Thursday , March 22 2018
Home / Impact Evaluations / The impacts of household inefficiency and lessons from boot camp: a gender world tour

The impacts of household inefficiency and lessons from boot camp: a gender world tour

Tomorrow is international women’s day and in anticipation I’m taking a look at two interesting new papers that shed insight into why thinking about gender dynamics matter and an innovative way to deal with gender stereotypes.      When we think about how households function there are two key distinctions.    First, do they act like a ...

Markus Goldstein considers the following as important:

This could be interesting, too:

Jp Koning writes Fiatsplainin’

Amol Agrawal writes The Rothschilds, a pamphlet by ‘Satan’ and anti-Semitic conspiracy theories tied to a battle 200 years ago

Amol Agrawal writes Masazumi Wakatabe, an economic historian appointed as Deputy Governor of Bank of Japan…

Bank of Japan writes Average Interest Rates by Type of Deposit

Tomorrow is international women’s day and in anticipation I’m taking a look at two interesting new papers that shed insight into why thinking about gender dynamics matter and an innovative way to deal with gender stereotypes.   
When we think about how households function there are two key distinctions.    First, do they act like a unit – i.e. when we model them, can we consider them a single “actor” of sorts (aka the unitary model).   Second, are they acting in a pareto efficient sense (e.g. there is no slack in the way things are allocated such that one person could be made better off without making someone else worth off).  
When we get into the empirics, the case of a unitary household is a subset of the larger class of pareto efficient households (aka the cooperative models).   One of my favorite papers took look into how households (mal)function is Chris Udry’s 1996 paper where he shows that women have lower yields than men in the same households in Burkina Faso and that these households could increase their total yield simply by reallocating inputs from male fields to female fields.   But they don’t, so basically they’re leaving money on the table (pareto inefficiency).  
So let’s jet over to Mexico and an interesting recent paper in this literature from Manuela Angelucci and Rob Garlick.    In a neat and creative case of using existing data, Angelucci and Garlick are using datasets around the Mexican cash transfer program Progresa/Oportunidades.  They start by looking for unitary-ness and inefficiency in households.   Unlike Udry’s work in Burkina, they’re focusing on consumption.   
I am not going to get into the details of the test -- testing for consumption inefficiency is significantly more complex than for production and if you are interested in the mechanics, I recommend the discussion in the paper as well as the literature they cite.   Using consumption data from the Progresa evaluation sample, they reject not only the unitary model, but also efficiency.  
Now, this is an average result for the whole sample and that’s where most (or even all?) of the literature stops.   Angelucci and Garlick take us one important step further.    They disaggregate the households into older households and younger households (based on the age of the household head).   And then they show us that the inefficiency is concentrated among the younger households.    For the older households, they can reject the unitary model, but they cannot reject efficiency.  Point 2:  there is meaningful heterogeneity behind some of these averages.    
So what’s going on?  Is it that the older are just wiser in the ways of household allocation?   Or that we’re only observing marriages that survive, and efficiency is a key to survival.   Or is it just something special about this cohort of rural Mexican households.   Here again, Angelucci and Garlick get creative with their use of data.    Making use of a 2007 follow up data set (the original was from 1998/9), they look for these patterns again.   In this dataset, the older cohort is born later than the one in the original dataset and so, if there is no difference in inefficiency, we’re look at a potential explanation of a cohort effect rather than some life cycle characteristics.    And indeed, in this later dataset, they find that the younger and older look the same – both are inefficient.  
Now, when I talk to policy folks and we get to the topic of intrahousehold allocation and why efficiency matters, I usually rely on Udry’s argument that there is a clear implication for welfare and growth.   Angelucci and Garlick give us another argument to use.   Going back to the earlier dataset, they can now compare households which are efficient with those that are inefficient.   They show that human capital investment (in the form of completed primary and secondary education) is significantly higher in the efficient (older) households than in the inefficient ones.   Moreover, the responsiveness of the inefficient households to the cash transfer treatment (designed to increase educational outcomes) is significantly lower than in the efficient households.   So, not only does intrahousehold inefficiency matter for investment in public goods like children’s education, the response to policy interventions to address educational deficits is going to be different based on whether households are efficient or not.  
Now intrahousehold inefficiency has a lot to do with gender dynamics (more on this in a later post).   When we think of trying to shift gender dynamics, a lot of people throw their hands up in the face of persistent gender norms.   In the next stop in our gender world tour, there was some recent exciting news about how to tackle these in a new paper by Gordon Dahl, Andreas Kotsadam, and Dan-Olof Rooth.
Dahl et. al. take us to Norway where they partner with the Norwegian military to randomly assign female recruits to some squads and not others for boot camp.   The goal here is an important one – to see how intensive exposure to folks who are not like you changes your attitudes.   
The set up is pretty simple (aside from understanding military unit organization).   Norway, like every country in the world, has significant occupational sex-segregation.   But, both men and women must register for potential military service.  If you’re selected (about 1 in 6) and a man, you have to serve.   In 2014 (when Dahl and co. are doing their study), if you were selected and a woman, you could choose to serve (it’s now mandatory for women too).  
The Norwegian military had already been moving to mixed-sex squads.   Dahl and co. work with them to randomly assign women (usually 2) to squads of around 6 people going into boot camp.   These folks will live together and work together for eight weeks. In fact, they’re not allowed to leave base during this time, so the interaction is intensive.      
Dahl and co. do a baseline with the recruits at entry and then a follow up at the end of boot camp.  It’s important to note that they are careful to avoid any kind of priming – both in the set up of the experiment (the soldiers didn’t know) and in the survey (which covered a range of questions).    Dahl and co.  look at three primary outcomes.    The first of these is whether folks agree with the statement “teams perform better when made up of the same sex.”  At baseline 63 percent of men (and 90 percent of women) disagree with this statement.    At the end of boot camp, the men who were in mixed-sex squads are 13 percentage points (24 percent) more likely to disagree with this statement.   (It’s interesting to note that the endline mean is lower – with only 55 percent disagreeing with the statement.)
The second dimension they look at is a broader gender attitude.   Here they ask whether the respondent agrees with the statement “it is important for men and women to share household work equally.”  At baseline, two thirds of men (and 88 percent of women) agree with this statement.   In this case, the exposure to mixed-sex squads results in a 9 percentage point (12 percent) increase in men agreeing with this statement.   So, the impacts seem to not be confined to just their view of gender dynamics in the military.  
As a third dimension, Dahl and co.  get into self-perception.   Here they ask respondents about a bunch of personality traits and put “I am feminine” into the list.   Close to no men say this fits them completely, so Dahl and co. code this as zero if the respondent says “does not fit at all” and 1 if they say anything else (i.e. does not fit well and upwards).  By this metric, 58 percent of men (and all women) get a 1 at baseline.     Here, the mixed-squad treatment results in a large increase, with a 14 percentage point increase in the men who do not completely disavow any femininity.  
Taken together, these results indicate that the experience of living and working together for eight weeks had a significant impact on men’s gender attitudes.   Dahl and co. go two steps further to look at other outcomes.    First, they look at whether this exposure has changed attitudes towards women in higher positions in the military.    The exposure to mixed-sex squads has no significant impact on these (note that these positions were not part of the experiment and most of them were occupied by men). 
Second, Dahl and co. try to tackle the question of how this mixing of the sexes might affect morale.   In sum, they find that exposure to mixed-sex squads has no significant impacts on enthusiasm for the military, satisfaction with military service, feeling of qualification for military service or desire to continue in the military.   So basically, this gender integration isn’t going to damage morale or preparedness. 
That’s it for the world tour – some thoughts to take forward into International Women’s Day.  And a tip of the hat to the Norwegian military for helping with this.   
Markus Goldstein
Markus Goldstein is a development economist with experience working in Sub-Saharan Africa, East Asia, and South Asia. He is currently the Gender Practice Leader in the Africa Region and a Lead Economist in the Research Group of the World Bank. His current research centers on issues of gender and economic activity, focusing on agriculture and small scale enterprises. He is currently involved in a number of impact evaluations on these topics across Africa. Markus has taught at the London School of Economics, the University of Ghana, Legon, and Georgetown University. He holds a PhD from the University of California, Berkeley.

Leave a Reply

Your email address will not be published. Required fields are marked *