Theorising mortality convergence
In this post I summarise and reflect on the theory of mortality convergence by French demographers Jacques Vallin and France Meslé. Their work is very much in conversation with Abdel Omran’s theory of epidemiologic transition (ET), which I considered in my last post.
Vallin and Meslé develop their theory of mortality convergence over a number of publications in the early 2000s, but I find that the theory finds its clearest articulation in their 2005 paper in Genus. They agree with much of Omran’s theory and highlight the power of ET to explain much of what had happened until the mid-20th century. However, they break with other critics (and Omran’s updates of ET theory) by arguing that simply adding on new transition stages is insufficient to explain the contemporary patterns in mortality. They propose instead to more thoroughly refigure the theory by shifting from ET to the concept of a health transition. They offer two arguments for the necessity of this step. First, they argue that Omran’s theory implicitly assumes a continuous process of convergence between all populations, which has been empirically shown not to be the case since the 1960s. The second is that the developments in population health since the 1970s rely mostly on medical technologies and changes in individual behaviours, thereby intimating that these issues are missing from Omran’s account.
Placing Julio Frenk’s concept of the health transition at the core of the theory manages to: “include not only the development of epidemiologic characteristics within the overall health situation, but also the ways in which societies respond to the health situation and vice versa.” While I think that Omran’s original paper and the two updates by no means ignore the issues of societal and technological change - they are present as underlying determinants of ET in all three -, I agree with Vallin and Meslé that using the more expansive term may more successfully highlight them as both determinants and outcomes of population health processes. It may be more useful to think of a broader health transition when considering contemporary patterns in international mortality where economic and medical technological development have become core aspects of the phenomenon.
From my point of view, the key innovation of mortality convergence theory lies in the explicit consideration of phases of successive convergence and divergence between mortality trajectories of populations. (The populations here are mostly national, but Vallin and Meslé also generalise their theory to inequality in mortality trajectories between subnational areas, genders, and socioeconomic groups.) Vallin and Meslé argue that these patterns are the consequence of mortality-relevant innovations appearing in certain populations, which leads to increases in life expectancy and positions the population at the vanguard. An example of such an innovation may be a medical procedure like bypass surgery. Other populations then copy these innovations at various rates and with varying amounts of success in an attempt to match the vanguard population’s life expectancy gains. Vallin and Meslé emphasise that this is a population-specific journey and the ability to innovate and copy differs between countries.
The differential ability to innovate and copy leads to the observed patterns of convergence and divergence between mortality trajectories. Divergence occurs either because vanguard populations figured out a new way to reduce mortality, or because the laggard populations are unable to copy the vanguard’s innovations for some time. Conversely, convergence may only occur when there are no substantial mortality-relevant innovations being developed and the laggard populations are able to successfully copy the vanguard’s innovations. In their own words:
“In reality, each time a major change occurs in the way to fight or to prevent diseases, some pioneer populations are ready to take benefit immediately, according to their socio-economic or political situation while others need time to fulfil the required conditions to benefit. And when, finally, they do benefit, some of the first one or possibly others get new ways to make new progresses and a new gap opens between those who are able to take benefit and those who are not ready.”
The quintessential example of the theory in action is the development of the east-west mortality divide in Europe. Vallin and Meslé often argue that the mortality gap arose due to the inability of the centralised health system in communist states to make available technologies that revolutionised cardiovascular care. They argue that (1) these countries lacked the resources necessary to procure such technologies, and (2) the political configuration was unable to inspire the necessary behavioural change in terms of diet and other cardiovascular risk factors. Between 2004 and today, many empirical studies that explicitly or implicitly use the theory emerged. Examples include a comparison comparison of West Germany with East Germany and Poland (Nolte et al., 2002) that identified the quality of healthcare as a potentially key determinant of convergence; comparison of Russia, Lithuania, and Belarus in terms of the speed of economic transition (Grigoriev et al., 2010); comparison of the Baltic States (Jasilionis, 2011) that identified implementation of health system reforms as potentially key determinants of convergence. More recently, the 2017 special edition of European Journal of Population Studies focuses on CEE as an interesting case study and includes papers by Fihel and Pecholdova on Czechia & Poland, Grigoriev and Pecholdova on German reunification, and Timonin et al. on Russian regions.
While it is possible to interpret the empirical evidence gathered by these studies in line with the general theory, none of these case studies constitute direct tests of the theory. For that, some important details about the mechanism of described effect of convergence and divergence are still missing. An important example is the missing description of the processes of communication and exchange between the populations that could function as a mechanism of convergence. One possible way forward would be to combine the diffusion of innovations theory (see Rogers, 1962 and Dearing and Cox, 2018) with the Vallin and Meslé theory of innovation-and-diffusion-driven cycles of mortality divergence and divergence. With this, we could start to produce concrete testable hypotheses:
- Convergence would happen first in the wealthy populations as they have the ability to access medical advances through travel or small-scale importation (assuming free market and free movement after 1990-1991)
- Convergence would happen fastest in regions with higher connectivity to the vanguard (e.g., geographic proximity, cultural/historical proximity, capital cities, university towns, high FDI-areas)
- Convergence will be sped up by exogenous events that increase connectivity (e.g., trade deals, freedom of travel, more expert meetings; all associated with joining the European Union)
- Convergence will be more equitable in places with stronger protections of healthcare access for whole population (e.g., no cost at point of service, good geographical coverage of services, etc.)
- Convergence will be more geographically unequal within countries with highly decentralised systems of governance (e.g., trade/import rules, welfare rules, healthcare systems)
- Convergence will focus on causes of death that are clearly preventable by the application of medical technology or identifiable social innovations (laws, behaviours), which can be measured by amenable mortality.