Richardson and Shiell provide helpful overviews of our article, make a number of important points about the difficulties and challenges of research in this area, and offer some thought-provoking suggestions on the priorities for research. We focus below on a few minor points of disagreement, where we choose to give a slightly different emphasis.

Shiell’s starting point is the 2008 WHO Commission Report on the Social Determinants of Health (Commission on the Social Determinants of Health, 2008). As Shiell points out, this supra-national report – for perfectly understandable reasons – takes a ‘single issue’ approach that focuses on the objective of reducing health inequality, treating all other equity considerations (e.g. liberty) as merely a constraint on the pursuit of the focal objective. Of course, national government agencies actually responsible for decision making may be forced to take a more nuanced – or, if you prefer, messy – approach that considers multiple, potentially conflicting, efficiency and equity objectives of the diverse kinds outlined in our paper.

Shiell advocates the development of authoritative health equity ‘checklists’ for the conduct of primary studies, systematic reviews, and economic evaluations of equity considerations – perhaps building on the Cochrane Health Equity Field checklist for systematic review authors (Morris, Reference Morris2008). Although this sounds a sensible medium-term aim, we suspect it may be premature to start developing a checklist for cost–equity analysis. It may be wise to accumulate a body of practical experience in performing applied cost–equity analyses before attempting to produce specific methodological guidance.

We fully agree with both Shiell and Richardson that much difficult methodological research remains, including research that starts to unpack the multiple pathways and dimensions of socio-economic status. Shiell’s remark in this respect may prove to be highly prescient – in the analogous case of early research into efficiency measurement using Quality Adjusted Life Years 20 or 30 years ago, issues of multiple pathways and dimensions along with many other daunting challenges did not prove insurmountable barriers to progress (Williams and Cookson, Reference Williams and Cookson2006). One minor disagreement with Shiell in this respect is that in the case of applied research on measuring inequality impacts, we do not think that ‘all of the easy research has been done’. On the contrary, we think that plenty of easy applied research remains to be done, using standard techniques to perform pilot inequality impact analyses. What has been lacking hitherto is funding and political interest.

We agree with Shiell that research on ‘upstream’ determinants is important. However, we cannot agree that research on ‘downstream’ determinants ‘will only make things worse’. Downstream policies like increasing uptake of statins and smoking cessation among deprived communities can make a real tangible difference to health inequality; as can policies like increased taxation on cigarettes and alcohol (if only public health bodies are allowed by finance ministries to evaluate and recommend them).

We also agree with Shiell (and therefore perhaps slightly disagree with Richardson) that research on health inequality impacts has more immediate potential benefit to human health and well-being than research on equity weighting. Research on equity weighting is not an entirely ‘academic’ activity, however. It is germane to the interests of bodies like the National Institute for Health and Clinical Excellence (NICE) in the UK, which undertakes appraisals of public health interventions. Neither is it unimportant from the perspective of Shiell, the WHO Commission, and other advocates of a strong policy focus on ‘upstream’ determinants of health inequality. If they want to convince policy makers to engage in substantial redistribution of the ‘upstream’ determinants of health inequality, they will need to present convincing, thoroughly validated and robust evidence that the general public support such redistribution even when it appears to hamper other more politically pressing policy objectives, such as economic growth.

A minor point of disagreement with Richardson is his claim that opportunity cost analysis ‘cannot be attempted’ unless there is a fixed budget constraint. A form of opportunity cost analysis can always be attempted, so long as it is possible to quantify how far a policy objective (such as income, or health, or educational attainment) will have to be sacrificed in order to satisfy an equity consideration. The magnitude of the sacrifice – for instance, the money or health or educational attainment forgone – represents the opportunity cost of the equity constraint. The advantage of having a fixed budget, as in the health care case, is that it may be possible to identify a ‘benchmark’ magnitude of opportunity cost that is generally acceptable to decision makers in the sector – such as the NICE ‘cost per QALY threshold’. However, opportunity cost analysis can still be worthwhile, and informative to decision makers, even when no such benchmark exists.

We are grateful to Richardson for citing a number of counter-examples to the general finding that the public value child health more highly than adult health; there is indeed plenty of ‘noise’ in this area and hence our mutual advocacy of further research. Richardson also cites the use of age weights in the WHO Burden of Disease work (Murray and Lopez, Reference Murray and Lopez1996), as a counter-example to our claim that equity weights are rarely used in economic evaluation. However, the age weights in question were not really ‘equity weights’ but rather ‘efficiency weights’ designed to allow for the relatively high productivity of people in their child-rearing years.

Despite these minor academic differences, we find ourselves largely in agreement with the thrust of both Richardson and Shiell’s commentaries, and share their sense that this is an important and under-researched area, ripe for change and progress, whose time may finally have come. Our hope is that in 20 or 30 years time, we will not only have the answers to many of the methodological challenges posed by Richardson and Shiell, but that decision makers will be taking note of those answers and using them to inform, strengthen, and improve policy on health equity.