kitchen table math, the sequel: 8/2/15 - 8/9/15

Wednesday, August 5, 2015


Speaking of recessions and teachers, the Hamilton Jobs Gap calculator now projects a return to pre-recession job levels sometime between January and August 2017.

The recession began in December 2007 (the crash came nine months later, in September 2008).

So ten years.

The Hamilton projection surprises me because everyone else I read seems to think we'll never return to pre-recession job levels -- "never" meaning not in this generation, at least. A big chunk of workers has been sidelined, the thinking goes, and they're not coming back.

But Hamilton seems to be assuming we can get back to the employment-population ratios of 2007.

At least, that's what I think Hamilton assumes. I don't have the patience to work through their explanation tonight. Or maybe ever.



In January 2007, 80.3% of people age 25 to 54 were employed.

Today that figure is 77.2%, up from a low of 74.8% in November 2010.

Prime working age.

Weaker economy, better teachers

Department of silver linings.


Markus Nagler
Marc Piopiunik
Martin R. West
1050 Massachusetts Avenue
Cambridge, MA 02138
July 2015


How do alternative job opportunities affect teacher quality? We provide the first causal evidence on this question by exploiting business cycle conditions at career start as a source of exogenous variation in the outside options of potential teachers. Unlike prior research, we directly assess teacher quality with value-added measures of impacts on student test scores, using administrative data on 33,000 teachers in Florida public schools. Consistent with a Roy model of occupational choice, teachers entering the profession during recessions are significantly more effective in raising student test scores. Results are supported by placebo tests and not driven by differential attrition.


How do alternative job opportunities affect teacher quality? This is a crucial policy question as teachers are a key input in the education production function (Hanushek and Rivkin, 2012) who affect their students’ outcomes even in adulthood (Chetty et al., 2014b). Despite their importance, individuals entering the teaching profession in the United States tend to come from the lower part of the cognitive ability distribution of college graduates (Hanushek and Pace, 1995). One frequently cited reason for not being able to recruit higher-skilled individuals as teachers is low salaries compared to other professions (e.g., Dolton and Marcenaro-Gutierrez, 2011; Hanushek et al., 2014).

Existing research provides evidence consistent with the argument that outside options matter. A first strand of the literature has used regional variation in relative teacher salaries, finding that pay is positively related to teachers’ academic quality (e.g., Figlio, 1997). A second strand has used long-run changes in the labor market – in particular, the expansion of job opportunities for women – finding that the academic quality of new teachers is lower when job market alternatives are better (e.g., Bacolod, 2007). However, both bodies of evidence suffer from key limitations. First, relative pay may be endogenous to teacher quality. Second, measures of academic quality are poor predictors of teacher effectiveness (cf. Jackson et al., 2014). This important policy question therefore remains unresolved.


We find that teachers who entered the profession during recessions are roughly 0.10 standard deviations (SD) more effective in raising math test scores than teachers who entered the profession during non-recessionary periods. The effect is half as large for reading value-added. Quantile regressions indicate that the difference in math value-added between recession and non-recession entrants is most pronounced at the upper end of the effectiveness distribution. Based on figures from Chetty et al. (2014b), the difference in average math effectiveness between recession and non-recession entrants implies a difference in students’ discounted life-time earnings of around $13,000 per classroom taught each year.2 Under the more realistic assumption that only 10% of recession-cohort teachers are pushed into teaching because of the recession, these recession-only teachers are roughly one SD more effective in teaching math than the teachers they push out. Based on the variation in teacher VAMs in our data, being assigned to such a teacher would increase a student’s test scores by around 0.20 SD.


Our finding that the effect of recessions on teacher effectiveness is twice as strong in math as in reading is consistent with evidence that wage returns to numeracy skills are twice as large as those to literacy skills in the US labor market (Hanushek et al., 2015).


Our results also suggest that recent improvements in cognitive skills among new teachers in the US documented by Goldhaber and Walch (2013) may be attributable to the 2008-09 financial crisis, rather than an authentic reversal of long-term trends.


To our knowledge, ours is the first paper to document a causal effect of outside labor market options on the effectiveness of entering teachers in raising student test scores.


We find that teachers who started their careers during recessions are more effective.


Chetty et al. (2014b) find that students taught by a teacher with a one SD higher value-added measure at age 12 earn on average 1.3% more at age 28. Using this figure, our preferred recession effect translates into differences in discounted lifetime earnings of around $13,000 per classroom taught each school year by recession and non-recession teachers (evaluated at the average classroom size in our sample).