

Tagged as: ,

Ancestor Worship is Efficient

By Robin Hanson · March 14, 2010 10:55 am · Comments (40)

Maybe not “worship” exactly, but at least great respect and deference.  By “efficient” I mean that it increases economists’ standard “cost-benefit” concept of welfare.  That is: as usually estimated, the benefits of deferring greatly to distant ancestors far outweigh its costs.  And while this does suggest that we should defer more to ancestors, it also shows just how much distorted prices can break economists’ favorite tools.

The economic welfare of a proposed change is the benefits minus the costs of that change, translated into cash terms, though of course changes don’t have to actually be cash transactions.  When available, market prices are commonly accepted as estimates of the benefits and costs of things gained and lost.  Economic welfare is a powerful heuristic for finding win-win deals: in many kinds of situations, the strategy of consistently making the changes that increase economic welfare tends to be usefully close to an actual win-win deal that gives most everyone more of what they want.

The efficient ancestor worship problem arises from two key facts:

  1. Economic welfare cares not about giving people experiences but about satisfying their preferences, i.e., giving them what they want.  And even long dead people still have (or “had” if you prefer) preferences that we could now better satisfy.  If we do something a dead person would have wanted, that counts as a benefit.
  2. At standard market interest rates, the magic of compound interest quickly gives astronomical priority to the preferences of folks who lived long ago.  For example, in historical records near risk-free interest rates (e.g., land rents over prices) consistently exceeded 9%/yr from 3000BC to 1350AD, for a total factor of over 10162.

Together, these facts suggest we would increase economic welfare if we spent less than 10162 dollars today to do anything for which a 3000BC ancestor would have been willing to pay a dollar (equivalent in their currency).

Clearly we would quickly bankrupt ourselves if we tried to implement such “efficient” changes, and doing so would not be remotely close to a win-win deal with our ancestors.  What goes wrong here?

Our contract law system refuses to enforce many win-win deals between distant generations.  Many folks would be willing to create trusts that accumulated funds long after their death and then paid distant descendants (perhaps indirectly) to do things like remember their ancestor’s name, pray to his gods, etc.  Unless stolen, such funds would eventually come to dominate the world economy and dramatically lower interest rates.  With lower interest rates, economic efficiency would count the preferences of distant ancestors as far less valuable, and as a bonus businesses and governments would have far stronger incentives to attend to the interests of distant future folks, such as via global warming policies.

But we in fact refused to enforce a great many such long term deals.  For example:

The rule against perpetuities at common law … prevents a person from putting qualifications and criteria in his will that will continue to control or affect the distribution of assets long after he has died, a concept often referred to as control by the “dead hand.”

Our unthinkingly repugnance at being controlled by the dead, and our eagerness to grab their resources, prevents us from enforcing long-term win-win deals.  This refusal to enforce deals increases interest rates, which distorts all our trade-offs across time, bringing economic welfare estimates into stark conflict with intuitive moral judgments about time trades (as in global warming), which then encourages people to turn to non-economic frameworks for policy analysis.

When policy distorts prices, it distorts calculation of economic welfare, which encourages people to ignore economic welfare when choosing policy, which reduces their reluctance to intervene to further distort prices, which leads to a sad spiral of increasing confusion.  Please, let’s enforce long-term win-win deals!

Added: A fascinating alternate history might start from a year 1300 English legal precedent enabling flexible growing long term trusts.  By 1800 early trusts grew a billion-fold, and trusts dominate the economy.  What else changed?!

Tagged as: , ,

Parable of the Multiplier Hole

By Robin Hanson · March 12, 2010 11:00 am · Comments (46)

Imagine that we discovered a “hole in space”, through which we could see an alternate Earth, filled with people recognizably like us, though different in many ways.  Those people could also see us.

While no objects could move from their side of the hole to ours, small items (but not humans) could move from our side to theirs.  Furthermore, the hole had the amazing property of multiplying everything we sent through by a factor F of a million!  That is, if you tossed a gold coin through the hole, a million identical coins would come out the hole on the other side.

How tempted would you be to toss useful items, like food, through the hole?   Remember, the cost to you, relative to the benefit to them, is 1/F, only one part in a million.  When considering the following variations, and their various combinations, consider not only F = a million, but also ponder what fraction F would make you indifferent to tossing or not:

  1. Your gift goes to a random person on the other side.
  2. Your gift goes to a government on the other side, which controls the hole.
  3. You can specify to whom your gift will go, using some simple descriptors like “poor”, “smart” etc.
  4. We could also do other things to help them, such as by studying a problem of theirs and sending them a report with suggested solutions.  But these other actions don’t get multiplied by F; a million copies of the report doesn’t help more than one copy.
  5. The hole isn’t very reliable, and only one time in a thousand does what you toss through the hole actually get to the people on the other side.  But when the hole does work 1000*F items come out the other side.
  6. You have very good theoretical reasons to think that most likely there are people much like us on the other side of the hole, but you can’t actually see through the hole (though they can see us).

The point of this parable is that interest rates would also greatly leverage any gift you gave the distant future folks.  For example, in 1785 a French author wrote a satire about Ben Franlkin, the most famous American to Europeans.  While Franlkin was famous for his Poor Richard’s Almanac, the satire mocked American optimism by having “Fortunate Richard” leave money in his will to be invested for 500 years before being given to charity.

Franklin responded by leaving £1000 each to Philadelphia and Boston in his will to be invested for 200 years.   He died in 1790,  and by 1990 the funds had grown to 2.3, 5M$, giving factors of  35, 76 inflation-adjusted gains, for annual returns of 1.8, 2.2%.  Why has Franklin’s example inspired no copy-cats?  Does no one care to help distant future folks through the multiplier hole of compound interest?

Tagged as: , ,

Hard Facts: Incentives

By Robin Hanson · March 12, 2010 1:00 am · Comments (11)

More wisdom from Hard Facts:

We … [did] research to discover if courteous clerks fueled sales.  … We ultimately found little if any evidence that courtesy increased store sales. …The main finding … was that clerks in stores with more sales were actually less courteous.  Apparently, the crowding and long lines in busy stores make clerks and customers grouchy. (p.39)

A survey of more than 200 human resource professionals from companies employing more than 2500 people … found that even though more than half of the companies used forced rankings, the respondents reported that forced ranking resulted in lower productivity, inequity and skepticism, negative effects on employee engagement, reduced collaboration, and damage to morale and mistrust in leadership. (p.107)

Individuals believe that others are motivated by money, even as they know that they are much less so. … A survey … of almost 500 prospective lawyers … revealed that 64 percent … said they were pursuing a legal career because it was intellectually appealing or because they were interested in the law, but only 12 percent thought their peers were similarly motivated.  Instead 62 percent thought that others were pursuing a legal career for the financial rewards. (p.115)

A survey of 205 executives from diverse industries found that 68 percent reported their companies had executive bonus plans because senior management believed tthat such plans would motivate executives.  These same executives reported, however, that they did not make daily business decisions based on how such decisions would affect either their bonus or those of other people. (p.116)

Students who are in school or who have chosen a major for instrumental reasons – in order to get a better job or to make more money – are much more likely to cheat than students who have chosen a course of study because of their interest in in the subject matter.  (p.124)

Tagged as: ,