Sunday, April 13, 2014

Random Thoughts

I have a bunch of  random thoughts, and can't be bothered stringing them together into a cohesive post. Perhaps they can't be strung together into a cohesive post. Either way, I'm posting them as little short independent sections.

--------------------------------

I wrote a Tungsten Hippo post this week! It is about how I don't think the answers to all of our problems will come from STEM, and how that  factors into my decision to start the Tungsten Hippo site.

Also, I posted a couple of great recommendations this week: Ex Libris, by Anne Fadiman, is a great book of essays for anyone who loves to read, and Why We Fly, by Evan Rails is a fun short ebook for anyone who likes to travel and wonders why.

--------------------------------

If I ever get really, stupidly rich, I have two projects I want to do:

1. I will buy a big empty lot, put down the squishy composite stuff from playgrounds, and then commission a bunch of sculptures of animals and other things that kids can climb on. Then I will put coffee carts and benches around for the parents. Probably with free wi-fi.

This idea comes from taking my kids to the zoo last week and spending at least half of the time watching them climb on statues of animals. The new koala tree at the San Diego Zoo is particularly attractive to children. It usually has several kids climbing on it and several grown ups standing near it saying "do you want to go look at some animals now?"

I no longer fight this, and just accept that seeing some animals is only a small part of what we'll do at the zoo, but I can't stop thinking that a sculpture climbing park would be much more efficient.

2. I will fund a study to find out whether or not just giving struggling people money is a more effective solution to a range of problems than our usual interventions.

This idea comes from watching people go into the paycheck lending place near my house and being tempted to just go over and hand them $500 instead. I could not currently afford to do this for more than one or two people, and just walking up to someone and handing them a significant sum of money would be weird. But if I could do it as part of a study, then it wouldn't be weird. And if I were stupidly rich, I could afford to do this for a significant sample size.

--------------------------------

I find myself wondering about sexism, racism, and other exclusionary behaviors in different fields. I wonder if it really is worse in tech than in other fields, or if the fact that people in tech are more likely to be active on social media and have a way to tell the world about the crap that happens.

I do not think tech is worse than science, but that is a personal impression, and I'd love to see some data. Does anyone know of data on this?

-------------------------------

I follow a few conservative pundit-types on Twitter. The idea was to keep myself from living in the inverse bubble from the Fox news people. I can't stand any of the Fox news people, though, so I follow David Frum and Ross Douthat. Generally, they do the job of informing me what conservatives are thinking about the news of the day without making me want to reach through my computer screen and strangle them.

Last week, though Ross Douthat got in a discussion with Matt Yglesias about the gender wage gap and since I follow Matt Yglesias, too, I saw the entire exchange. And I wanted to reach through my computer screen and if not strangle Douthat, at least shake him a little. He was spouting the line about how women just choose to work fewer hours in less well-paid fields and really, what could we do about that? If he had ever considered what factors drive those choices, he gave no hint of that.

This prompted a mini-rant from me on Twitter, but also made me wonder: are there any conservatives who have really examined their own privilege and come away still conservative? It seems like this is at least theoretically possible, but I cannot come up with anyone who would fit this bill. I'd actually be really curious to read what a conservative who recognizes the full inequalities of our society proposes to do about them- so if you have any suggestions, please leave them in the comments.

-------------------------------

In fact, leave me comments on any of the above!

26 comments:

  1. Dr. Confused10:24 PM

    I don't know if the "just give money to the poor" strategy has been tested in an urban, first-world context, but in developing countries it seems to be effective: http://www.oecd.org/dev/pgd/46240619.pdf
    http://www.economist.com/news/international/21588385-giving-money-directly-poor-people-works-surprisingly-well-it-cannot-deal

    What has been tested in the U.S.: giving homeless people homes. Also seems to work. http://www.nationofchange.org/utah-ending-homelessness-giving-people-homes-1390056183

    ReplyDelete
  2. Anonymous4:42 AM

    I don't study poverty, but I'm fairly sure just giving people money has been studied in the US. In fact, that's what EITC and Welfare both are, though with red tape involved. There's also methods where you give everybody money and then tax it away from specific people (from what I understand, this is more common in Europe) in an effort to control their behavior (ex. water usage).

    You can't just give people money when they're in a bad situation because of moral hazard. We all have moral hazard, not just poor people. That means in order to get the benefit, we will do the negative thing (or work less hard to stop the negative thing), such as get into a bad situation. That's why we have: in-kind benefits, targeting, stigma, hurdles, and a couple of other things I'm forgetting. Different programs will use different methods of combating moral hazard.

    (Also, in developing countries, *who* you just give the money to often matters. Giving it to mothers often has a more beneficial effect on the family than giving it to fathers, who often drink it away.)

    ReplyDelete
    Replies
    1. Anonymous4:50 AM

      Here's a book recommendation on poverty: http://amzn.to/1nmqqX9

      Strongly recommended if you're curious about what has been done and the different ways we think about poverty. Also I still haven't read Scarcity, but I know it is about poverty research as well. http://amzn.to/1iN54iL

      Delete
    2. Thanks for the recommendations! I will certainly read up before funding my experiment. :)

      I wasn't thinking about the amount of money that would be a huge moral hazard risk. More like $500 to get the car fixed so you wouldn't lose your job, or $100 to apply to college. You know, the things that middle class people can take in stride but that can derail someone working their way out of poverty.

      This is why I donate to Modest Needs, but I wonder how many people are put off by the requirement for someone to"qualify" your need and if it is a lot, how to fix that.

      Delete
    3. @Cloud- yes, they try to qualify to deal with the same moral hazard issue. If the reason you can't make your car payment this month is that you have a DUI fine, people aren't that sympathetic, and the charity loses its reputation if it comes out that's (or similar) what's happened with people. But the idea of just giving people money is oddly popular right now in conservative/libertarian circles. The idea is that all the apparatus put in to thwart moral hazard is enormously inefficient and creates an army of bureaucrats, and some people think the moral hazard part might be better than the bureaucrats.

      Delete
    4. Anonymous4:44 AM

      $500 is a lot for a lot of people. Without a screening (screening being one of the ways I forgot that they use to reduce moral hazard) mechanism in place (and its attendant stigma) I have at least one wealthy colleague making 200Kyear on his own who would put out his hand. Because he has strong moral hazard.

      IIRC, the studies that have been done on small windfalls show that they don't help much, though I don't know if anything has been done to formally study targeted screened windfalls.

      Personally, as I told my class when asked recently, I would prefer to give everyone a set amount of foodstamps (everyone) and raise taxes on the wealthy/middle class for the amount of foodstamps. That would get rid of the stigma of foodstamps. (Stigma being one way to remove moral hazard that yes, even modest needs uses.) I like the way that the Danish have that baby box that *everyone* gets, not just the poor.

      Delete
    5. Anonymous4:50 AM

      I should amend that-- small windfalls in the US. The international micro-finance literature is highly mixed, IIRC, with some microfinance projects showing improvements and some microfinance projects not showing much.

      Delete
    6. The problem I have with using stigma to reduce moral hazard is that it often attaches to kids.

      I'm curious how much money would be lost to jerks like your colleague and how it would compare to the money lost by paying for the screening we do, but I guess it doesn't really matter, because realistically a couple of prominent cases of fraud like that would sink the scheme.

      Maybe the idea of an income floor is the way to go. I also like the idea of child allowances, but I don't want people without kids to starve, either.

      Delete
    7. Anonymous7:56 AM

      Most programs that are aimed at kids try to use other methods besides stigma. So EITC has almost zero stigma. WIC has far less stigma than TANF. Government has been trying to reduce stigma in Foodstamps by switching to EBT.

      Instead they use hurdles (paperwork), targeting (WIC is only available to women, infants, and children), provide only in-kind benefits (specific types of food for WIC, food in general for foodstamps), and so on. (In-kind benefits being another type of way around moral hazard that I forgot.)

      It isn't fraud if it is legal. Like Mitt Romney says, if it's legal, he's going to take every tax break that's offered to him. It's on government to design appropriate programs that minimize moral hazard and maximize benefits to the targeted population.

      Delete
  3. As for climbing-on-animals, if you're ever in Philly check out Fitler Square. It's a small park with a selection of large metal animals that children (and grownups) can climb. No coffee cart though...but as I recall there is a coffee shop on one side of the square, so you could grab a drink before you set the kids loose on the park!

    ReplyDelete
  4. I grew up near D.C. and as a child spent many hours at the Smithsonian. Amazing as the museums were, my brother and I considered climbing on the fiberglass Triceratops to be a highlight of any trip. The only downside was that it got really hot in the summer.

    ReplyDelete
  5. Alexicographer9:24 AM

    Haha. Yes, our zoo has a playground, and actual mechanical equipment (e.g. a jeep) the kids can access and climb on. Like you, I've come to embrace the thought that these may actually BE the best part of the zoo -- who am I to say?

    Not exactly what you're looking for, but I was surprised a few years back to hear former (twice, I think) Libertarian candidate for NC governor Mike Munger advocating (in an academic talk he gave where I work) such things as universal health care and an (in effect) "floor" for income (I may misremember, but that is what I (mis)remember). He's not a pundit, but has traditionally been very conservative economically (not socially); you can read something of his more recent thinking on such subjects in an article titled, "Euvoluntary or not, exchange is just," in Social Philosophy and Policy, and at the blog he co-authors here: http://euvoluntaryexchange.blogspot.com/ (also some other blogging here: http://mungowitzend.blogspot.com/ , including not infrequent complaints about assorted liberal columnists, particularly economists, e.g. Krugman). I don't read either blog regularly, so YMMV, but I do sometimes skip over there when I'm bored or wanting a "different" perspective on things.

    ReplyDelete
  6. 1. I'm a fan of the climbing sculptures idea. In Madison, WI, we have two Splash Pads which are large areas of fountains for kids to run through. Free! They're awesome.

    http://www.saa-madison.com/sites/saa-madison.com/files/project-images/LVSP-1.jpg

    ReplyDelete
  7. 2. I love the Freakonomics podcast, and they did a poverty one a few months ago. "Fighting Poverty With Actual Evidence."

    http://freakonomics.com/2013/11/27/fighting-poverty-with-actual-evidence-a-new-freakonomics-radio-podcast/

    ReplyDelete
  8. Your "give struggling people money" idea made me laugh, because I just read Howards End. Have you read it? So enjoyable, and I can't believe I hadn't read it before. It's totally about class (and also other things ... art ... modernity) and features a lot of conversations by middle class ladies suggesting this very thing, which sound eerily like conversations that happen today (including this one). So, not a scientific test, but proof that the idea has been around for a while.

    ReplyDelete
    Replies
    1. I've never read Howard's End. It is sort of depressing to think that we're still have the same conversations about class and poverty.

      I was just thinking that we get so caught up in the risk that we might accidentally give money to someone who doesn't "need" or "deserve" it that we lose sight of the risk that we might not give money to someone who desperately does need it. And really, which one is worse?

      Delete
    2. Anonymous6:33 AM

      That's not all that moral hazard is though-- moral hazard includes the idea that the act of your making money available to people causes them to behave in ways that are not beneficial to themselves. So trying to spend down your assets to hit an asset threshold or keeping your hours limited so you don't hit an income threshold are also moral hazard.

      Yes, type I error is the concern picked up by Fox News (usually in the form of someone who is doing something illegal anyway, so shouldn't even count), but it's not the only problem with social insurance.

      And... communism didn't work. Probably the best bet for maximizing the total social welfare function is some balanced form of Socialism. If there were easy answers we would be doing it. If you really want to do some reading on the topic, Jon Gruber's Public Finance and Public Policy textbook is excellent. He does an amazing job of showing exactly why these problems are so difficult. I've also really enjoyed Legacies of the War on Poverty which just came out.

      Delete
    3. Yeah, I get that. My point is: I doubt we can totally remove moral hazard. But in our attempts to counteract it, are we letting too many people go hungry and suffer? And which way would we rather err. Personally, I'd rather err on the side of not having hungry people, even if some of my money is wasted and even if some people start gaming the system.

      Delete
    4. Anonymous7:50 AM

      "But in our attempts to counteract it, are we letting too many people go hungry and suffer? "

      I'm not sure I'm being clear... yes, there's type I error, and I think that it's better to let some bad guys slip through the cracks than to increase type II error. That's what you're talking about.

      HOWEVER, moral hazard also means that people will behave in ways that cause them to hurt themselves. Counteracting this type of moral hazard means that people are BETTER off rather than worse off when the government doesn't interfere. An example would be when Ronald Regan cut the disabilities rolls in the 1980s and a ton of people ended up getting jobs and when they were able to get back on disability, they chose not to because they were in a better place. No longer living in near-poverty on disability wages because they were employed. But they wouldn't have gotten to that place if they hadn't been kicked off because it was too scary to leave disability. (For that reason, government has tried to make it easier to earn money before being kicked off disability, and there's a lot of work on whether or not this works, but you also end up with the problem that more people are more likely to try to get on disability if they can both get disability and earn money, which means they'll still try to limit their hours... it's a tug of war with lots of empirical evidence on the size of things.)

      In terms of, why don't we just give everybody all benefits-- the answer is that would be lovely, but we have limited resources. We need people to be productive in the economy in order to tax them, and our ability to give benefits is limited by our ability to tax and our ability to borrow as a country. We don't want to be unable to give money to programs that are doing a lot of good because we're giving too much money to programs that are wasting it. That wasted money could instead go to a program that's benefiting people and improving the economy. Which programs are which are empirical questions.

      (At this point, the argument with my parents generally devolves into complaints about how the US gov't spends large amounts of money on wasteful things that help nobody, which is called changing goal posts, so to pre-empt that...) And yes, I would love it if the US spent less money on defense and all the other places it's "wasting" large sums of money on, but even without a defense budget, there's still limited resources. (And with that defense budget, our resources are even more limited.)

      So no, it's not as simple as saying let's get rid of type II error by maximizing type I error. There's big costs to both types of error and the size of the two errors is important. AND some programs "hurt the people they're trying to help" to quote the abstracts of most papers that find unintended negative consequences to government programs. The existence and size of these hurts are empirical questions for every policy.

      Most policy makers aren't Fox News. They get Type I and Type II error. They're trying to do the best that they can, to help the most people that they can, while hurting the fewest number of people they can, with the limited resources that they have.

      Delete
    5. I should have been more careful in my setup. I'd do an experiment like this not because I think our policy makers are bad or are operating under false guidelines, but because I think they are constrained by political realities and, as you say, finite resources. So I think it would be interesting to design an experiment that would remove those constraints and gather the data on what happens.

      This sort of thing happens in biomedical research quite a bit- rich guy gets a bee in his bonnet about something, and funds some research that would never have been funded under the usual mechanisms. Sometimes the data produced is really useful, and the rich guy has done the world a real favor. Sometimes not, and he has just contributed to keeping scientists employed and not much more.

      The more difficult thing in the experiment I was describing is that I would need to be more careful to design it so as not to cause harm. I do actually understand moral hazard, I'm just equally concerned about other harms.

      Obviously, this is not my field of expertise and I'd need experts to help me design an ethical experiment, but since in this fantasy, I'm fabulously rich, I can hire some.

      But even taking the thing that sparked the idea- what if you could do something approaching a random trial where some of the people heading in for paycheck loans just got handed the cash they needed instead. Would that make a difference or not? It would be an interesting data set, and I don't know that it would ever be done without some eccentric rich person to make it happen.

      Incidentally, I have worked on government projects and for the most part been really impressed with the people I met. Except for one dude at Homeland Security who I'm pretty sure was an utter moron, but 1 idiot out of the roughly 75 people I worked with is not a bad ratio.

      Delete
  9. OK, you guys have convinced me that my experiment would be impossible to design (and probably wouldn't produce actionable results). So how about this instead:

    Pick a county and pay a child allowance for every kid born in the county, until they are 18 (remember, I'm stupidly rich, so I can fund this). Get the names of the kids and their mothers from county birth records at the start, and then pay the allowance every month, even if they move out of the county.

    I'd rather take a "snapshot in time" and give allowances for all kids living in a county at a certain date, but I'm not sure I could get that data in a low overhead sort of way.

    I'd give the money to all parents, rich and poor. And I wouldn't announce my intention ahead of time, so it would just be people who happened to be living in that county when the experiment started,

    ReplyDelete
    Replies
    1. Oh, and I should say: I'd hire a team of social scientists to actually design and run the experiment. I'd just specify the broad outlines.

      Delete
  10. This is probably covered by some of the links above, but the signature cash-transfer program in the developing world is Give Directly, and it seems to be very effective. There is some debate over whether conditional cash transfers (for example, in return for sending kids to school, getting kids vaccinated, etc) are more effective. They certainly require a lot more bureaucracy, so when the sums involved are small, it may well be better just to give the money without conditions. GiveWell has a set of blog posts about it, if you are interested.

    ReplyDelete
  11. So the idea of a Basic Income, or Guaranteed Income, paid to everyone with no questions asked, was tried (with great success) in Manitoba a while back. It is being considered in parts of Europe and Canada.

    http://en.wikipedia.org/wiki/Basic_income

    The experiment in Manitoba:
    http://www.cbc.ca/news/canada/manitoba/1970s-manitoba-poverty-experiment-called-a-success-1.868562
    http://sociology.uwo.ca/cluster/en/PolicyBrief10.html

    People were healthier, kids stayed in school longer, a few people quit their jobs and went back to school. Turns out that most people like working and want to be productive members of society. Sure there are some folks who would rather sit around at home not doing anything...but the cost is far less than administering paperwork and deciding who is "deserving". It also removes the social stigma associated with welfare when everyone gets the same thing.

    ReplyDelete
  12. This comment has been removed by the author.

    ReplyDelete
  13. I just spent 30 mins trying to remember where I had heard about a similar financial idea. You always make me think! :-)
    http://www.npr.org/2012/07/13/155103662/struggling-families-lift-themselves-out-of-poverty

    ReplyDelete

Sorry for the CAPTCHA, folks. The spammers were stealing too much of my time.