To What Extent is Life Expectancy at Birth a Useful Long-term Indicator of Health and Human Development?

The Growing Role of Health-adjusted Life Expectancy in an Increasingly Long Living World.

43 min readMay 19, 2019

Abstract: Life expectancy at birth is often considered a useful measure of health and human development because of its affiliation with the Human Development Index (HDI). Since the creation of HDI in 1990, life expectancy at birth has succeeded in helping demographers, governments, and developers understand the developmental progress of a country. Interestingly, life expectancy at birth has been increasing globally, and the line between high-income and low-income countries is beginning to blur. Considering this global change in longevity, this paper examines the long-term usefulness of life expectancy at birth to measure a country’s health and human development. The strengths and weakness of life expectancy at birth as an indicator of health and development are examined at the independent level and in the context of the Human Development Index. Additionally, this paper explores the validity of health-adjusted life expectancy at birth as a potential alternative and cooperative measure to life expectancy at birth.

1. Introduction

In 1990, Mahbub ul Haq released the first Human Development Report with an intention to “shift the focus of development economics from national income accounting to people centered policies’’ (Ul Haq, 1995). In many ways, Haq achieved this goal by developing a new measure of development, the Human Development Index (HDI), with economist and philosopher Amartya Sen. This new measure was created as an alternative to Gross Domestic Product (GDP), a largely economic-focused measure that many studies used to determine a country’s developmental progress. Rather than focus only on economics, HDI attempts to take into account human well-being by using Sen’s capabilities approach, which proposed that poverty should not be seen as a lack of income, but rather a lack of capabilities (Sen, 1999). This HDI measure consists of three criteria; “the capability to survive and be healthy, to be knowledgeable, and to enjoy a decent standard of living” (Fukuda-Parr, 2003).

The focus of this dissertation is life expectancy at birth (LEB), a measure used to calculate HDI to understand a population’s health development. LEB is calculated for a country by examining the average number of years a person who is born and resides in that country is expected to live. Because of LEB’s association with HDI, it is often used an indicator of a country’s health development. However, to what extent will life expectancy be a useful indicator of health, and more broadly, human development in the near future? To answer this question, this paper will proceed in the following format. First, a literature review beginning with an overview of LEB in the context of HDI and as an independent measure. The literature review aims to address the strengths and limitations of LEB as a measure of a country’s health development over the next three decades, and examine proposed alternative measures such as health-adjusted life expectancy (HALE). Secondly, this paper will include a methodology section. This section will explain the correlation analysis, the life expectancy projection models, and the index calculations used to produce the results of this paper. The paper will conclude with a results and discussions section that will examine the strength of LEB as a measure of health development, the potential limitations of LEB in the next five decades, and the usefulness of alternative health indictors in assessing and comparing development. Ultimately each section will collectively determine the extent to which life expectancy at birth is a useful long-term indicator of health and human development.

Lastly, it is important to note that it is not the intention of this paper is to find a replacement for LEB as a measure of health and human development, but rather propose additional measures that with GDP and HDI can give governments a better understanding of the development of their country.

2. Lituerature Review

The primary focus of this section is to use existing literature on life expectancy at birth (LEB) and the Human Development Index (HDI) to identify potential strengths and limitations of LEB as a useful long-term indicator of health and human development for a country. These strengths and limitations will be tested using the methodology described in section 3.

2.1 Life Expectancy at Birth and Development

LEB is one of the most frequently used health status indicators, primarily due to its relationship with numerous signifiers of health development, such as increasing standard of living, improved education, greater access to health care, and reductions in infant mortality (OECD, 2018). LEB is calculated using available population and mortality data, is derived from life tables, and is based on age- and sex-specific death rates (WHO, 2015). LEB is often considered the most important measures of health for two main reasons: its relationship with other more complex health indicators and its ease of calculation due to the relatively widespread availability of mortality data (Khalsa, 2011).

In the context of HDI, LEB was selected as a measure of longevity simply because of the “common belief that a long life is valuable in itself and in the fact that various indirect benefits (such as adequate nutrition and good health) are closely associated with higher life expectancy” (UNDP, 1990). The 1990 Human Development Report adds that it is the association between LEB and other health measures that makes LEB a valuable indicator of human development (UNDP, 1990). The HDR report elaborates that the “present lack of comprehensive information about people’s health and nutritional status” makes LEB especially valuable due to its ease of calculation and availability as compared to other indicators of health (UNDP, 1990). Nearly three decades after the publication of the 1990 HDR, is LEB still the best measure for assessing a country’s health and human development? And is LEB still strongly correlated with other health indicators? This paper will explore this further in the results and discussions section.

2.2 Life Expectancy at Birth and Narrowing Goalposts

As stated earlier, HDI consists of three criteria: “the capability to survive and be healthy, to be knowledgeable, and to enjoy a decent standard of living” (Fukuda-Parr, 2003). Each of these criteria have an indicator and a dimension index calculated from that indicator (UNDP, 2016). As outlined by the United Nations Development Programme (UNDP), the indicator for “the capability to survive and be healthy” is life expectancy at birth (LEB), and the dimensional index is the life expectancy index (LEI) (UNDP, 2016). LEI is calculated using the LEB of a population and a minimum and maximum life expectancy at birth value known as “goalposts.” The minimum value acts as a natural zero, while the maximum value serves as an aspirational target. In the case of LEI, the natural zero, or minimum, is 20 years, “based on historical evidence that no country in the 20th century had a life expectancy of less than 20 years” (Maddison, 2010) (Oeppen and Vaupel, 2002) (Riley, 2005) (UNDP, 2016). The aspirational target, or maximum, is the highest expected life expectancy at birth for any country over the next three decades, which at the writing of this paper (2018) is 85 years (UNDP, 2016) (Ul Haq, 2003). Calculating LEI follows this formula and produces a value between 0 and 1:

More complete details on the calculation of LEI and HDI can be found in the methodology section 3.3.

At the inception of HDI, these “goalposts” would change each year depending on best and worst observed LEB values in the world. This methodology was changed, as Haq describes, to be the most extreme values over the previous three decades or expected over the next three decades (Ul Haq, 2003). This change was made to allow for a more “meaningful comparison of countries’ performances for over 60 years (Ul Haq, 2003). In the 2014 HDR, fixed “goalposts” were adopted (20 years for minimum and 85 years for maximum) with the intention of making HDI a more comparable measure that demonstrated a countries progress (UNDP, 2015). Considering that LEB is increasing globally, these “goalposts” may inevitably be adjusted to represent an updated natural zero and aspirational target for LEB. The question remains, how will the inevitable narrowing of these “goalposts” affect LEI, and therefore HDI. Theoretically speaking, if this range were to narrow over the next three decades small differences in LEB among countries will result in a much more dramatic difference in both LEI and HDI. LEB data recorded by the data visualization group, “Our World in Data”, demonstrates a narrowing LEB that by 2012 decreased to a range of 45 to 85 years (Roser, 2018). If the countries on the lower end are able to increase their LEB at a faster rate than those countries on the higher end, it is expected that this range will continue to narrow and significantly effect LEI and HDI calculations. In order to test the effects on narrowing LEB ranges on LEI and HDI calculations, this paper will project future LEB values in the methodology section and determine the long-term implications of this narrowing range in the results and discussions sections of this paper.

2.3 Life Expectancy at Birth as a Measure of Longevity and Good Health

According to the 2016 Human Development Report (HDR) technical notes, LEB is considered an indicator of living a long and healthy life (UNDP, 2016). By classifying an increase in LEB as an indicator of both longevity and good health, UNDP is making the assumption that an increase in LEB is correlated with a decrease in morbidity. In other words, UNDP is proposing that LEB indicates that a population is not only living longer, but is also spending additional years gained in good health as opposed to prolonged illness or disability. Interest in the relationship between mortality and morbidity has grown in the past few decades due to a recently observed decline in mortality in the elderly population (Robine & Ritchie, 1991).

Numerous theories about the relationship between mortality and morbidity have been debated in a bid to clarify whether living longer also means living additional years in good health. One theory argues that an increase in health care access and better medical treatment has resulted in the prevention of debilitating diseases, therefore resulting in the postponing of death from these diseases (Fries, 1989). This theory is referred to as the “compression of morbidity” and proposes that people will spend less time in poor health due to the postponement of illness and the inevitable plateauing of life expectancy (Robine & Ritchie, 1991) (Fries, 1989). This theory supports the assumption that an increase in LEB indicates both longevity and additional years of good health. The morbidity compression theory has gained increased attention over the past few years, particularly in the context of demographic aging (Geyer, 2016). Numerous studies have presented evidence for the morbidity compression theory, and it has gained widespread traction (Mor, 2005). There is, nevertheless, some opposition.

An opposing theory suggests that decreasing mortality will increase morbidity, stating that developments in healthcare have allowed people to live longer in a state of poor health (Robine & Ritchie, 1991) (Fries, 1989) (Gruenberg 1977). This theory contradicts the assumption made about LEB and suggests that people who are living longer may not be spending additional years in good health. Some studies have even demonstrated that in some populations morbidity is expanding rather than compressing due to an increase in disease prevalence (Crimmins & Beltrán-Sánchez, 2011). Although the LEB seems to generally be correlated with increased good health as suggested by the morbidity compression hypothesis, this opposing theory and evidence suggest that this is not always the case.

Overall, it appears that LEB may not always be a good predictor of “living a healthy life” as suggested by UNDP in the HDR. For this reason, alternative measures have been suggested to allow for a better understanding of a population’s health and human development. One of the first alternative measures discussed was disability-free life expectancy (DFLE), sometimes referred to as Sullivan’s Index (Sullivan, 1971). DFLE is calculated using life expectancy and disability data to calculate a value that demonstrates how many years a person in a population is expected to live free of disability (Jagger et al., 2001) (Sullivan, 1971). One limitation of DFLE is that it gives “zero weight to years lived in less than full health” (Mathers et al., 2003). For this reason, a slightly adjusted measure was created that “counts all years of life, but with a weight that varies with how ill or disabled an individual is in each year” (Wolfson, 1996). This measure is known as health-adjusted life expectancy (HALE) at birth and is defined by the WHO as the “average number of years that a person can expect to live in “full health” by taking into account years lived in less than full health due to disease and/or injury” (WHO, 2016). HALE is calculated using Sullivan’s method and uses data from country life tables and severity-weighted prevalence estimates (Mathers et al., 2003). The severity-weighted prevalence estimates are adjusted for comorbidity for each age group in a given population in order to estimate the number of years lived in full health by a given age cohort in the life table (Mathers et al., 2003).

In 1996, Michael Wolfson, the Director General of the Institution and Social Statistics Branch of Canada, referred to HALE as a good indicator of health that provides a more direct way to measure the quality of life, in addition to the quantity of life, as LEB measures (Wolfson, 1996). On its own, HALE is a great comparable measure that can demonstrate in which countries people are expected to live in good health the longest. Additionally, when comparing the difference between HALE at birth to LEB within a country, the two measures together become a useful indicator of the burden of ill health and serve as a useful way of testing for the compression of morbidity (Wolfson, 1996). The WHO has also seen value in HALE and has published country and region-specific HALE data since the year 2000 (WHO, 2018a). In 2015, the WHO estimated a global HALE at birth of 63.1 years, 8.3 years lower than global LEB (WHO, 2018a). This statistic suggests that, on average, poor health results in a loss of 8.3 years of healthy life (WHO, 2018a). As the value of HALE continues to be demonstrated, the question remains: Will HALE at birth be a better indicator of “living a long and healthy life” than LEB? And if so, how would incorporating HALE at birth in HDI calculations change the way the international development community thinks about a country’s health and human development? This paper aims to further investigate these questions in the results and discussion section.

3. Methodology

A diverse set of methods is used in this paper to answer the proposed research question. This section of the paper will discuss the various methods used in independent subsections. The first subsection (3.1) will explain the correlation test methodology. The second subsection (3.2) will explain the LEB country projection methodology. The third subsection (3.3) will explain the ranking analysis and the Health-adjusted Human Development Index (HHDI) calculations. Finally, the fourth subsection (3.4) will discuss limitations in the methods used in this paper.

3.1 Life Expectancy at Birth and Health-adjusted Life Expectancy Correlation Methods

In section 2.1, this paper discussed the value of LEB as a measure of health and human development as an independent measure and in the context of HDI. A major value attributed to LEB was its strong correlation with other health indicators. In the nearly three decades after the publication of the first HDR in 1990, few correlation analysis studies have been conducted to determine how correlated LEB is to other health indicators. Additionally, no study has been conducted on the extent to which alternative measures to LEB, such as HALE at birth, are correlated with other health indicators. This section will describe the methodology used to test the correlation of LEB and HALE at birth with other health indicators.

First, health indicators were selected based primarily on which health indicators were more commonly used, and the availability of data for each indicator. Ultimately six health indicators were selected for correlation analysis. These indicators are infant mortality rate (per 1,000 live births), maternal mortality ratio (modeled estimate, per 100,000 live births), neonatal mortality rate (per 1,000 live births), under-5 mortality rate (per 1,000 live births), prevalence of anemia among children (% of children under 5), and prevalence of undernourishment (% of population).

Next, data was collected for both LEB and HALE at birth, as well as for the six selected health indicators. LEB and HALE data was collected from the WHO Global Health Observatory data repository. LEB data is available from the years 2000–2016 on both global, regional and country level. HALE data is a bit more limited and is only available for the years 2016, 2015, 2010, 2005 and 2000. For this reason, only data from these five years was collected for the other health indicators to allow for regression analysis. The data for the six health indicators is collected from a variety of organizations from the World Bank data compiler website (data.worldbank.org). Estimates of infant mortality rate (per 1,000 live births), neonatal mortality rate (per 1,000 live births), and under-5 mortality rate (per 1,000 live births) is developed by the UN Inter-agency Group for Child Mortality Estimation (UNICEF, WHO, World Bank, UN DESA Population Division) at childmortality.org. Estimates for maternal mortality ratio (modeled estimate, per 100,000 live births) comes from WHO, UNICEF, UNFPA, World Bank Group, and the United Nations Population Division. Estimates for the prevalence of anemia among children (% of children under 5) originates from the World Health Organization Global Health Observatory Data Repository. Lastly, estimates for the prevalence of undernourishment (% of population) is compiled from the Food and Agriculture Organization. This data was collected at the global level, as well as for three representative counties from each of the four income categories. This paper uses the updated 2018 classifications by income specified by the World Bank. GNI per capita estimates are also taken from the World Bank data page. The countries selected were chosen primarily due to the availability of data, representation of a region, and diversity of income level. Switzerland, the United States, and Japan were selected to represent high-income countries (GNI/capita greater than USD 12,055) (World Bank, 2018e). Costa Rica, China, and South Africa were selected to represent upper-middle income countries (GNI/capita between USD 3896–12,055) (World Bank, 2018e). Indonesia, India, and Kenya were selected to represent lower-middle income countries (GNI/capita between USD 996–3895) (World Bank, 2018e). Finally, Zimbabwe, Afghanistan, and Burundi were selected to represent low-income countries (GNI/capita less than USD 995) (World Bank, 2018e). The raw data collected for this correlation analysis can be found in Appendix 1.

In order to determine how correlated LEB and HALE are with other health indicators, the Pearson correlation coefficient (r) is calculated for each relationship. The Pearson correlation coefficient (r) is the “measure of the strength of the linear relationship between two variables” (Lane, 2018). The value of this coefficient ranges from -1 to 1, where a value close to -1 represents a more correlated negative relationship between two variables, a value close to 1 represent a more correlated positive relationship between two variable, and a value close to zero indicates a less correlated relationship between the two variables. This correlation value, r, was calculated for the relationship of each of the six health indicators with both LEB and HALE. These values were calculated using Microsoft Excel Regression Data Analysis function which uses the following formula to calculate r.

This Microsoft Excel Regression Data Analysis also calculates a p-value that denotes if the r-value calculated is statistically significant with a 95% confidence level.

3.2 Life Expectancy at Birth Projection Methods

In section 2.2, this paper discussed narrowing “goalposts”: the idea that in the near future the minimum and maximum values of LEB will move closer together and require demographers, developers, and policy makers to re-think the calculation of the life expectancy index (LEI) and the human development index (HDI). In order to test how LEB values would change over the next five decades, this paper ran LEB projections till the year 2070.

The projections in this paper are created using the demography module (DemProj) in Spectrum v5.7. Spectrum is a program developed by Avenir Health and is a “suite of easy to use policy models which provide policy makers with an analytical tool to support the decision-making process” (Avenir Health, 2018). The DemProj module in Spectrum uses the cohort-component method to create future projections. This method applies available data on births, deaths, fertility, and migration to existing measures of population age groups to create future population projections (Measure Evaluation, 2018). More specifically, the DemProj module applies age-specific fertility rate to five-year age cohorts for a given population to estimate how many infants are expected to be born. Next, infant mortality rates are applied to the calculated infant birth estimate to determine how many infants will survive. Next, future age cohorts are estimated using adult mortality data for each five-year age cohort. Finally, international migration data is applied to determine how many people from each five-year age cohort will enter or leave the country.

In order to estimate the narrowing of “goalposts” for LEB, projections are conducted for the current highest-ranking counties in LEB and the current lowest-ranking countries in LEB. The top ten and bottom ten countries were selected for this projection based on 2016 LEB data collected by the United Nations Population Division. The ten highest-ranking countries, ranging from LEB values from 82 to 84 years in 2016, are Japan, Hong Kong (SAR, China), Macao (SAR, China), Switzerland, Spain, Singapore, Italy, Norway, Australia, and Iceland (World Bank, 2016). The ten lowest-ranking countries, ranging from LEB values from 52 to 57 years in 2016, are Sierra Leone, Central African Republic, Chad, Nigeria, Cote d’Ivoire, Lesotho, Somalia, South Sudan, Guinea-Bissau, and Burundi (World Bank, 2016).

A base year of 1990 was selected, and LEB values are projected till the year 2070 for all projections to allow for a three-decade comparison from the year 2020, 2030 and 2040. No model life tables were used for any projection, and projected data was compared to United Nations Population Division estimates to ensure the accuracy of projected values. All data used in the projections is sourced from the 2017 World Population Prospects by the United Nations Population Division. Life expectancy at birth data for each country is based on official estimates of life expectancy available through 2015. The age pattern of mortality is based on life tables through 2013 from the Human Mortality Database.

Once the narrowing of “goalposts” for LEB was determined using the projected values, the goal was to use these projected values to better understand the implications of this narrowing on the calculation of the life expectancy index (LEI) and the Human Development Index (HDI). Three years were selected: 2020, 2030, and 2040. For each of these three years, the LEB projection for the highest and lowest ranked countries is checked in order to find the minimum LEB value in past three decades (rounded down to nearest whole number) and maximum LEB value in next three decades (rounded up to nearest whole number). As mentioned in section 2.2, the minimum and maximum values of LEB within a three-decade range have been historically used to calculate LEI. The equation for LEI is as follows:

For this analysis, the minimum and maximum values are taken from the projections, and the actual value is substituted with a theoretical LEB value between 50 and 85 years. LEI values were calculated for 2020, 2030, and 2040 for each of the theoretical LEB values. Additionally, these values were compared to LEI calculations using the currently accepted minimum and maximum values of 20 years and 85 years respectively. This paper will analyze this comparison of LEI values in the results and discussion section.

3.3 Health-adjusted Life Expectancy Index and Health-adjusted Human Development Index

In section 2.3, this paper proposed the idea of health-adjusted life expectancy (HALE) at birth being an alternative measure to life expectancy at birth (LEB). In order to test the validity of HALE at birth as a measure of health and human development, two country ranking analyses are conducted. For both of these ranking analyses, the countries Hong Kong (SAR, China), Liechtenstein, Andorra, Palau, Sint Kitts and Nevis, Dominica, Palestine, and Swaziland are not included due to lack of available HALE data for these countries in the WHO Global Health Observatory data repository.

The first ranking analysis aimed to compare LEB with HALE at birth directly. Country-level data on LEB and HALE at birth was collected for the year 2016 from the WHO Global Health Observatory data repository. Countries were ranked according to LEB and HALE at birth respectively. Next, the countries’ rankings for LEB were placed alongside the countries’ rankings for HALE at birth. Lastly, a ranking analysis was conducted, by finding the difference between a countries HALE at birth rank and its LEB rank, to examine how a country’s rank had changed when ordering countries by HALE at birth as compared to LEB.

The second ranking analysis aimed to compare the effect of LEB and HALE at birth on the human development index (HDI). First, 2015 country level data was collected for each of the four components required to calculate HDI. These four components are life expectancy at birth (LEB), expected years of schooling, mean years of schooling, and gross national income (GNI) per capita. All the data required for this calculation was available on the UNDP Human Development Report Table on Human Development Index and its components. LEB data is sourced from the United Nations Department of Economic and Social Affairs (UNDESA) 2015 World Populations Prospects (WPP). Expected years of schooling and mean years of schooling data are sourced from the United Nations Educational Science and Cultural Organization (UNESCO) Institute of Statistics, the ICF Macro Demographic Health Surveys, the UNICEF Multiple Indicator Cluster Surveys, and a dataset for educational attainment. Finally, GNI per capita data is sourced from the World Bank, the IMF, and the UN Statistical Division.

Next, dimension indexes are calculated using the protocols highlighted in the 2016 HDR technical notes. The general formula for each dimension index is the same formula described and used in section 2.2 and 3.2:

For the life expectancy index (LEI), a maximum and minimum values of 85 and 20 years are used respectively. For the expected years of schooling index, a maximum and minimum values of 18 and 0 years are used respectively. For the mean years of schooling index, a maximum and minimum value of 15 and 0 are used respectively. Lastly, for the income index, a maximum and minimum value of USD 75,000 and USD 100 are used respectively. The arithmetic mean of the expected years of schooling index and the mean years of schooling index is calculated to determine the education index. Additionally, “because each dimension index is a proxy for capabilities in the corresponding dimension, the transformation function from income to capabilities is likely to be concave — that is, each additional dollar of income has a smaller effect on expanding capabilities” (UNDP, 2016) (Anand and Sen, 2000). For this reason, when calculating the income index, the natural logarithm of the actual, minimum, and maximum value is used. Furthermore, if the actual value in the calculation of a dimension index is greater than the maximum value, the index is given a value of 1, as the calculation of HDI does not allow for any dimension index to have a value greater than 1. HDI is calculated by taking the cube root of the product of the three dimension indexes (life expectancy index, education index, income index) (UNDP, 2016). The calculated HDI values are compared to the 2015 HDI values listed on the UNDP Human Development Report Table on Human Development Index and its components to ensure accuracy. An example calculation of HDI can be seen in Figure 1 below:

**Figure 1: HDI example calculation.** Source: (UNDP, 2016)

In order to test the effect of HALE at birth on the HDI, a new dimension index was calculated using 2015 HALE at birth data from the WHO Global Health Observatory data repository. This paper will refer to this new dimension index as health-adjusted life expectancy index (HALEI). HALEI was calculated using the standard dimension index equation with a maximum and minimum value of 77 and 12 years respectively. The minimum and maximum value are determined by reducing the minimum and maximum values used to calculate LEI by eight years. An adjustment of eight years was selected because the 2015 global average of HALE at birth is about eight years less than that of LEB (WHO, 2018a). This new HALEI is then used in the HDI equation, replacing the LEI value. This paper will refer to this modified HDI as Health-adjusted Human Development Index (HHDI). An example calculation of HHDI can be seen in Figure 2 below:

**Figure 2: HHDI example calculation.** Modified HDI calculation using HALE at birth instead of LEB.

To see the effect of HALE at birth on HDI countries are ranked according to HDI and HHDI. Next, the countries’ rankings for HDI are placed alongside the countries’ rankings for HHDI. Lastly, a ranking analysis was conducted, by finding the difference between a countries HHDI rank and its HDI rank, to examine how a country’s rank had changed when ordering countries by HHDI as compared to HDI.

3.4 Limitations

Before discussing the results of the methods described above, this paper will explain some notable limitations in the methodology used to create these results. In the methodology section 3.1, the calculation of the correlation value (r) was outlined. An important factor to note about r is that the calculation assumes a linear relationship between two variables being tested. For all correlation tests, values were graphed to ensure a linear relationship, but because only five data points were used it is possible that the relationship between some of these health indicators may not be best described as linear. Unfortunately, the correlation test was limited by the HALE at birth data that was only available for five different years. Future correlation analysis should aim to use more data points to ensure more meaningful r-values. Additionally, HALE at birth data was not available for each of the four income categories (high-income, upper-middle income, lower-middle income, and low-income). For this reason, the correlation test could only be applied to individual countries with the data that was available. Although three countries have been selected to represent each income category, readers should not assume that r-values calculated for these countries are indicative of other countries in its category. The purpose of the country-specific data is to gain a better understanding of the relationship LEB and HALE at birth have with various health indicators in individual countries as compared to global level data.

In section 3.2, this paper discussed LEB projections created using the projection tool Spectrum. The limitations of these projections are not unique to the LEB projections showcased in this paper but apply to all population projections. When creating a projection for a population, demographers must operate under defined assumptions. These assumptions are usually based on existing population data, available literature, and past population trend. Although the assumptions for the LEB projection are outlined in section 3.2, it is important to mention that these projections could result in different values if operating under alternative assumptions. The assumptions made in this paper are subject to individual bias and should not be viewed as guaranteed future events. These projected LEB values should be viewed as a plausible occurrence, assuming certain variables remain constant or change predictably as they have in the past. Additionally, unforeseen future events (such as new policy, natural disasters, an outbreak of disease, etc.) could have a dramatic effect on the future LEB of all the projected counties. Because of the unpredictable nature of these events, it is beyond the scope of this paper to account for unforeseen circumstances when creating these LEB projections.

In section 3.3, this paper created a new dimension index, HALEI, and a modified version of HDI, HHDI. The creation of these indices is unprecedented and are not endorsed measure by the UNDP. The purpose of these new indices is to examine the potential use of HALE at birth as an alternative to LEB in understanding the health and human development of a country. Readers should focus on the comparison between HDI and HHDI as opposed to interpreting HHDI as an independent measure.

4. Results and Discussion

In this section, this paper aims to showcase the results obtained from the methodology described in the previous section. This results and discussion section will be separated into three subsections that match the subsections in the literature review and methodology sections. The first subsection (4.1) will showcase the results from the correlation methodology (described in section 3.1) and discuss the implications of these results on the value of LEB and HALE at birth as an indicator of health and human development. The second subsection (4.2) will showcase the results from the LEB projection and LEI calculation methodology (described in section 3.2) and discuss the future implications of narrowing goalposts on the LEI values. The third subsection (4.3) will showcase the results from the LEB, HALE, HALEI, and HHDI calculations and ranking analysis methodology (described in section 3.3) and discuss the potential value of HALE at birth as an alternative measure of health and human development.

4.1 Correlation Results and Discussion: The Value of LEB and HALE as Indicators of Health and Human Development

Table 1 below indicates correlation values (r) and statistical significance (p) for LEB and HALE at birth with each of the six health indicators tested at the global level. Correlation value tables for the 12 tested countries and raw data used to calculate all r-values and p-values can be found in Appendix 1.

**Table 1: Correlation values (r) and statistical significance (p) for LEB and HALE at birth for six health indicators.** Data sources: (WHO, 2018b) (World Bank, 2018a)(World Bank, 2018b)(World Bank, 2018c)(World Bank, 2018d)(World Bank, 2018f)(World Bank, 2018g).

When examining the r-values depicting the correlation between LEB and each of the six health indicators (rLEB), it is evident that LEB is very strongly correlated to these six health indicators. All rLEB values fall between -0.97 and -1, indicating that LEB is strongly correlated with each of the health indicators. All the r-values are negative, indicating a negative correlation. This negative correlation is logical, considering that any reduction in mortality, undernourishment, or anemia should theoretically result in an increase in expected LEB. The same is true for the correlation between HALE at birth and each of the six health indicators (rHALE). In this case a negative correlation is also logical. The rHALE values range from -0.97 and -1 as well, indicating a strong negative correlation. Both rLEB and rHALE values are supported by p-values less than 0.05, indicating statistical significance with a greater than 95% confidence level. Interestingly, the rHALE values are all slightly closer to -1 for each health indicator, suggesting that HALE at birth is slightly more correlated to each indicator than LEB on a global level.

Looking at the individual country data (Appendix 1), it is evident that the correlation between these variables at the global level is generally the same for that at the country level. Despite this, there are some notable counter examples in the data that should be discussed. For a majority of the countries examined, prevalence of anemia among children was the least correlated to either LEB or HALE at birth. A majority of countries tested had fluctuating values of anemia data that prevented a strong correlation with either indicator. In some cases, positive correlations were detected between variables. These positive correlations are due predominately to a lack of data points and should not be considered as significant relationships. Overall, the country-specific r-values do not showcase much meaningful understanding of correlation due to a lack of data points. What these country-specific r-values do demonstrate is the vast differences in the meaningfulness of LEB and HALE at birth as health indicators depending on a country’s stage of development. It is important to note that the stage that a country is in has a significant implication on the correlation between variables. Countries that are low to lower-middle income tend to have more correlation between variables due to higher mortality, undernourishment, and anemia values that tend to progress in a more significant and predictable manner than high and higher-middle income countries. Countries that are considered more developed tend to have lower mortality, undernourishment, and anemia values that fluctuate more on a yearly basis and decline at a slower and less predictable rate. This suggests that LEB and HALE at birth data may be less predictive of other health indicators in more developed countries. For this reason, LEB and HALE at birth should be considered as composite scores that allow for a broad understanding of a country’s health development as opposed to a detailed and narrow understanding.

Overall the correlation data, specifically at the global level, suggests that LEB is a good indicator of the health development of a country. As mentioned in the UNDP 1990 HDR, LEB is a convenient measure of health development because of its availability and relationship with other health indicators (UNDP, 1990). It is ultimately the correlation between LEB and other health indicators that make it a valuable measure of health and human development, as opposed to the common belief that living longer is valuable in of itself. Additionally, HALE at birth appears to also be a good indicator of health development, in all cases showing a slightly stronger correlation with the six tested health indicators than LEB at the global level. The correlation data supports the notion that HALE at birth may be a potential alternative measure to LEB.

Although HALE at birth appears to be as good a health indicator as LEB, it is important to note that, in terms of data availability, LEB values are more easily determined for various countries. HALE at birth calculations require disability and illness data that may not be well documented in every country. The requirement of this additional information makes HALE a less convenient proxy for health development than LEB. As countries continue to develop and gain more sophisticated monitoring systems, HALE at birth has the potential to be easily calculated. Ultimately, better availability of disability and illness data will allow HALE at birth to be utilized as an effective alternative to LEB as a measure of health and human development.

4.2 LEB Projections and LEI Calculations Results and Discussion: The Implication of Narrowing LEB Goalposts

In section 2.2, this paper discussed the possibility of narrowing goalposts for LEB. This theory expressed that advancements in health and technology would allow people to live longer globally, thus decreasing the difference in LEB between the highest and lowest ranking countries. As mentioned earlier, the calculation of the LEI, and therefor HDI, has historically relied on the minimum LEB value recorded in the past three decades and the maximum LEB value expected in the next three decades. For this reason, LEB projections were created till the year 2070 in order to understand the effect changes in the minimum and maximum LEB values could have on future LEB calculations.

Figure 3 below depicts projected LEB values from 1990 to 2070 for the ten highest ranking and ten lowest ranking countries in LEB in the year 2016. The higher ranking countries are depicted in a variety of brighter colors and the lower ranking countries are depicted in a variety of muted colors. The dashed purple lines in the figure indicate the 85 years and 20 years LEB goalposts used by the UNDP to calculate LEI. The raw projection data used to create Figure 3 can be found in Appendix 2. The projected data below visually demonstrates the narrowing on LEB. The ten lowest ranking countries in LEB increase at a faster rate than the ten highest ranking countries in LEB. In 1990 the LEB values for these 20 countries ranged from 37.9 years (Sierra Leone) to 78.8 years (Japan). The minimum and maximum LEB values in 1990 are separated by 40.9 years. In 2018, (the current year at the writing of this paper) LEB values range from 52.9 year (Sierra Leone) to 84.4 years (Hong Kong SAR, China), a difference of 31.5 years. In 2070, the projected LEB values range from 62.7 years (Sierra Leone) to 88.5 years (Hong Kong SAR, China and Macao SAR, China), a difference of only 25.8 years.

**Figure 3: Life expectancy at birth projections 1970–2070.** Projections created using DemProj in Spectrum v. 5.7. Data sourced from 2017 World Population Prospects by the UN Population Division. Life expectancy at birth data for each country is based on official estimates available through 2015. The age pattern of mortality is based on life tables through 2013 from the Human Mortality Database.

The declining difference between the maximum and minimum LEB values appears to be a consistent trend. As these values continue to converge the independent value of LEB to measure the health development of a country becomes more limited. This is especially true when attempting to compare countries that have the same, or very similar, LEB values. For the projected 2070 LEB values, the top ten ranking countries range in LEB from 86.6 years to 88.5 years, while the bottom ranking countries range in LEB from 62.7 years to 68 years. Considering this small range of values, specifically for the top-ranking countries, it may be difficult to compare the health development of countries on LEB alone in the near future. An additional measure will be necessary to gain a fuller understanding of the developmental progress for these plateauing LEB countries. This paper will discuss the potential of HALE at birth becoming this additional measure in the next subsection (4.3).

Additionally, this paper aims to discover the effect of converging LEB values on the future calculation of LEI, a main component in the calculation of HDI. As mentioned previously, the maximum and minimum values used in the calculation of HDI have historically come from the lowest recorded value in the last three decades and the highest expected value in the next three decades. Table 2 showcases the minimum and maximum values revealed by the projections for the years 2020, 2030 and 2040. Table 2 also shows calculated LEI values for each of these years and the current year (2018) for LEB ranging from 50 years to 85 years. The results from the LEI calculations are striking, demonstrating significant changes in LEI values in short periods of ten years. LEI values change significantly even when only adjusting goalposts slightly. The dramatic change in LEI values as a result of goalpost adjustment can best be demonstrated by looking at an example country. Sierra Leone, with a current LEB of 53 year has a LEI values of 0.508 if calculated using current goalposts. In 2040, if the minimum and maximum values are adjusted as in the Table 2, Sierra Leone (which is projected to have a LEB of about 60 years) will have a LEI of 0.293. This is a significant drop in LEI despite a continued increase in LEB.

**Table 2: LEB and HALE at birth ranking analysis.** LEB and HALE at birth values are from 2016. Data source: (WHO, 2018b)

This LEI and goalpost analysis demonstrates an impending problem that UNDP will have to address in the near future. As LEB continues to rise globally, UNDP will have to make a decision on how to adjust the minimum and maximum values used to calculate LEI and ultimately HDI. Adopting the previous method, which requires the use of recorded LEB values for the previous three decades and projected LEB values over the next three decades, does not appear to be a valid solution. A rapid increase in LEB will cause the goalposts to shift dramatically and ultimately diminish the usability of LEI and HDI to assess human development. The alternative possibility, of keeping the goalposts fixed at 20 years and 85 years, also imposes an impending problem. According to the LEB projection data showcased earlier in this subsection, Hong Kong (SAR, China) will pass the current maximum of 85 years in 2023, only five years from the current year (2018). Additionally, all of the current top ten LEB ranking countries are projected to pass a LEB of 85 years by the year 2036.

This means that after the year 2036, each of these countries will have a calculated LEI over 1 if the goalposts are not adjusted. Because UNDP does not allow any dimension index to have a value over 1 for the calculation of HDI, all of these values will be rounded down to 1 and HDI will function as a less meaningful indicator and comparison tool for these top ranking countries. At this point, UNDP will have two potential options. Their first option is to change the goalposts, more specifically the maximum value, to better reflect the inevitable progression of LEB beyond 85 years in numerous countries. If UNDP decides to do this, they should slowly increase the maximum value, and potentially the minimum value, for the LEI calculation. By increasing these goalposts slowly, UNDP will prevent drastic changes in HDI values and will make HDI values more comparable with previous HDI data before the shifting of the goalposts. The second option, assuming UNDP decides to keep the goalposts fixed at 20 years and 85 years, is to create a modified health and human development index that allows for the comparability of countries that are either above the maximum threshold or too similar in value. Potential modified indices that use HALE at birth instead of LEB will be discussed further in the next subsection (4.3).

4.3 Country Ranking Comparison Results and Discussion: The Value of HALE as an Alternative Measure of Health and Human Development

In the previous two subsections this paper has discussed, in specific situations there is a need for an alternative measure to LEB is to better understand both the health and human development of a country. It has been proposed that HALE at birth is a logical alternative because it measures not only the quantity of life but also the quality of life. The goal of this section is to understand the implications and practicality of using HALE at birth as an independent measure and in the context of HDI.

First, in order to understand the value of HALE at birth as an independent measure of health development, 183 UN recognized countries were ranked by LEB and HALE at birth. The results of this ranking analysis are displayed in Table 3. The country names written in blue in the table indicate countries that have a considerable increase in ranking (10 or more places) from LEB to HALE at birth, while the country names written in purple indicate countries that have a considerable decrease in ranking (10 or more places) from LEB to HALE at birth. All HALE at birth and LEB values come from the same year (2016). At first glance, the information portrayed in this figure is striking, considering that almost every country changes its ranking to some degree. At the top of the figure, two notable examples of changes in rank are Singapore and Cyprus. Singapore, tied for fourth overall in LEB with a value of 82.9 years, jumps to first overall in HALE at birth ranking, with a value of 76.2 years. Although a jump of three places does not seem dramatic, the story that these values tell about the country is. According to LEB rankings, Singapore is a country with high health development, with its citizens expected to live well over 80 years. According to the HALE at birth rankings, Singapore should be considered a country with the highest health development, reducing disability and illness more than any other country (with citizen living in full health for an average of 76 years). The other notable example at the top of the figure is Cyprus. Cyprus has the most dramatic increase in ranking from LEB to HALE at birth. With a LEB value of 80.7 years and a HALE at birth value of 73.3 years, Cyprus increase its rank from 28th to 6th respectively. This increase in rank of 22 places is staggering and changes the perception of Cyprus from a country with impressive health development, to a top ten country pushing the boundaries of health for its citizens. Other notable examples, such as Antigua and Barbuda, increase their rank by over 20 places, or Jordan and China, that increase their rank by 17 and 13 places respectively. On the other end of the spectrum, we can also see countries that drop in rank dramatically when going from LEB to HALE at birth. Of all these countries Oman is the most affected, dropping an astonishing 39 places in rank. With a LEB of 77 years and a HALE at birth of 65.6 years, the data suggests that although Oman citizens are expected on average to live long lives, they are also likely to spend over ten years in less than full health. This suggests that Oman’s health development may not be as advanced as observed by their well above average LEB value. Other notable countries that decline in rank are Morocco and Algeria, which both drop over 30 places, and Turkey, Iran, and the United Arab Emirates, which all drop over 20 places.

In addition to changing the perception of a country’s health development, HALE also function as a convenient measure for comparing countries that have very similar LEB values. For example, Slovenia and Cyprus have similar LEB values of 80.9 and 80.7 years respectively. Using the LEB values alone, it is reasonable to believe that these countries are in a similar stage of their health development. When examining the HALE at birth values of Slovenia and Cyprus, which are 70.5 years and 73.3 years respectively, it becomes more apparent that Cyprus has progressed more in health despite similar LEB values because the citizens of Cyprus are expected on average to be in full health almost three years longer than the citizens of Slovenia. It is important to note that LEB is rarely used as a single measure to draw a comparison in the developmental progress of countries. LEB is more notable as a health indicator of development in the context of HDI. For this reason, an alternative index, Health-adjusted Human Development Index (HHDI), was created to allow for a more meaningful comparison utilizing HALE at birth data.

Table 4 showcases the ranking analysis for HDI and HHDI. Similar to the previous figure, the country names written in blue in the table indicate countries that have a considerable increase in ranking (5 or more places) from HDI to HHDI, while the county names written in purple indicate countries that have a considerable decrease in ranking (5 or more places) from HDI to HHDI. All data used to calculate HDI and HHDI comes from the same year (2016) and can be found in Appendix 3. Notable increase in rank include Angola, Azerbaijan, and South Africa, which increase by 11, 9, and 8 places respectively. Notable decrease in rank include Lebanon, The Syrian Arab Republic, and Oman, which decrease by 12, 10, and 10 places respectively. Despite these notable examples, few countries experience as dramatic changes in rank as compared to the previous ranking analysis. This is due primarily to the inclusion of the education index and income index that diminish the effect of the calculated health-adjusted life expectancy index (HALEI). Nonetheless, HHDI functions as a useful comparison tool for countries that have similar HDI values. For example, Oman and Belarus have an identical HDI value of 0.796, but have HHDI values of 0.779 and 0.805 respectively. This data suggests that Belarus has more human development than Oman, especially when considering health as a notable factor. This is a fair assumption to make considering the capabilities approach central to the idea of human development. At a fundamental level, HDI is attempting to calculate the capabilities of a population in a country by considering their ability to be healthy, educated, and have a decent standard of living. HHDI more directly considers disability and illness by using HALE at birth instead of LEB. In other words, HHDI more directly calculates the capabilities of an individual to be healthy that HDI. Therefore, it is reasonable to conclude that on average the citizens of Belarus have more human development than the citizens of Oman.

In addition to functioning as a complementary tool for understanding a country’s human development, HHDI may also help solve UNDP’s impending goalpost problem discussed in the previous subsection (4.2). Assuming UNDP decides to keep the LEI goalposts fixed at 20 years and 85 years, the creation of a modified human development index will allow for countries, with LEB values above the maximum threshold, to be more easily compared and evaluated. Ultimately, HHDI will allow for more information about the human development of a country by allowing for the assessment of the quality of life in addition to the quantity of life.

Overall, HHDI does not fundamentally change how HDI assesses human development. Just as HDI, HHDI aims to put a quantifiable value on human development, and more specifically the capabilities of people in a country to be healthy, educated and to have a decent standard of living. Nonetheless, the use of HALE at birth gives HHDI the potential to change the way the international community thinks about health and human development. The comparison between HDI and HHDI begs the question: Is it more important to consider the longevity of a population or is it more important to consider the overall health and wellbeing of a population? As demonstrated by the correlation data in section 4.1, LEB and HALE at birth have nearly the same capacity to function as an aggregate measure of the health development of country, but ultimately HALE at birth functions more directly as a measure of quality of life because of the inclusion of disability and illness in its calculation. Nonetheless, LEB is easier to calculate and compare than HALE at birth due primarily to data limitation. For this reason, HHDI should not be considered a replacement for HDI, but rather function as a complementary measure that aids in the understanding of the human development of a country.

Table 3: LEB and HALE at birth ranking analysis. LEB and HALE at birth values are from 2016. Data source: (WHO, 2018b)

Table 4: HDI and HHDI ranking analysis. Calculation for HDI and HHDI can be found in Appendix 3. Calculation are created using 2015 data. Health-adjusted Human Development Index (HHDI) uses HALE at birth instead of LEB as an indicator of health. Data sources: (Barro & Lee, 2016) (ICF Macro, 2015) (IMF, 2016) (UNDESA, 2015) (UNDP, 2015a) (UNESCO, 2016) (UNICEF, 2015) (WHO, 2018b).

5. Conclusion

As stated in the title, this paper aims to understand the extent to which life expectancy at birth (LEB) is a useful long-term indicator of health and human development. Overall, the previous sections focused on three main topics to answer the proposed research question. First, the strength of LEB as an indicator of health development was discussed and tested using correlation analysis. This analysis concludes that LEB correlates strongly with other health indicators on a global level and is, therefore, a well-suited measure for understanding a countries health development. This correlation analysis also demonstrated the value of health-adjusted life expectancy (HALE) at birth as a promising alternative measure to LEB. Next, the future of LEB was examined by projecting LEB in order to better understand the implications of a diminishing difference in LEB between the current top and bottom ranking countries in LEB. The projections indicated a narrowing on goalposts, or the minimum and maximum LEB values, that are used to calculate the life expectancy index (LEI) and the Human Development Index (HDI). The narrowing values indicated the diminishing value of LEB, considering that as LEB values become more similar globally, the measure will have less comparability power independently. This narrowing also indicated an impending problem for UNDP and the calculation of HDI. As the LEB in countries begins to exceed the current maximum of 85 years, the UNDP must decide to either adjust the goalposts or create a modified HDI measure that allows for a meaningful comparison of countries that exceed this maximum. Calculations of LEI using adjusted goalposts demonstrated that shifting the minimum and maximum values of LEB causes dramatic changes in LEI and may not be the best solution to the impending goalpost problem. A modified HDI appears to be a better solution to this problem. For this reason, the third topic that was explored was the creation of a new modified human development index, Health-adjusted Human Development Index (HHDI), that allowed for a better comparison of countries that had HDI values that were too similar or had a LEB value over the maximum threshold of 85 years. A ranking analysis was conducted to compare HALE at birth to LEB, and to compare HHDI to HDI. These ranking analyses concluded that HHDI could be used with HDI to better understand a country’s human development because of the inclusion of HALE at birth, a measure that more directly considers the quality of life in addition to the quantity of life than LEB.

Together, these results dictate that LEB will remain a valuable indicator for the foreseeable future, but may rely on complementary measures in some cases to allow for a more complete understanding of a country’s health and human development. As LEB continues to increase for lower performing countries and continues to plateau for higher performing countries, alternative measures such as HALE at birth and HHDI will become more prevalent. In the future, it will be necessary to consider a hierarchy for these measures. Based on the findings of this paper, HDI and LEB should remain as the basis for understanding a country’s health and human development, and HHDI and HALE at birth should be used as a supplemental measure to gain a better understanding of a country’s developmental progress. Ultimately, the increased use of HALE at birth and HHDI to determine a country’s development has the potential to incentivize countries to focus more on the quality of life of their citizens in addition to their longevity. Just as HDI altered the way the international community viewed development, HHDI and HALE at birth have the potential to change the way the international community views health development.

Bibliography

Anand, S., and A. Sen (2000). The Income Component of the Human Development Index. Journal of Human Development and Capabilities (1)1: 83–106.
Avenir Health (2018). Spectrum. Avenir Health.
Barro, R.J., and J.W. Lee (2016). Dataset of educational attainment, February 2016 revision. barrolee.com.
Crimmins, E. M., & Beltrán-Sánchez, H. (2011). Mortality and morbidity trends: is there compression of morbidity?. The Journals of Gerontology: Series B, 66(1), 75–86.
Fries, J. F. (1989). The compression of morbidity: near or far?. The Milbank Quarterly, 208–232.
Fukuda-Parr, S (2003). The human development paradigm: operationalizing Sen’s ideas on capabilities. Feminist economics, 9(2–3), 301–317.
Geyer, S. (2016). Morbidity compression: a promising and well-established concept?.
Gruenberg, E. M. (1977). The failures of success. The Milbank Memorial Fund Quarterly. Health and Society, 3–24.
ICF Macro (2015). Demographic and Health Surveys. The DHS Program.
IMF (International Monetary Fund) (2016). World Economic Outlook database. IMF.
Jagger, C. et al. (2001). Health Expectancy Calculation by the Sullivan Method: A Practical Guide. European Concerted Action on the Harmonization of Health Expectancy Calculations in Europe (EURO-REVES).
Khalsa, Dharma (2011). Life Expectancy: The Most Important Measure of Health. Alzheimer’s Research & Prevention.
Lane, David (2018). Values of the Pearson Correlation. Online Stat book.
Maddison, A (2010). Historical Statistics of the World Economy, 1–2030 AD. Paris: Organization for Economic Co-operation and Development.
Mathers, Murry, Salomon (2003). Methods for Measuring Healthy Life Expectancy. Health Systems Performance Assessment: Debates, Methods and Empiricism Chapter 33. World Health Organization Geneva.
Measure Evaluation (2018). Lesson 8: The Cohort Component Population Projection Method. USAID and PEPAR.
Mor, V. (2005). The compression of morbidity hypothesis: a review of research and prospects for the future. Journal of the American Geriatrics Society, 53(9s), S308-S309.
OECD (2018). Life expectancy at birth. OECD Data.
Oeppen, J., and J.W. Vaupel (2002). “Broken Limits to Life Expectancy.” Science 296: 1029–1031.
Robine, J. M., & Ritchie, K. (1991). Healthy life expectancy: evaluation of global indicator of change in population health. Bmj, 302(6774), 457–460.
Roser, Max (2018). Life Expectancy. Published online at OurWorldInData.org. Retrieved from: ‘https://ourworldindata.org/life-expectancy'
Riley, J.C. (2005). Poverty and Life Expectancy. Cambridge, UK: Cambridge University Press.
Sen, Amartya (1999) Development as freedom, Oxford: Oxford University Press
Sullivan, D. F. (1971). A single index of mortality and morbidity. HSMHA health reports, 86(4), 347.
Ul Haq, M. (1995). Reflections on human development. Oxford University Press.
Ul Haq, M. (2003). The birth of the human development index. BIRTH, 2, 1.
UNDESA (United Nations Department of Economic and Social Affairs) (2015). World Population Prospects: The 2015 Revision. UNDESA.
UNDP (1990). Human Development Report 1990. United Nations Development Programme.
UNDP (2015a). Table 1: Human Development Index and its components. UNDP.
UNDP (2015b). Training Material for Producing National Human Development Reports. UNDP.
UNDP (2016). Technical Notes: Calculating the human development indices — graphical presentation. Human Development Report 2016.
UNESCO (United Nations Educational Scientific and Cultural Organization) Institute for Statistics (2016). Data Centre. UNESCO.
UNICEF (United Nations Children’s Fund) (2015). Multiple Indicator Cluster Surveys. UNICEF.
WHO (2015). Life expectancy at birth indicator. UN. Retrieved from: http://www.who.int/healthinfo/indicators/2015/chi_2015_24_life_expectancy.pdf
WHO (2016). Healthy life expectancy (HALE) at birth. Indicator Metadata Registry.
WHO (2018a). Healthy life expectancy (HALE) at birth. Global Health Observatory (GHO) data repository.
WHO (2018b). Life Expectancy and Healthy life expectancy data. Global Health Observatory (GHO) data repository.
Wolfson, M. C. (1996). Health-adjusted life expectancy. Health Reports-Statistics Canada, 8, 41–45.
World Bank (2016). Life Expectancy at Birth, total (years). Data.
World Bank (2018a). Maternal mortality ratio (modeled estimate, per 100,000 live births). Data.
World Bank (2018b). Mortality rate, infant (per 1,000 live births). Data.
World Bank (2018c). Mortality rate, neonatal (per 1,000 live births). Data.
World Bank (2018d). Mortality rate, under-5 (per 1,000 live births). Data.
World Bank (2018e). New country classifications by income level: 2018–2019. The Data Blog.
World Bank (2018f). Prevalence of anemia among children (% of children under 5). Data.
World Bank (2018g). Prevalence of undernourishment (% of population). Data.

Appendix 1: Correlation Analysis Data

Data sources: (WHO, 2018b) (World Bank, 2018a)(World Bank, 2018b)(World Bank, 2018c)(World Bank, 2018d)(World Bank, 2018f)(World Bank, 2018g).

Appendix 2: LEB Projection Data 1990–2070

Projections created using DemProj in Spectrum v. 5.7. Data sourced from 2017 World Population Prospects by the UN Population Division. Life expectancy at birth data for each country is based on official estimates available through 2015. The age pattern of mortality is based on life tables through 2013 from the Human Mortality Database.

Appendix 3: HDI and HHDI Calculations

Calculation are created using 2015 data. Health-adjusted Human Development Index (HHDI) uses HALE at birth instead of LEB as an indicator of health. Data sources: (Barro & Lee, 2016) (ICF Macro, 2015) (IMF, 2016) (UNDESA, 2015) (UNDP, 2015a) (UNESCO, 2016) (UNICEF, 2015) (WHO, 2018b).