Thesis Thursday: Sara Machado

On the third Thursday of every month, we speak to a recent graduate about their thesis and their studies. This month’s guest is Dr Sara Machado who graduated with a PhD from Boston University. If you would like to suggest a candidate for an upcoming Thesis Thursday, get in touch.

Essays on the economics of blood donations
Daniele Paserman, Johannes Schmieder, Albert Ma
Repository link

What makes blood donation an interesting context for economic research?

I’m generally interested in markets in which there is no price mechanism to help supply and demand meet. There are several examples of such markets in the health field, such as organ, bone marrow, and blood donations. In general, all altruistic markets share this feature. I define altruistic markets as markets with a volunteer supply and no market price, therefore mainly driven by social preferences.

In a way, the absence of a price leads to a very traditional coordination problem. However, it requires not-so-traditional solutions, such as market design, registries, and different types of incentives, due to many historical, political, and ethical constraints (which leads us to the concept of repugnant markets, by Roth (2007)). The specific constraints for blood donations are outlined in Slonim et al’s The Market for Blood, which also outlines the main experimental findings regarding the effects of incentives on blood donations. The blood donations market is the perfect setup to study altruistic markets, not only because of its volunteer supply but also due to the fact that it is a potentially repeated behaviour. Moreover, the donation is not to a specific patient, but to the supply of blood in general. Social preferences, as well as risk and time preferences, play a key role in minimizing market imbalances.

How did you come to identify the specific research questions for your PhD?

I was quite fortunate, due to an unfortunate situation… There was a notorious blood shortage, in Portugal, when I started thinking about possible topics for my dissertation. It got a lot of media coverage, possibly due to political factors, since the shortage happened shortly after a change in the incentives for blood donors. My first question, which eventually became the main chapter of my dissertation, was whether there was a causal relationship.

The second chapter is the outcome of spending many hours cleaning the data, to tell you the truth. I started to realize that there are many other factors determining blood donation behaviour. All non-monetary aspects of the donation process are very relevant in determining future donation behaviour (also highlighted by Slonim et al (2014) and Lacetera et al (2010)). I show that time can be a far more important currency than other forms of incentives.

Finally, I realized how important it would be for me to be able to measure social preferences to continue my research on altruistic markets and joined a team lead by Matteo Galizzi, who is working on measuring preferences of a representative sample of the UK population. My third chapter is the first installment of our work in this domain.

Your research looked at people’s behaviour. How does it relate to the growing recognition that people make ‘irrational’ choices?

The more I look into this, the more I think that we have to be careful about a generalization of irrationality. There is nothing “irrational” in blood donors’ behaviour, for the most part. So far, I have only resorted to very neoclassical models to explain donors’ behaviour – and it worked just fine.

The way I see it, there are two separate aspects to take into account. First, the market response. It is worrisome if we find market responses that are only possible if the majority of agents are making “irrational choices”. Those markets need tailored interventions to inform the decision-making process.

The second aspect zooms in into individual decision-making. In this case, it is important to determine whether there are psychological biases leading to suboptimal, or irrational, choices.

One might argue that a blood donation due to an emotional response to some stimuli is “irrational”. I strongly disagree with that categorization. For example, there is nothing suboptimal in donating blood as a sign of gratitude to previous blood donors.

The main message is that it is important to identify behavioural biases that lead to inefficient market outcomes, but “irrational choices” is too wide an umbrella term and should be used with caution.

Are any of your key findings generalisable to settings other than blood donation?

I think two key findings are quite general. The first one is the fact that it is possible to design incentive schemes that bypass the question of the crowding out of intrinsic motivation. This is a fairly general issue, that ranges from motivating employees at the workplace in general to the design of incentive schemes for physicians, to the elicitation of charitable giving, just to name a few examples. As long as it is a repeated behaviour, the result holds. This highlights a different aspect, the importance of placing lab and isolated field experimental evidence into perspective when informing policy making. There is extensive experimental literature on the crowding out of intrinsic motivation, but very little has been done at the market level and with a longitudinal component. This has limited the ability to take into account the advantages of focusing on repeated blood donation, on the one hand, and of incorporating demand side responses, on the other hand (namely by increasing the number of blood drives).

The second key aspect is the advantage of using time as the main opportunity cost faced by a volunteer supply, in the context of prosocial behaviour.

Based on your research, what might an optimal blood donation policy look like?

I believe there are two key ingredients in the design of the optimal blood donation policy: 1) promoting blood donation as a repeated behaviour; and 2) increasing the responsiveness of blood donation services in order to minimize demand and supply imbalances.

The first aspect can be addressed by designing incentive schemes targeted at repeated donors, with no rewards for non-regular behaviour. The second would greatly benefit from the existence of a blood donor registry, similar to the one already in place for bone marrow donation. This registry would allow for regular blood donors to be called to donate when their blood is needed, minimizing waste in the system. The organization of blood drives would also be more efficient if such a system was in place.

These two components contribute to the development of the blood donor identity, which guarantees a steady supply of blood, whenever necessary.

Chris Sampson’s journal round-up for 30th May 2016

Every Monday our authors provide a round-up of some of the most recently published peer reviewed articles from the field. We don’t cover everything, or even what’s most important – just a few papers that have interested the author. Visit our Resources page for links to more journals or follow the HealthEconBot. If you’d like to write one of our weekly journal round-ups, get in touch.

Solving shortage in a priceless market: insights from blood donation. Journal of Health Economics [RePEcPublished 16th May 2016

I’m drawn to studies about blood donation. As a regular donor, I’d like to better understand why I bother. This study is only loosely about individual donors’ behaviour; looking at ways of increasing donations in times of shortage. In the UK, there is no (legal) market for (human) blood. This means that a shortage can’t simply be solved by increasing prices. This study evaluates two approaches to increasing donations: i) sending out a shortage message to past donors and ii) family replacement, whereby a patient in need of blood is given the opportunity to recruit a family member for donation. Data are taken from a blood bank in China, with information on 330,000 donors and 447,357 donations from 2005-2013. Large scale shortage messages were sent out on 2 occasions to 7,858 and 3,102 past donors. Family replacement was introduced in 2010 and is associated with around 4% of donations. The identification strategy for shortage messages relies on the fact that these related to specific blood groups, and a difference-in-differences analysis with matching is used. It’s less clean for family replacement because the data cannot identify people who were invited on the basis of family replacement but did not donate. The study finds that both methods are effective, but with differing short- and long-term implications and with heterogeneous impact across different populations. This is a mammoth paper that presents a lot of analysis and provides a lot of discussion about the trade-offs between alternative strategies. While probably of huge value to blood donation policy-makers, I found myself at a loss trying to identify the main conclusions.

A comprehensive algorithm for approval of health technologies with, without, or only in research: the key principles for informing coverage decisions. Value in Health Published 11th May 2016

When NICE consider new treatments, it isn’t simply a matter of approve or reject. There are two other important options: ‘only in research’ and ‘approve with research’. The first means that the treatment can be used in a research setting, while the second means that the treatment can be used (more broadly) while research is being carried out. This study presents a decision process to determine whether a new technology should be approved, rejected, approved only in research or approved with research. The authors argue that the starting point for the decision process is whether or not the technology has been shown to be cost-effective, but that this is not the end of the story. The other factors that should determine the decision (from the 4 options) relate to whether there are significant irrecoverable costs associated with introducing the technology (e.g. set-up costs), whether further research seems valuable and could be carried out with/without approval, and expectations about how uncertainty might change over time. A complete process is presented that outlines the necessary assessments and decisions that need to be implemented along the way to achieve the best (i.e. health maximising) approach. I expect this will represent an extremely useful tool for determining guidance outcomes from assessments by the likes of NICE, and the transparency that it facilitates in the process could make for some very legitimate decision-making. My only query is whether it is worth it. Following the process is likely to lead to a lot of research, and the kinds of research that is not usually carried out. It might be that in practice there is a lot to be said for ‘muddling through’, which might lead to the same outcomes without the same research burden. I expect real practice will fall somewhere in the between.

Health insurance and income inequality. Journal of Economic Perspectives [RePEcPublished May 2016

Part of the reason we have the NHS in the UK is that we expect it to reduce (or prevent increases in) inequality, which might arise if health care was funded through voluntary insurance. The same goes for Medicaid and (to a lesser extent, perhaps) Medicare in the US. But when we think about inequality we often focus on income, and the benefits ‘in-kind’ from the likes of Medicaid are overlooked. This study considers whether – and the extent to which – Medicaid and Medicare reduce inequality. The authors review the literature and use data from a number of national surveys to estimate medical expenditure in different income groups. Key to the question is how we value Medicaid and insurance coverage: is $1 of insurance or Medicaid coverage worth $1 to recipients? If the average cost of the programmes is added to people’s incomes, then inequality is reduced by about 25-30%, but if individual expenditures are instead used then the results would be quite different. It’s also important to consider how the tax (or lack thereof) levied on health insurance affects inequality. This will lead to an increase in inequality because coverage increases with income and marginal tax rates. The authors find that taxing employer-provided insurance in line with income would reduce the ratio of the 90th to the 10th income percentile by about 4%. On net the effect of Medicaid dominates and overall health policy in the US probably reduces inequality. But from where I’m sitting, this is really the tip of the iceberg. The definition of inequality being discussed here is a narrow one. The authors rightly identify the importance of the health impact of these programmes on a broader interpretation of well-being, but note that health spending is not strongly related to health outcomes. But if we are more concerned with inequality in health (specifically) rather than income inequality, then that may not matter.