August 31, 2016 | News | Interview
The next million-dollar question in demography
An interview with Mikko Myrskylä, the new director of the Max Planck Institute for Demographic Research (MPIDR) in Rostock, about the future of our longer lives, the riddle of low fertility, gender equality and why demographic science should care about people’s well-being.
The results of MPIDR director Mikko Myrskylä's basic research are often not only interesting for demographic science, but also for politics. © MPIDR
Demographic change keeps being a top priority on the political agenda. What is the most important open question in demography?
Myrskylä: Historically, the million-dollar question in demography has been: How long can we live? Today we know that life expectancy continues to increase to levels that we haven’t experienced before. There are no thresholds looming in sight that would allow us to say that this is going to stop anytime soon. Today, the million-dollar question is: How are we going to spend these extra years? Healthy or sick? Productive and working or in retirement? Will we live independently or spend these years in nursing homes?
There is a scientific controversy, whether the time we spend disabled at the end of our lives is getting longer or shorter as life expectancy increases. That is, if there will be an expansion or a compression of “morbidity”. What do you believe is true?
It is true that the evidence has not shown a consensus, but in a certain sense morbidity is being compressed. This is not a forecast for the future, and it’s not an observation of the past. But we see from comparisons between sub-populations within nations that those with the highest life expectancy tend to also have the lowest time spent in disability.
A 2015 study in the medical journal “The Lancet”, based on the WHO’s huge Global Burden of Disease data set, seems to have proved an expansion of morbidity for the last 25 years for most countries in the world, including Germany. Is this wrong?
This expansion in time is not surprising. Much of it is likely to be driven by how health behaviors change; in the discussion of compression of morbidity one also needs to keep in mind whether we measure the number of disabled years or fraction of life that is spent in disability. For the absolute number of years, most important is the development of smoking and obesity. Obesity is becoming more frequent, whereas the number of smokers is decreasing. While smoking tends to shorten life span a lot, it doesn’t add much to the number of disabled years. The opposite is true with obesity. Those of older age today, who became obese only after mid-life, live almost as long as non-obese, but often suffer from disabling conditions, diabetes chief among these. That creates an overall increase of disabled years.
So there really is an expansion of morbidity?
Looking across sub-populations and from those with low to those with high life expectancy, there is rather a compression of morbidity. We observe it when comparing populations with bad versus good health behaviors. Then we are seeing shorter lives with longer periods with disability versus longer lives with shorter periods of disability. This not only shows that long life with a short disabled period can be attained already today, it also shows that if health behaviors were to improve this is possible not just for selected sub-populations but for large fractions of national populations.
The key message here is that disability at the end of our lives is not a matter of pure fate. The future of morbidity depends critically on the future of smoking, obesity, and other health behaviors. And they can be changed towards the better. If that in fact happens depends a lot on politics.The most impressive example is smoking. Bans of smoking and public service campaigns on its damaging effects have led to smoking going out of fashion in many countries. We now need to think about policies to tackle obesity, too.
Aren’t there already various programs against obesity?
Unfortunately, none of these programs are really effective. I do not think anyone knows how to stop the obesity epidemic. It appears, however, that trends are changing slightly. In the most obese countries the growth-rate of obesity is slowing down, if it hasn’t already stopped. But why is this happening? This continues to be something we still don’t know much about.
Are good health and long life possible for all? Or only for the privileged?
It is true that the poor die young. It is one of the most persistent findings in health sciences that those less educated and with less resources and lower occupational positions tend to be in much poorer health and have shorter lives. It’s close to a universal finding throughout modern societies. It doesn’t matter if medical care is financed by the state or privately, or if income equality is high or low. In a recent study with data from Finland we showed that the difference in life expectancy among the richest and poorest 20% of men was almost 12 years. The gap had grown by 4 years over the last 30 to 40 years. More than two thirds of that gap is explained by differences in smoking and alcohol consumption. So again, health behaviors are crucial – also in the context of rich compared to poor.
Is the social divide in lifespan unjust? Or is bad health behavior purely an individual’s fault?
It is a profound and complex question what is unjust and what is not. To me it seems obvious that there is a clearly unjust inequality in developed nations, where the length of your life strongly depends on the education of your parents.
What should politics do about this?
Invest in the education of our children. In the education of each of them, now and in the future.
Will the widening gap in mortalities one day stop life expectancy from rising altogether?
This is unlikely. Despite the growing mortality differentials, the trends of life expectancy are positive for most parts of society, also for those with the least resources, with only few exceptions. So it is unlikely that growing inequality would halt life expectancy increases.
Many politicians feel that rising life expectancies put an unbearable burden on social systems. Will we have to do with much lower living standards in the future?
There might indeed be challenges for social systems, especially the pension and health care systems, if the worker-to-non-worker ratio decreases more and more. But first we need to find out how far this decrease is actually the case! It seems that there are developments that can at least mitigate the problems. Not only are official retirement ages being raised in several countries, but we also see a higher fraction of people who are willing and able to continue working up to official retirement age or even longer.
We are setting up a new research focus on this at MPIDR called „Labor Demography“. We want to find out how working life expectancy changes over time and across high-income countries. We are taking the full life-course perspective, including not only retirement but also periods of parental leave and educational timeouts. We want to know how changes in health at older ages, as well as policy changes and shifts in economic opportunities, have influenced the timing of entry and exits from the labor force, and ultimately the length of the full working life. Answers to these questions will shed new light on how aging nations can sustain social systems.
When discussing aging of societies, low fertility is often seen as a key problem. Is that a correct assumption?
The public and policy discussion on fertility automatically implies that there is a problem. What many see as the ideal birth rate to be achieved, namely the replacement level of approximately two children per woman, only means that the current cohorts have the same number of children as the number in the cohort itself. This concern focuses on the size of birth cohorts and the subtext is that less people would be bad. This is misguided because of several reasons. First, with increasing longevity and fluctuating migration levels, the level of fertility is only one factor that influences the population size and age structure. Second, the public discussion on fertility focuses on the quantity of lives. Instead we need to think about the quality of life.
But isn’t a shrinking population a real problem for the society as a whole, especially for the economy?
Low or decreasing fertility may have many effects, and it is not necessarily true that low fertility levels will do damage to society. It is in fact a difficult argument to make that declining population size would be bad. It is easier to argue that the changing age structure – more older people and less working age people – is a problem. To what extent this is an issue depends on how healthy and productive the aging population is. It is also possible that through declining fertility we are able to invest more in the education, or human capital, of the smaller cohorts and hence get a more productive population.
Is there an explanation for low fertility?
There is no mystery on low fertility today if that means the mere fact that birth rates have been below the replacement level for decades now. Human fertility was very high 150 years ago, before birth rates started to drop universally in the developed world, in most cases following a preceding universal drop in death rates. This is what we call demographic transition.
In the latter half of the 20th century fertility fell below the replacement level in many countries. I do not think fertility will reach replacement levels again. In some countries it has, but currently I see no force that would be driving birth rates that high again on a broad scale. Individuals are deciding if having children fits for them given the constraints they face and given their preferences. On the whole, these decisions may aggregate to fertility levels that are well below two births per woman. That’s what we have observed in Western Germany for close to half a century now.
Do we understand what keeps people from having more children today?
Unfortunately, we do not yet completely understand what we see by comparing different low fertility countries or countries over time. Why does UK have higher birth rates than Germany? Why is Swedish fertility going up and down like a roller coaster? Why does fertility continue to be fairly high in France but fairly low in Italy and other Mediterranean countries?
You can easily set up a large model to explain cross-country differences, packed with many potential factors which all might matter. But that doesn’t help much. One has to distill from the potentially large number of factors the handful that really matter.
Is there a chance to better identify relevant causes for low fertility?
Yes, if we pay attention to two things: Firstly, the crucial question is not why birth rates might be low from a historical perspective. We know that people are having less children than they state they would optimally like to have. We need to answer the questions where do these preferences come from and what is keeping people from realizing their wishes. Secondly, I believe that the key to understanding fertility decisions is to link them to well-being. Both individual level and societal level measures of well-being will be critical to understand why people have or do not have children.
On the macro level the reasons lie in societal structures, which include gender equality, and in overall social development. For individuals, fairly mundane factors are also important, such as the availability of child care, or the career implications of taking time off from work to care for the children. Scientifically, we are facing the challenge to operationalize and define the societal factors, that either stimulate or inhibit fertility by interacting with the individual factors. We are currently looking into how to model and simulate this.
But there must already be some clues as to what the important factors are? Why do people not realize their desired number of children?
I think low birth rates in Germany to a considerable extent are a consequence of the constraints that women and couples face when they try to combine families and their jobs. A really critical factor is: How is having children going to influence my career? Can I take time off from work to look after my kids? Practically, these questions very much come down to gender equality. For instance, countries which manage to share the career effects of children through gender equalized parental benefits, appear to have higher fertility than countries that don’t.
Is gender-equality a problem especially in Germany with its persistently low birth rates?
Germany is still lagging behind in gender equality. Politics have made some progress, but attitudes change slowly. For example, the new parental leave scheme introduced in 2007 (“Elterngeld”) is totally neutral with respect to gender. It allows for both women and men to take parental leave of the same length, say seven months both. In practice though, the overwhelming majority of men only takes up to two months which they would forfeit if they didn’t. At the same time, women regard it as natural to take one full year off.
But more and more men in Germany do dare to take parental leave. Isn’t that a good sign?
Yes, it is. But if the father takes as much time off as the mother – which would be truly equal – you are still an outlier, far beyond the norm of what working men are “supposed” to do. Attitudes have not yet changed that strongly. Neither have those of mothers and fathers, nor those of employers.
What is Germany supposed to do about this lack in gender equality? Just wait?
Partly yes, because the direction of change is probably towards more equal sharing of parental responsibilities. But of course people making decisions on childbearing are making those decisions now. Many would prefer modern, more gender equalized circumstances now. And politics can certainly help by removing the barriers for more gender equality by developing the Elterngeld system further. It could promote shifting leave months back and forth from mother to father, or it could even require an equal amount of time for both, say seven or eight months. Sharing the time more equally could be something parents might be very happy with, but perhaps they just don’t dare to ask their employer for it yet – or their partner. Removing the cap of leave allowances at two thirds of the former income, or 1,800 Euro per month could be another improvement to Elterngeld. As long as men earn more, this cap means that when the father takes the leave, the household income suffers a bigger drop than when the mother takes the leave. Also, the cap not only causes parents financial difficulties during the crucial first year of their children, it also sends an unfavorable message: Yes, we do support you when having children – but not fully.