By Li Liu, Xiaoxue Liu, and Austin Krompasky
Throughout modern history, mortality has steadily decreased. Remarkably, this improvement can be seen for all ages and almost all major causes of deaths. In 2015, however, there was an unusual increase in mortality, which stayed relatively high through 2016.
According to the Centers for Disease Control and Prevention (CDC), the age-adjusted population mortality rate in the United States increased by 1.2 percent from 2014 to 2015. In the United Kingdom during this same time period, it increased by 4.2 percent, according to the Office for National Statistics (ONS). In a significant deviation from the recent trend, both resulted in a decrease in life expectancy. Similar increases in deaths have been reported in many other countries including France, Spain, Denmark, and Switzerland.
This dramatic deviation from the trend line demands further investigation. Is this just a short-term pattern, or is it the beginning of a changing trend? As with most questions related to population mortality, there are multiple potential explanations and perspectives.
Which Causes of Death Increased?
In 2015/2016, there was no calamity such as a natural disaster, pandemic, or terrorist attack big enough to significantly change the population mortality in the Western world. Nonetheless, the increase in mortality can be seen for almost all age groups and major causes of death (see Figure 1).
Some reports blamed the 2015 mortality increase on the flu, especially in the U.K., where the flu vaccine was less effective than in the past. It was reported that this relative ineffectiveness was due to a unique strain of flu, which was widely circulated but was not included in the vaccine.
But flu mortality only explains a small fraction of the overall mortality increase. Flu alone is a small cause of death. Flu and pneumonia together represent 7 percent of total mortality and are responsible for 15.4 percent of the total increased deaths.
In the United States, the 2014-2015 flu was only moderately severe, and flu mortality was even less obvious than in the U.K. to influence the overall mortality.
In the United States, mortality due to drug poisoning—which fall into the category of unintentional deaths—has increased significantly in recent years. From 2010 to 2015, drug poisoning mortality increased by 33 percent and today is higher than road accident mortality.
Despite this worrying trend, it represents less than two percent of total deaths in the United States. Extra mortality from drug poisoning explained roughly 6.6 percent of the total increase.
In the U.K., the drug poisoning mortality rate is less than half of that in the United States, although the U.K. also saw an increasing trend in recent years. Overall, it comprises a smaller portion, roughly 1.2 percent of the 2015 mortality increase.
The largest cause of increase in the U.K. is dementia. According to ONS, from 2013 to 2015, deaths from dementia increased by 24 percent for those ages 75 and older. Dementia diagnosis rates, however, increased just as dramatically. In early 2013, one out of every two dementia cases were properly diagnosed. By early 2016, two out of three cases are properly diagnosed as a result of strong government initiative in recent years to increase dementia diagnosis and awareness.
Due to these improvements in diagnosis rates, much of the increase in dementia mortality is not caused by an increase in prevalence. Had these deaths not been categorized as dementia, they would have been spread among other causes. Nonetheless, as dementia largely affects the elderly, much of the additional mortality in 2015 comes from the older population.
The breakdown of mortality by cause makes it clear that the 2015 mortality increase is unlikely from a random cause-driven event. Rather, it is an overall change in mortality, indicating that drivers more fundamental than randomness may be the culprit. Looking into the drivers of mortality may provide additional insight.
Drivers Changing Mortality
One of the largest challenges facing public health today is obesity. Today, the United States has an adult obesity rate of more than 35 percent; in the U.K, that rate is more than 25 percent. While obesity rates are still climbing, the increase has been less persistent in the U.K., while in the United States, the youth population has shown signs of leveling off.[3, 4]
In order for obesity to be a direct driver of the higher 2015 mortality rate, however, there would have to be a sharp increase in the obesity rate significantly higher than historical increases. Obesity rates in recent years did show increases, but are well within the historic trend.
It should be cautioned that heterogeneity in obesity rates in different age groups may show uneven patterns as younger obese cohorts age. However, it is unlikely that cohort movement would explain a sudden short-term change in mortality felt across age groups.
Developed countries had substantial success in curbing smoking, but the rate of quitting has slowed down. In the United States, smoking prevalence remained relatively stable in the 1990s, although it dropped further afterward. The U.K. also had much slower cessation rates in recent years, compared to the 1970s and 1980s.
As smoking cessation slows down, it gradually becomes a less powerful mortality improvement driver. Smoking cessation, like obesity, has a long-term impact rather than creating a short term swing.
There are many other smaller drivers of mortality such as climate change and antimicrobial resistance. There is also a worrying trend of rising mortality among middle-aged white Americans. Although these could have a material impact on population mortality, they tend to have an impact over the longer term. As a result, they are not adequate to explain the sudden mortality increase.
With these drivers not quite fitting with the 2015 mortality increase, other event-based possibilities were considered. One that stood out was the recovery from the Great Recession, which is event-based and has a systematic impact. We discovered that historically, mortality experienced a brief increase in the few years following a major recession. To understand this better, let’s first look at the mortality after the Great Depression.
Mortality After the Great Depression
The Great Depression remains the longest and the most severe depression experienced in industrialized Western history. In 1927, there was a mild economic recession in the United States followed by a stock market crash in October 1929. Panic in the financial world quickly spread through the entire economy. When the Great Depression reached its nadir in 1933, 13 million to 15 million Americans were unemployed. Impacts spread to Canada and Europe.
Contrary to the intuition that mortality would increase during a depression, population mortality did not see an uptick during the Great Depression. A report by the League of Nations showed no appreciable effect on the aggregate mortality rates of the countries included in the study, which included the United States, the U.K., Germany, Italy, and a few others. This can also be seen from historical life expectancy in Figure 2. Despite the uneven pattern prior to the Great Depression in 1929, life expectancy increased afterward (green box). In 1929, there was a decrease in life expectancy, especially in the U.K., but it was mostly attributed to an influenza epidemic in that year. An exception to this was suicide mortality, which increased significantly.
Surprisingly, mortality experienced a significant increase several years after the Great Depression. In fact, mortality increases that lasted two to four years occurred during the economic recovery. This is especially obvious in 1936 and 1937 (gray box, Figure 2).
Mortality After the Great Recession
History tends to repeat itself. The Great Recession, which began in the late 2000s and lasted into the early 2010s, was the largest and most severe economic crisis experienced by developed countries since the Great Depression. The United States entered the Great Recession in December 2007. Unemployment rate rose to 10 percent and the economy reached its lowest point in June 2009, when the U.S. recession officially ended.
During the Great Recession, the tough economic situation did not significantly impact mortality in the United States. Age-adjusted mortality continued its downward trend across both genders and among most age cohorts (Figure 3).
As recovery began, mortality improvement began to slow. In 2015, there was an increase in mortality followed by a high mortality year in 2016, which was six years after the depth of recession.
Many developed countries also experienced the mortality increase around the same time following their recovery from the Great Recession. The U.K., France, Spain, Denmark, Switzerland, and Italy all saw higher mortality in or around 2015.
Mortality During Economic Recovery
The observation that mortality improves during major economic recessions but then slows or even temporarily stops improvement a few years later is counterintuitive and resulted in a collection of studies.[7, 8]
Mortality increases during economic recoveries are common after the most severe economic depressions or recessions. Table 1 shows “the big five” of the most severe bank-related financial crises since World War II in high-income Organization for Economic Co-operation and Development (OECD) countries and their post-recession mortality year, together with some recent data from the Great Recession.
It should be noted that most populations are not as large as the United States’, so the respective mortality trends have been less stable. This makes a year of mortality increase less rare than in the United States. There are also many factors that could cause increase in mortality. An increase in mortality may not necessarily indicate a direct effect from economic recovery.
Also, not all recessions had an observable post-recession mortality increase. The most severe ones tend to be impactful enough to overshadow random factors. Milder ones may have a mixed impact as other factors dominate.
Nonetheless, it is notable to see a post-recession mortality increase to some degree after all these severe economic downturns.
Why Does a Post-Recession Mortality Increase Happen?
Most of the additional mortality in 2015 occurred at older ages. This is not surprising given this age group has by far the highest mortality rate.
Longevity at older ages is largely supported by paid care and unpaid informal care. Paid care includes care in communities, nursing homes, hospices, and paid home care. Government social care spending adds to paid care. Informal care is typically unpaid care provided by family members, including spouses and adult children.
Informal care plays a critical role in elder care in most societies. In the United States, the annual unpaid adult care is estimated to be worth $470 billion, according to AARP. Meanwhile, health spending on nursing home, home health care, and personal care combined is $409 billion. In the U.K., the estimated total unpaid care is £119 billion per year, while the overall health care spending is £150 billion per year according to ONS. Any material change in informal care will have a significant impact on society.
The U.K. tracks social care spending and informal adult care hours. Since the recession, U.K. government spending on social care has decreased. This decrease adversely affected care for the elderly, which led to potential higher mortality.
On the other hand, during a recession, more people have time for informal care at home. Recessions may also create more part-time jobs that made informal care more feasible. As a result, informal care has increased steadily (see Figure 4). This phenomenon counters the impact from government spending cuts. There hasn’t been any obvious increase in older age mortality until 2015.
As the economy recovers, however, we see the reverse in informal care. People return to work, resulting in a decrease in informal care. Meanwhile, government spending continued to decline. The combined effect created an elderly care shortage that led to the increase in mortality soon after.
In addition, many retirement accounts were adversely affected during the recession. In this climate, informal care should play a larger part than usual in retirement, compounding its impact during economic recovery.
Economic recovery can also cause mortality increase for the working population.[7, 8] Economic expansions are associated with reduction in sleep and increase in alcohol consumption, overtime, and work stress. These factors have adverse impacts on health, especially for those with chronic diseases. Less-experienced new hires may also have higher likelihood of injuries.
Post-recession mortality is often observed after major economic recessions. The 2015 mortality increase is such an example after the most severe systemic economic recession since World War II. There are multiple reasons for this mortality increase. Some, such as flu, are random in nature. Drug poisoning is concerning but remains relatively small. Others, including obesity, smoking cessation, and dementia are unlikely to vary greatly from year to year. These headwinds only partially explain the disruptive mortality increase in 2015.
If history can be a guide, however, this effect will be short-lived. Post-recession mortality following the Great Depression lasted two to four years. We will likely see mortality return to its downward trend soon. When we as actuaries set mortality-related assumptions, we should of course consider the clear increase in 2015, but it alone does not mean a reversal of overall mortality improvement.
We should also identify the headwinds against future mortality improvement. Decreasing benefits from smoking cessation, increasing obesity, and many other factors will weigh down improvement. After the Great Recession, the decline in government social care spending continues to impact mortality improvement. In fact, the worsening mortality in middle-aged white Americans showed alarming signs for certain population segments.
On the other hand, medical advances continue to reveal new and promising fronts. Segments of the population are improving their lifestyle behaviors with increased awareness on health, all facilitated by quickly evolving electronic devices. In a few years, the 2015 mortality increase will likely be seen as a blip on the mortality trend. We will likely discuss less whether mortality will keep improving in the foreseeable future and more about the magnitude of the improvement.
LI LIU, MAAA, FSA, PhD, is a director and actuary at Prudential Financial.
XIAOXUE LIU is a senior associate at Prudential Financial.
AUSTIN KROMPASKY, MAAA, ASA, is a senior actuarial associate at Prudential Financial.
 “Final Flu Vaccine Effectiveness Data Published”; Public Health England; Sept. 11, 2015.
 “Recorded Dementia Diagnoses”; U.K. Health and Social Care Information Centre; 2017.
 The State of Obesity—Better Policies for a Healthier America. Retrieved on Feb. 7, 2017, from stateofobesity.org.
 “Statistics on obesity, physical activity and diet: England, 2015”; Lifestyles Statistics Team, U.K. Health and Social Care Information Centre; 2015.
 “Rising morbidity and mortality in midlife among white non-Hispanic Americans in the 21st century”; Proceedings of the National Academy of Sciences of the United States of America; 2015.
 “The Economic Depression and Public Health Memorandum Prepared by the Health Section” League of Nations; Quarterly Bulletin of the Health Organization; 1932.
 “Increasing mortality during the expansions of the U.S. economy, 1900-1996”; International Journal of Epidemiology; 2005.
 “Life and death during the Great Depression” Proceedings of the National Academy of Sciences of the United States of America; 2009.
 “Episodes of Systemic and Borderline Financial Crises”; World Bank; 2003.
 “National Health Expenditures 2015 Highlights”; Centers for Medicare & Medicaid Services; 2016.
 Valuing Carers 2011: Calculating the value of carers’ support; Centre for International Research on Care, Labour & Equalities; 2011.
 “The 1918 Influenza Epidemic’s Effects on Sex Differentials in Mortality in the United States”; Population and Development Review; 2000.