Ragnar Frisch in the early 1960’s had high hopes for future Soviet economic development:
The blinkers will fall once and for all at the end of the 1960s (perhaps before). At this time the Soviets will have surpassed the US in industrial production. But then it will be too late for the West to see the truth. (Frisch 1961a)
Contemplating whether or the number of children to have? Take a look at “A Global Perspective on Happiness and Fertility” by Margolis and Myrskylä. The authors use data from 25 years of the World Value Survey, totalling 86 countries and over 200 000 respondents. They are interested in what the relation between what people answer on the question “Taking all things together, would you say you are very happy, quite happy, somewhat happy, or not at all happy?” and their number of children.
Margolis and Myrskylä find that a higher number of children is associated with lower happiness, but stress that looking at this in the aggregate is highly misleading. This is shown by breaking the data down by subcategories and plotting the results. In particular, they “find that the association between happiness and fertility evolves from negative to neutral to positive above age 40,” as shown e.g. here:
FIGURE 3 Happiness and number of children by age and sex, from Margolis, R. and Myrskylä, M. (2011), A Global Perspective on Happiness and Fertility. Population and Development Review, 37: 29–56. doi: 10.1111/j.1728-4457.2011.00389.x
So more children may pay off in the long run. Though it must be said that this is just descriptive, but valuable and interesting nevertheless. There are more graphs like this one, and the results can be understood simply by looking at the graphs.
Those are the words of James Heckman, from a lecture (slides, paper) at the University of Oslo last week. In particular, it is Trygve Haavelmo’s 1943 paper The statistical implications of a system of simultaneous equations (pdf) that gets the honor of being “the first rigorous treatment of causality”. A summary:
According to Heckman, Haavelmo built on Marshall’s general idea of ceteris paribus to define fixing (“an abstract operation that assigns independent variation to the variable being fixed (p. 8)”), that is to be distinguished from classical statistical conditioning (“a statistical operation that accounts for the dependence structure in the data (p. 8)”). This fixing occur hypothetically, thus causality becomes defined in terms of thought experiments, along the earlier thoughts of Ragnar Frisch. In Heckman’s words: “Causal effects are not empirical statements or descriptions of actual worlds, but descriptions of hypothetical worlds obtained by varying – hypothetically – the inputs determining outcomes. (pp. 2-3)”.
Much of the lecture and paper is a polemic against Pearl’s do-calculus. Those interested in that debate can read Heckman and Pinto’s paper and Pearl’s comments on it, watch a conference discussion they had last year, or read stuff that more able people than me havebloggedabout before. Not debatable, though, is that Heckman knows to please his hosts.
Discussing economists’ reliance on measures of growth and GDP, Joel Mokyr asks:
So you go to somebody who is about to have surgery and you ask him, How much would you demand to be paid if I took out your appendix without anesthetizing you, without putting you to sleep? Nobody would agree. The sum would be infinite. What can anesthesia contribute to GDP when it was introduced in the 1850s and 1860s? Russ:Could not be very much. Guest: Nothing. It’s very small. But that is exactly the kind of thing we fail to account for in our calculations. So that’s why I gave that whole list of things; and we could make this list infinitely large. It is the small things that actually don’t amount to an awful large part of our income and product that actually have improved life a great deal and that we really wouldn’t want to do without any more.
“Open data” and “open government” have become buzzwords in recent years, but are often conflated. Mark Headd at the Civic Innovations blog refers to the “Open Data vs. Open Government” debate, which he considers an issue of data for transparency vs for “operational” needs. He emphasizes the need to keep in mind that the original objective of the open data movement was transparency, and warns about getting lost in bus schedules, etc.
As an example of data for transparency he gives the City of Philadelphia’s release of complaints against Philadelphia police officers. Most probably agree that police complaints are not a good thing and might provide clues into bad behavior. However even in this case one should be careful about stressing the transparency/accountability angle, since this easily creates the conception that the goal is to find wrongdoers, whereas what should be the goal is to learn. If there are systematic factors affecting complaints, it would be valuable to learn about them. And even in the absence of these, a complaint does not imply a presumption of guilt, that is for other investigations to determine.
If someone fears being subjected to unfair criticism, that is a legitimate motive for non-cooperation.
Two collections of interesting urban government datasets can be foundhere.
Chris Blattman rants against the resistance to trying to estimate cost-effectiveness that he has encountered in the aid world. One thing he writes about is the “we do not experiment on people” argument (counter: there are always some who gets the stuff and some who do not).
Another expression of reluctance to measurement that I have encountered is: “We understand that we should be held accountable to donors, but why the need for such tight control? Don’t they trust us?” But this gets wrong the rationale for measuring, which primarily is to learn about the effects of what we do in order to do it better. Even if you are not accountable to anyone, measurement may help you learn what you do best and to improve.