Friday, July 24, 2015

A Less Controversial 3ie Replication Study: The Impact of India's JSY Conditional Cash Transfer Programme



The impact of India's JSY conditional cash transfer programme: a replication study

3ie Replication Paper 6, 2015

 
Natalie Carvalho and Slawa Rokicki

Abstract

Conditional cash transfer programmes are becoming increasingly popular in low- and middle- income countries, with the goal of improving access to health and social services and reducing inequities in access and outcomes for the poor and marginalised. India’s conditional cash transfer programme, Janani Suraksha Yojana (JSY), established in 2005, is one of the largest such programmes in the world. Along with small payments to community health workers, it provides financial incentives to pregnant women to encourage them to deliver in health facilities. Lim et al.’s Lancet article, ‘India’s Janani Suraksha Yojana, a conditional cash transfer programme to increase births in health facilities: an impact evaluation’ (2010), was the first formal statistical impact evaluation of the programme across the whole of India. This replication study, through robustness checks and additional model specifications, re-examines this recent work on the effect of financial incentives for women through the programme on reproductive health coverage indicators and perinatal and neonatal mortality. Of three analytic approaches taken by Lim et al. (2010), this replication focused on exact matching analysis, using data from round three of India’s district-level household survey (DLHS-3). We found the original authors’ results to be replicable and robust to various changes in model specifications and analysis. We were able to replicate quite closely the national and subnational results reported by Lim et al. We conducted several additional analyses as robustness checks including alternative matching estimators, analyses to account for differential implementation of the programme and random effects models to examine state-level heterogeneity in results. We found meaningful heterogeneity across states and districts in the effects of JSY on reproductive health coverage indicators and mortality outcomes. Accounting for state-and district- level heterogeneity has important implications for understanding the effectiveness of this programme.

Keywords: replication study, conditional cash transfers, maternal health, neonatal mortality, skilled birth attendance

Sunday, July 19, 2015

Diary Date: Workshop on Well-Being and Recent Irish Economic History

Kevin Denny and I are organising a half-day workshop to look at Changes in Well-being in Ireland over the last 10 to 15 years. Obviously this has been an eventful time in Ireland and we think it would be very useful at this stage to draw together what is known about the Irish case.  Our specific aim is to consider a  wide range of possible outcomes including physical and mental health as well as subjective well-being. Moreover we want to draw on a range of approaches from epidemiology, psychological medicine and the social sciences as well as different types of data. Our intention is to have  around 7 presentations with plenty of time for discussion. The event will be held in the UCD Geary Institute on Tuesday November 17th from 12pm to 4pm. A light lunch would be provided. Further details of the talks and how to register will be provided here in due course. Suggestions on the programme still welcome. 

Thursday, July 09, 2015

PhD Studentships: Developing Statistical Models for Estimating HIV Prevalence


My colleague Rosalba Radice is advertising a PhD studentship at Birbeck to work on developing statistical models for estimating HIV prevalence. Further details are available below and here: http://www.jobs.ac.uk/job/ALN048/bei-anniversary-phd-studentships-developing-statistical-models-for-estimating-hiv-prevalence.

Examples of the type of analysis involved include:

Marra, G., Radice, R., Till, B., Wood, S., McGovern, M., 2015. A Unified Modeling Approach to Estimating HIV Prevalence in Sub-Saharan African Countries. Research Report 324, Department of Statistical Science, University College London. http://www.ucl.ac.uk/statistics/research/pdfs/rr324.pdf
 
McGovern, M., Bärnighausen, T., Marra, G., Radice, R., 2015. On the Assumption of Bivariate Normality in Selection Models: A Copula Approach Applied to Estimating HIV Prevalence. Epidemiology 26, 229–327. http://www.ncbi.nlm.nih.gov/pubmed/25643102

BEI Anniversary PhD Studentships: Developing Statistical Models for Estimating HIV Prevalence
Birkbeck College, University of London - Department of Economics, Mathematics and Statistics

Qualification type: PhD
Location: London
Funding for: UK Students, EU Students
Funding amount: £16,057
Hours: Full Time
Placed on: 6th July 2015
Closes:  7th August 2015

Applications are invited for a BEI Anniversary studentship in the Department of Economics, Mathematics and Statistics, starting in October 2015. The studentship will cover fees at (Home/EU rate), plus an annual stipend of approximately £16,057. Closing date for applications is 7 August 2015.

The BEI Anniversary studentship is for applicants who are interested in the research area of Developing Statistical Models for Estimating HIV Prevalence. Candidates for the BEI Anniversary Studentship are expected to have a Master’s qualifications at  distinction level or, where this has yet to have been achieved, be able to provide evidence of high performance that will lead to a distinction. Applications are welcome from a range of cross-disciplinary backgrounds including statistics, biostatistics, computer science, epidemiology, public health.

Monday, July 06, 2015

Labelling NHS Prescriptions: Preliminary Thoughts

Disclaimer: These are preliminary thoughts on a policy issue. They are intended to stimulate some discussion of an interesting policy that has not yet had much public debate or more detailed analysis. The views, as with all posts on this blog, do not represent any official institutional position. They are my views..personal views..preliminary views..they are my personal preliminary views!!

Labelling Prescription Medication in the UK: Some Personal Preliminary Views 

The UK government is proposing to add a label to prescription medications valued twenty pounds or more stating how much they cost along with the words "funded by the UK taxpayer" (see BBC and Guardian articles for details). The stated rationale (extract from the Minister's speech below) behind this is to make salient to patients, in the context of a mostly free-at-point-of-use healthcare system, that the medications they are using are valuable and are funded from tax money. This may then discourage misuse or waste of the medication. It may also presumably bring cost considerations into the discussion between the patient and doctor. I haven't seen this stated but I also presume the rationale is to make the cost of different types of medication more salient in general public debate particularly in the context of negotiations with pharmaceutical companies.
But there is a role for patients here too. There is no such thing as a free health service: everything we are proud of in the NHS is funded by taxpayers and every penny we waste costs patients more through higher taxes or reduced services. Yet estimates suggest that missed GP appointments cost the NHS £162 million each year and missed hospital appointments as much as £750 million a year. That is nearly £1 billion that could be used for more treatments or the latest drugs. On top of which we spend £300 million a year on wasted medicines. People who use our services need to know that in the end they pay the price for this waste.
So today I can announce that we intend to publish the indicative medicine costs to the NHS on the packs of all medicines costing more than £20, which will also be marked ‘funded by the UK taxpayer’. This will not just reduce waste by reminding people of the cost of medicine, but also improve patient care by boosting adherence to drug regimes. I will start the processes to make this happen this year, with an aim to implement it next year.
The purpose of this post is to discuss some potential issues with this policy.  I have nothing to do with this policy and no special information other than what is available publicly as to how it was constructed and for what purpose and am posting purely to raise discussion about a policy issue of clear relevance to the public more generally and also the readers of this blog who are interested in the application of behavioural science in policy.

The first question to be raised is how can this policy be tested. If it is rolled out universally and without any time staggering then there are limits to what can be done as any changes could be due to anything else changing during the time period. Perhaps with precise timing it will be possible to see whether usage of medication changes consequent on the policy being rolled out. 

The potential for unintended negative consequences arising is worth discussing across a number of different facets. Firstly, it is possible that providing cost information will make people sensitive to it in ways other than wanting to reduce cost and waste of medication or being receptive to doctor's reluctance to prescribe very expensive medication. For example, cost may act as a signal of quality and make patients more likely to want to have expensive medications. It may even act as a signal that the medications are valuable and potentially can be resold. This may seem unlikely in a free-at-use system but given internet trade and the potential for non-medical uses of a number of prescription drugs it is certainly possible that reminding people for the first time that the medications they are using are potentially worth hundreds of pounds may make this more salient to them. I have absolutely no idea how to estimate the extent this will happen in advance and raise it as a possibility worth thinking about and discussing rather than something for which there is concrete evidence to believe it will happen. 

Furthermore, there is potential heterogeneity in how this label would affect people. It is possible that some patients will interpret the signal as a statement that they are being a burden. Will there be groups of particularly dutiful people who will feel guilty about taking medication they are legitimately entitled to due to this stark message being salient everytime they take their medication from the box? You could argue that it is also making salient the fact that taxpayers are paying for it. So one benign outcome would be people internalising the cost information but also internalising the fact that they as "taxpayers" have paid for it. But there is clearly a potential for certain groups to perceive themselves as being a burden and it is worth some research or at least the impression of doctors and pharmacists on the extent this is happening. There is arguably a particular duty of care not to stigmatize people who are marginalised due to poor physical and mental health. 

More generally, there is now a wide literature on the perverse effects of disclosure that should be thought through when making policies like this. Simply put, there is a strong element of interaction between a message, who delivers the message and the recipient of the message in disclosure contexts. The potential unintended consequences listed above are only a subset of the types of effects these interactions might produce. While I do not have a strong a-priori sense of the relative magnitudes of the different effects it is certainly worth testing them particularly if this is likely to be a long-run policy and potentially applicable to other areas. 

It also raises a question as to why prescription medications are a particular focus. Should labelling be introduced for a wide range of public services e.g. pointing out the full economic cost of a college place, a BBC drama or a session of the Commons? It would be worth a lot more discussion as to the principles underlying when such communications are deployed. If only a random subset of public services are subject to this labelling then it may make cost issues more salient in those domains than in others. 

There is also the potential for such messaging to be used not for the purpose of promoting efficient allocation but rather to convey broader political messages. The word "taxpayer" in this context has the potential to convey a wide range of messages. In essence pretty much everyone in the UK is a taxpayer. If you purchase a pack of biscuits you pay tax on the transaction. But it is surely possible that the connotation of the word may suggest to some that there is a difference between those in formal employment and those outside the formal labour market. The potential for the message to convey a view that those not working for reasons of illness are somehow not "taxpayers" is certainly worth thinking about. Also the potential for the continuous priming of this message to promote negative sentiment among those in the formal labour market to those not in formal employment is also worth thinking about. The counterargument is that "taxpayer" may genuinely be used to convey "all of our taxes pay for this" and is thus intended to tap into norms of community and reciprocity. Again, I think this is debatable and probably testable. 

There is no question that the UK government has been leading in attempts to use non-price forms of intervention in policy. But there is also a potential for such communication-based forms of intervention to be used to make political points, with the case of the "immigration vans" in 2013 being one prominent example. In this particular case, there needs to be more discussion about why this messaging is being used as opposed to other forms and the particular outcomes intended and extent to which the above consequences have been thought-through.


Addenda: 

1. The BIT document "Applying behavioural insights to reduce fraud, error and debt" contains a number of principles on how communications of this nature might be used. The document is very interesting and highlighted on this blog before. Though it is still well worth debating the connection between the broader principles and the specific case above.

2. The 2012 "Action Plan For Improving The Use Of Medicines And Reducing Waste" document contains a number of suggestions for doing this. It also includes the following paragraph which is interesting in the context of the current plan. It clearly needs to be reconciled why this policy was not considered suitable for this reason three years ago but has now been announced as going ahead.

"We have considered the suggestion of disclosing the cost of NHS medicines to patients, perhaps by including information on packs or labels, to raise awareness of their value to the NHS and reduce wastage. It was agreed not to pursue this. We have noted independent research commissioned by the Department of Health, which suggests that some patients, for example older people in need of effective treatment could be deterred from taking the medicines they need because they are worried about the impact on the public purse. However, we see no reason why national or local information campaigns could not highlight the total cost of medicines used in the NHS."


3. The previous research I can find on this topic in the UK context is below. It is based on the views of service providers and tends to reflect the concerns outlined above. It is difficult to conclude though that this is direct evidence to suggest the policy would not work. However, if people directly involved in this area share these concerns it puts more importance on a more worked-out rationale for why these concerns will not materialise.

Trueman et al (2010) Evaluation of the scale, causes and costs of waste medicines. Report of DH funded national project. York Health Economics Consortium and The School of Pharmacy, University of London.: York and London.

"Some sources have advocated putting the supply price on every NHS medicine pack in order to raise awareness of costs, and so perhaps to reduce wastage and/or increase consumption. This research sought to understand the views of respondents on this suggested way of reducing the cost and volume of residual medicines supplied by the NHS. It found rather less immediate support for this proposal than there was spontaneous advocacy in favour of increased direct charging. Some interviewees said it would ‘make patients realise the cost of medicines’. But in rather more cases, service user respondents and others felt that pack pricing might cause some patients in need of effective treatment to become worried about its cost to the public purse. It might also, it was suggested, in some instances encourage the ‘selling on’ of some medicines and/or deliberate requests for more costly treatments. One GP said that the people who he most wanted to target probably would not care about the price of their medicines. Messages that appeal to a sense of responsibility are by definition unlikely to affect the genuinely irresponsible. Interviewees also mentioned the problems that could emerge in situations where (as is the case with approaching a half of the items used in the community) a medicine’s ingredient price is lower than the prescription charge: In conclusion, the interpretation offered here is that even it were possible within the European setting to implement such an idea, relatively few people are on consideration likely to judge the option of printing UK prices on all NHS medicine packs advisable. Most may be more inclined to think that it would cause more problems than it would solve"

Sunday, July 05, 2015

Behavioural Economics and Financial Behaviour Links

Below are some links I promised to send to various people. Not intended to be exhaustive but if you get through the several thousands of pages of introductory material below happy to send more!

1. The FCA Occassional papers provide a very interesting source of information about the potential application of behavioural economics and related areas in regulatory contexts.

2. I have kept a broad reading list about behavioural economics and policy. Some of the key readable introductions are below

Popular texts:
1. Ariely (2008), Predictably irrational
2. Kahneman (2011), Thinking, fast and slow
3. Thaler & Sunstein (2008), Nudge: Improving Decisions About Health, Wealth, and Happiness
4. Akerlof & Kranton (2010), Identity Economics: How Our Identities Shape Our Work, Wages, and Well-Being.
5. Sunstein (2013) Simpler: The Future of Government
6. Akerlof and Shiller (2009) Animal Spirits
7. Lunn (2008) Basic Instincts: Human Nature and the New Economics

3. The Behavioural Insights Team approach developed with many UK academics is available in these papers (MINDSPACE paper and Journal of Economic Psychology paper). 

4. Bubb and Pildes "How Behavioural Economics Trims its Sails and Why" argues that much of BE has artificially restricted the domain of potential public policies by focusing on choice-preserving mechanisms - response by Sunstein gives a good sense of where the current debate is focused. 

5. My short piece on Which? conversation blog outlines some of the issues in the debate around Nudge and financial regulation. 

6. Nice VoxEU on the potential for behavioural financial regulation. 

7. Very interesting page outlining recent reports commissioned by the Australian Securities and Investments Commission and carried out by the Queensland Behavioural Economics Group

July 22nd Behavioural Science and Policy Meeting in Dublin

Since 2008 a number of us have organised an annual conference for people working at the interface of economics, psychology and related areas. Speakers have included international thought-leaders in this area including David Laibson, David Halpern, Robert Sugden, Arie Kapteyn, Ruth Byrne and John O'Doherty as well a diverse range of speakers from across economics, psychology and policy in Ireland and they have contributed to maintaining an active discussion of the potential for this area in Ireland. The next one will take place at the ESRI in Dublin on November 27th. At the previous session we agreed to organise some more adhoc meet-ups in between the events partly to disseminate new ideas and also with a view to establishing a more structured network in this area in Ireland. The first of these meetings takes place in Dublin on July 22nd organised by myself and Sean Gill. It will take place at 7pm sharp at the Roasted Brown coffee shop in Temple Bar. There will be 5 short presentations (to be listed here in the next couple of days) and some discussion about future events. Meet-ups around this area are now taking place in several cities including London and Sydney. I spoke at the Sydney event recently and it was extremely lively and led to several useful follow-ups. There are many people interested in this broad area in Dublin and Ireland more generally. This is intended to a broad forum and we welcome attendance and contribution from academics interested in exchanging ideas with a broad audience, people across different areas including students and people with business and policy interests in this area. For now we envisage the events being structured around short talks where a speaker describes briefly an idea they are working on or thinking about and potentially some suggestions for collaboration. Though there are many other event formats that could be considered. If you are attending please drop me an email at liamdelaney2011@gmail.com

Speakers:

Liam Delaney: Overview of behavioural economics, policy and business.

Michael Daly:  Psychology, Self-Control and Policy

Sean Gill: Behavioural Economics and Health

Audience Contributions, Q+A and Suggestions for Development of Network