Crossing the Bridge: Do the Wealthy Live Longer?

A recent study on longevity provides intriguing data on life expectancy (LE) in the United States. Despite the U.S. having the highest health expenditure per capita, life expectancy in the US trails that in most other developed countries.

Life expectancy is a measure of a nation's or community's health that summarizes current mortality statistics by answering the following question: Assuming all current conditions remain unchanged, how long could children born this year be expected to live on average? In 2007, the US ranked 37th in the world in terms of LE at birth, with 75.6 years for men and 80.8 years for women. Across US counties, however, LE ranged from 65.9 to 81.1 years for men and 73.5 to 86.0 years for women. To assess the extent of these disparities, the authors used a benchmark based on ten countries with highest LE in the world. Then they ranked each US county based on how many years it is behind or ahead of the benchmark. For example, if county A has LE of 75 years and it took the benchmark countries years ten years to go from LE of 75 years to the current average of 80 years, then county A is ten years behind the benchmark.

The analysis determined that very few of the US counties are ahead of the benchmark, and most are behind. Some counties are decades behind, ranking close to less developed countries such as Peru and El Salvador. What is perhaps most surprising is that large disparities exist even between neighboring US counties. Take for example two California counties, both in the San Francisco Bay Area: Santa Clara, home to Stanford University, and Alameda, home to UC Berkeley. In 2007, based on LE for men, Santa Clara county was almost a decade ahead of the international benchmark and Alameda county was at least five years behind. An allegory comes to mind: By crossing the Bay Bridge, we jump 15 years back in time! For women, the time travel would be shorter, a decade.

The authors of the cited study are health researchers primarily interested in demographic factors and life style choices that create medically preventable deaths caused by obesity, smoking, and alcohol. Economists have a different interest in these statistics: the link between wealth and health. In 1975, demographer Samuel Preston first reported a positive relationship between GDP per capita and LE. The graphical representation of this relationship is now called the Preston curve. Two properties of the Preston curve are of special interest to policy makers: (1) Life expectancy at birth rises quickly at low levels of per capita income but flattens at high levels of income; (2) The Preston curve shifts upward over time, which is largely explained by improvements in health care technology. The shape of the Preston curve resembles that of a production function, suggesting that health, measured by LE at birth, is a product of a healthcare system where the only input of interest is per capita income.

Some factors that produce health from wealth operate on individual level. Higher income leads to better nutrition, which in turns creates better health outcomes, especially in children. Some operate on the community level (sanitation and other public health measures), and some on the national level (health care system coverage and production of medical knowledge). However, the causality in the Preston curve is unclear, and an alternative explanation is possible: The Preston curve may reflect an impact of health on income. That is, healthier people are able to work more and thus earn more, which enables them to take a better care of their children. Healthier children spend more time studying and thus become more productive workers, etc. This may explain the steeper slope of the Preston curve for the less developed countries where mortality is likely to affect productive members of labor force, while in developed countries, mortality largely affects retirees.

Regardless of the interpretation, the Preston curve remains an empirical observation that holds across countries and suggests that the link between health and income is more important for developing countries than for developed ones. In the case of the United States, does it matter at all? Quite a bit, it turns out. This graph shows a strong relationship between average personal income and LE across California counties. Specifically, average income per capita in Santa Clara county is 16% higher than in Alameda county, $36.5K versus $31.5K. In 2007, LE in Santa Clara county was 80.6 for men and for 83.9 women while in Alameda county it was 77.7 for men and 82.3 for women. So, the Preston curve is relevant even at the county level. Holding all else constant, baby boys and girls born in a relatively wealthier county are expected to live longer.

Discussion questions:

1) Why are researchers from different disciplines interested in life expectancy statistics?

2) What factors might be responsible for the US ranking 37th in the world?

3) What factors could be responsible for the differences in LE in two neighboring US counties with similar demographics and health care systems?