Aging in the United States

A tweet by Paul Krugman, 2008 Nobel Prize in Economic Sciences, on December 31, 2017 caught my attention. As life expectancy is such an important factor in determining the pace of aging in the population, I thought the United States should have a very distinct population aging pattern. Hence, after looking at aging in Japan, Sweden and Canada, it made sense to check on the US case.

The chart in Krugman’s tweet, originally included in the referred Washington Post’s article, portrays a rather dim situation: from 1979, life expectancy in the United States has grown much less rapidly than on average through the OECD countries. At the same time, the National Center for Health Statistics (NCHS) published data for 2016, signalling a second year of decline in life expectancy in a row. “Preliminary data suggests that U.S. life expectancy may drop even further in 2017 — a three-year decline not seen since World War I and the global influenza pandemic a century ago”, writes Thomas R. Frieden, Director of the Centers for Disease Control and Prevention from 2009 to 2017. To explain this situation, he makes a direct link with “the dramatic increase in deaths from illegal opioids — particularly illicitly manufactured fentanyl”.

How does this play out in the long term process of aging? With our chart below and the evolution of the forces that power aging, let’s review the situation in the United States since 1950.

In the United States, aging does not appear as a monotonic process just more or less rapid than in countries we previously studied. Back in 1950, it was at a level (8.2%) marginally higher than in Canada (7.6%), lower than in Sweden (10.2%) and two-thirds higher than in Japan (4.9%) [1]. Then, by 1995, it reached 12.7%, progressing at a pace similar to Canada but much lower than in Sweden and Japan. Instead of continuing a smooth progression, it plateaued at that level, only getting to 13% by 2010. Only from then it progressed fast, projected to gain 7.4 percentage points in the 20 years to 2030, its fastest growth in any 20 year period – much faster than in Sweden, but not as fast as in Canada and Japan, over the same period. This pattern of aging has been mostly modelled on the behaviour of the fertility rate.

The evolution of the fertility rate from 1960 to the late 1970s has been largely similar to that in Canada: starting from a high rate (3.7 births per women aged 15-49 years), then declining rapidly to 2 by 1972 and getting stable at 1.8 until 1986 – not moving below as it did in Canada [2]. In the following years, unlike in most other countries, the fertility rate moved up again, reaching 2 by 1989 and then hovering at 2-2.1 until 2009. After that year, it started a slow slope down to get to 1.8 in 2015. It is noteworthy that from 1989, the U.S. fertility rate was much above the OECD average, only with countries such as Israel, Mexico and Turkey posting a higher standing – in contrast with the relative performance on the life expectancy front mentioned above.

Naturally, such a pattern of the evolution of the fertility rate is the main explanation behind the evolution of the youth population, that we see in the chart. The rapid decline of the fertility rate in the 1960s and early 1970s has caused the diminution of the population aged less than 15 years between 1965 and 1985 (8 millions fewer in 1985 than in 1965). Then the return of the fertility rate to a relatively higher and stable level generated growth again in the youth population. As a result, in 2010, the share of the youth population in the United States was still just above 20%, three to four percentage points higher than what it was in Japan, Canada and Sweden.

We mentioned the role of immigration in mitigating the aging process in Canada and in Sweden, but not in Japan. What was its magnitude in the United States and has it likely been a significant factor in shaping the aging process? Clearly not to the extent seen in Canada and Sweden, of course in relative terms, as the numbers were much higher [3]. Incoming migrant flows represented annually 0.33% of the total population, on average during the period 1996-2014. This is less than half what was observed in Canada and Sweden. This means that its impact is likely lower than in these countries, while not insignificant (we did not find data on annual flows of immigration by age).

So, compared to the situation in the other countries reviewed, the evolution of the fertility rate was definitely favourable to a slower progress of aging in the United States. We have seen at the beginning of this post that the other major factor, life expectancy, has also clearly had a relative dampening effect on the aging process as longevity grew much more slowly in the United States than in most OECD countries. Let’s review this more closely.

Our review period starts in 1960. In that year, life expectancy at birth was 69.9 years, 3.9 years behind the lead country then, Norway, 3.2 years short of Sweden’s, about one year behind Canada’s, but 2.1 years more than in Japan [4]. As the chart in Paul Krugman’s tweet showed, it is in 1979 that the United States achieved it best position relative to the OECD average – life expectancy at birth was then 73.9, a gain of four years since 1960. It is interesting to measure the pace of progress in life expectancy in days of longer life gained per year. The longer life expectancy in the United States in 1979 compared with that of 1960 is then the result of gaining 77 days of longer life per year, on average. It is easy to interpret this as the additional days of life a new born in a given year can expect to live compared to a baby born in the previous year. An annual gain of 77 days in the United States was a substantial one, narrowing a bit the life expectancy gap with Sweden (gain of 50 days per year), just keeping it with Canada (gain of 77 days per year as well), but far below Japan’s, a country that passed ahead of all OECD countries in this period (with gain of 161 days per year). Thereafter, while life expectancy at birth was still progressing in the United States, it was definitely at a slower pace than in most other OECD countries: reaching 78.8 years in 2015, with an average annual gain down to 50 days. In the same period, all three other countries, while already posting a higher life expectancy than the United States in 1979, continued to experience faster growth: 67 days per year in Sweden, 71 in Canada and 78 in Japan.

As it is suggested in recent reporting of life expectancy in the United States, based on 2015 and 2016, and on early estimates for 2017, there is a possible disconnect between the evolution of life expectancy at birth and life expectancy at 65 years. Circumstances of sudden increase in mortality through violent deaths and opioid overdoses affect people younger than 65 years the most . It possibly reduces life expectancy calculated at birth while not affecting possible increase in longevity for those beyond 65 years. In the last few years of data to 2015, we already see this: life expectancy at birth does not increase while life expectancy at age 65 continues to increase, even if marginally [5].

In all four countries we reviewed so far, young people aged below 15 are more numerous than old people aged 65 and over in the 1950s through the 1990s. But at some point later, through combined lower fertility and greater longevity, the situation reverses. Of the four countries, it is in the United States that this reversal happens last. Actually, it has not even happened yet. It is just projected to happen in a year between 2020 and 2025. In Japan, more older than young people happened in 1998, in Sweden it happened in 2006 and in Canada in 2015.

How all this plays out as a context for broad labour force development?

• The working-age population (15-64 years) as a proportion of the total population reached its highest point in 2007 with a ratio of 67.3%. At its peak, this ratio is more than two percentage points lower than in either Japan and Canada, and just a bit higher than in Sweden. It has declined since 2007 and is projected to decline by slightly more than a percentage point every ten years in this century.
• Just like in Canada, despite the decline in this ratio, the working-age population continues to grow in the United States and is projected to grow for decades to come. Hence the labour force will continue to grow if the participation rate is maintained.
• In 2016, the labour force was at its highest level ever, with an activity rate of 73% [6]. Much time has passed since this rate had reached a peak of 77.4% in 1997 – showing that there may be scope for higher participation ­– but the trends have not been in this direction in recent years.

References

[1] United Nations, Population Division, Department of Economic and Social Affairs, World Population Prospects: The 2017 Revision.

[2] OECD (2018), Fertility rates (indicator). doi: 10.1787/8272fb01-en (Accessed on 25 January 2018).

[3] OECD (2018), Permanent immigrant inflows (indicator). doi: 10.1787/304546b6-en (Accessed on 25 January 2018).

[4] OECD (2018), Life expectancy at birth (indicator). doi: 10.1787/27e0fc9d-en (Accessed on 25 January 2018).

[5] OECD (2018), Life expectancy at 65 (indicator). doi: 10.1787/0e9a3f00-en (Accessed on 25 January 2018).

[6] OECD (2018), Labour force participation rate (indicator). doi: 10.1787/8a801325-en (Accessed on 25 January 2018).

 

---

Non solum data – Data sine monito oculo nihil sunt.

---

Through this blog, we invite constructive comments and constructive article contributions – review our blog policy and have your say!