Women in politics and spending on welfare

After another brain-draining exam season, my brain is finally ready to do something other than try and remember the content of this ginormous stack of flashcards:

The bidding starts at 15€.
Now, I shall spend the next few months hectically trying to put together my dissertation and to put myself in the mindset to start writing again, I dedicate this blog post to my dissertation topic: The relationship between women in politics and spending on welfare.

As you all know, I love feminism and politics so I thought why not write a dissertation about it. In the past, I've written about gender quotas - a policy instrument that aims to increase female representation by setting a minimum requirement of representation for legislatures. There's an endless back and forth about whether quotas are good or bad, but today many countries have implemented some form of the gender quota. But, what this policy actually suggests is that maybe, female and male policymakers are quite different and it could be beneficial to have some more women on the legislature. While to many of us this might seem like an obvious statement to make, social scientists are not always so easily convinced. What they need is some evidence that legislator gender actually matters for outcomes and that's what I'm trying to show with my dissertation. Because I only have a few months' time to put this thing together I should focus on one specific area where it might make a difference. Conventional wisdom (read: stereotypes that tell girls from a young age how they should behave) suggests that women are more caring, and experimental evidence finds that in many settings women tend to be fairer and more altruistic than men. In the political context, this could mean that women prefer policies that favor public goods. Public goods are a type of good or service that benefits a huge amount of people but nobody really wants to pay for it so the government provides these services.  Examples of public goods are infrastructure, education or health. Furthermore, because women spend a disproportionate amount of their time on childcare one might think that female legislators would prefer childcare policies (because why would male legislators use their power to improve equality, no, that would be ridiculous). So, these are the areas I focus on. Overall, I want to estimate the relationship between the share of female representatives in the legislature and spending on welfare. 

As you might have already thought, I am not the first one to come up with this idea. Indeed, Nobel prize winner Esther Duflo (yes, women win Nobel prizes in economics now, deal with it) did this exact same thing in a 2004 paper with coauthor Raghabendra Chattopadhyay. They find that the gender of the legislator does matter for spending allocation. In particular, they find that female council heads tend to invest more in the types of goods that benefit their own gender. Another paper, by Helena Svaleryd, on municipal spending decisions in Sweden finds that municipalities with a higher share of women spend relatively more on childcare/education compared to elderly care. If these papers already confirm our suspicion that legislator gender matters then why am I doing the same thing again you may ask. Well, it's because this type of research doesn't always find a significant gender effect. Indeed, on their paper on municipal spending in Italy, Massimiliano Rigon and Giulia Tanzi, find no significant effect of gender on spending allocation. Fernando Ferreira and Joseph Gyourko in their paper on U.S. mayors also find no effect of gender on policy decisions, spending, or crime rates. So, who's right? We don't really know. Maybe, whether gender has an effect depends on the particular context, the specific country, level of observation or time period. Or, maybe women can only become effective once a specific threshold of representation is met. Either way, more research is needed. That's where I come in.

In my dissertation, I look at U.S. state legislatures. The United States has a federal government that makes decisions for the entire country, but the individual states have their own legislatures with their own revenues, expenditures, and programs that provide services for a large number of people. I look at whether states that have more women in their legislature spend relatively more on welfare programs. This state-level analysis is useful because I can easily compare the state legislatures as they all operate within the same political system. If I compare the United States with Germany, for example, there are many things that could explain why one country spends more on welfare than the other that might have nothing to do with the number of female representatives. This example illustrates the difficulty of identifying a causal effect. Ideally, I would want to know if female representation causes the change in welfare spending. In a perfect world, I would have a number of states that are exactly the same (same number of people, same history, same attitudes towards women, same income,...) except for the number of female representatives they have in their legislature. Sadly, I don't live in that world, but in reality, I try to get as close as possible to this scenario. The main problem is that many things that might influence both how much states spend and how many women are elected I can't really observe. For example, a state's unique history might influence how much they spend on welfare but also whether they see women as fit for politics. You can't directly control for these types of things in your analysis, so you have to find a different way. There are many different ways you could try and do that, the papers I mentioned above all take a different approach. Given my skill set and time constraint I opt for an easier (but also less robust) version: panel data analysis with fixed effects. Basically, I look at many (50) states over many (~30) years and hope that any differences between states "average out" over time. This hopefully allows me to eliminate the effects of factors such as state history. 

Of course, there are many ifs and buts with my analysis, so I need to be careful when I interpret my results. For example, I might find a gender gap, but is legislator gender actually the reason for this gap? It could be that all the women have higher education levels, or all the men have higher incomes, and in the end, it is these differences in education/income that drive differences and not the gender of the legislator. So, to summarise, there are a lot of things to look out for, and a lot of ways I could go wrong, but hopefully I will be able to say something about the relationship between women and welfare spending in the end. Stay tuned. 

The papers I mentioned above are all super cool, have a look: 

Chattopadhyay, R. and Duflo, E. (2004), Women as Policy Makers: Evidence from a Randomized Policy Experiment in India. Econometrica, 72: 1409-1443. doi:10.1111/j.1468-0262.2004.00539.x

Ferreira, F. and Gyourko, J. (2014). Does gender matter for political leadership? The case of U.S. mayors. Journal of Public Economics, 112, pp.24–39. https://doi.org/10.1016/j.jpubeco.2014.01.006

Rigon, Massimiliano and Tanzi, Giulia, Does Gender Matter for Public Spending? Empirical Evidence from Italian Municipalities (April 27, 2012). Bank of Italy Temi di Discussione (Working Paper) No. 862. Available at SSRN: https://ssrn.com/abstract=2057858 or http://dx.doi.org/10.2139/ssrn.2057858

Svaleryd, Helena, Women's Representation and Public Spending (2009). European Journal of Political Economy, Vol. 25, No. 2, 186-198; IFN Working Paper No. 701. Available at SSRN: https://ssrn.com/abstract=988885

No comments:

Post a Comment