Health of Nations

Introduction

The “Wealth of Nations” can easily be compared by their Gross Domestic Product (GDP) and for a comparative purpose, we use Per Capita GDP by Purchasing Power Parity (GDP per Cap PPP). Looking at the world map, most countries range from low to medium (colored red to yellow) while only a few countries are high (green). Wealth is surely concentrated.

However, in this paper, we are interested to understand the “Health of Nations”. Health is based on various parameters that we intend to examine here. But before we deep dive, we would like to see some sort of an overall indicator. United Nations Development Programme (UNDP) calculates the Human Development Index (HDI) based on four variables − life expectancy at birth, expected years of schooling, mean years of schooling, and per capita PPP gross national income. HDI displays a much better picture with a lot more countries on the higher side (towards the green).

Most often HDI is used as an indicator of health, but as we can see, only one parameter of life expectancy is covered in this. Other parameters of health like mortality, immunization, disease prevalence, doctor/hospital access, aging parameter, etc. also need to be considered for understanding the Health of Nations.

Section 1: Deep Dive into Various Parameters of Health

In this section, we will examine various other parameters of health and understand their association with Wealth (GDP) and Development (HDI). But before that let us examine the association between Wealth and Development itself.

As per capita GDP rises from low to medium, the HDI also rises. Once a GDP threshold is reached, the HDI also has reached its peak. There is a flattening seen around $15k. At this level, an HDI of 0.75 is achieved. So, $15k might be a desirable minimum level that a country could strive to achieve in Development terms.

Parameter: Life Expectancy

Life Expectancy is also a parameter for HDI. Hence it correlates most. There are various indicators under Life Expectancy. These are — “Life expectancy at birth (years)”, “Life expectancy at birth, male (years)”, “Life expectancy at birth, female (years)”, “Life expectancy index, Inequality in life expectancy (%)”, “Inequality-adjusted life expectancy index”. These are quite correlated to each other. So, we will examine two out of these — “Life expectancy at birth (years)”, “Inequality in life expectancy (%)”. The first one will help us compare the level across countries and the second one will help us understand gender bias on this parameter.

Indicator: Life expectancy at birth (years)

The picture quite resembles the HDI picture. This is expected as this indicator is a part of the index.
The map shows that most African nations have low life expectancy while Europe, the Americas, Oceania have a high life expectancy.

This indicator has a strong positive influence on HDI. The pattern is the same across continents.
This is expected, as we know, this indicator forms a part of HDI.

Here we attempt to answer whether “wealth” or per capita GPD by PPP has any influence on Life Expectancy at birth. There appears a positive association but not very strong and that too in the lower per cap GDP countries. The picture is similar across continents except for Africa, which sees a sharper slope.
A threshold wealth is required for nations to attain a healthy life expectancy (about $15k).

Indicator: Inequality in life expectancy (%)

This indicator is about the inequality between genders on life expectancy. The picture is a little better than life expectancy itself. However, in most of Africa and Indians Subcontinent has more gender bias than most nations.

The indicator has a strong negative influence on HDI.
This can be seen across continents as well. Even Europe which has far lower inequality also shows a sharp negative pattern.

GDP per capita has quite a strong influence on this indicator. Once the threshold is reached, which is again about $15k, the inequality stabilizes.
In Africa, the influence is strongest, while Europe where per cap GDP is mostly above the threshold, show no influence.

Parameter: Age of Population

An aging population is a healthier population as people can live longer. This parameter is not considered in HDI. Let us examine how would this parameter relate to Development and Wealth. Here we have two indicators — “Median age (years)” and “Population ages 65 and older (per 1,000 people)”. We will consider the 65-plus population indicator as it would discriminate longevity more than the median age.

Indicator: Population ages 65 and older (per 1,000 people)

The aged population is far concentrated in Europe while the rest of the world has a medium to a low proportion of the aged population. The lowest aged populations are in Africa, the Indian Subcontinent, and Latin America.

The aged population has a positive association with HDI. However, somewhere around 9%, it starts plateauing.
Africa shows a similar and stronger association up to the threshold. Americas+Oceania has even an influence on HDI beyond the threshold. Asia does not appear to affect HDI much. Europe with a high proportion much beyond the threshold has no association with HDI.

GDP per cap dues affect the aging of the population but again up to a threshold, which in this case is $50k.
Africa having the lowest per cap GDP and the lowest proportion of aged shows a strong association. Similarly, GDP affects aging in Americas+Oceania. In Asia, there is a weak influence up to a threshold of $15k. In Europe, there is almost no influence.

Parameter: Mortality Rate

There are various indicators under Mortality Rate. These are — “Mortality rate, infant (per 1,000 live births)”, “Mortality rate, under-five (per 1,000 live births)”, “Maternal mortality ratio (deaths per 100,000 live births)”, “Mortality rate, male adult (per 1,000 people)”, “Mortality rate, female adult (per 1,000 people)”.
We will examine the infant and maternal mortality rates viz: “Mortality rate, infant (per 1,000 live births)” and “Maternal mortality ratio (deaths per 100,000 live births)”

Indicator: Mortality rate, infant (per 1,000 live births)

Infant Mortality Rate is high in central parts of Africa and the Western part of the Indian Subcontinent.
It is low in most of the rest of the world.

Infant Mortality Rate has a strong negative association with HDI.
This pattern can be seen across continents.

GDP per cap has an influence on Infant Mortality up to a threshold of around $15k.
In Africa where GDP is low, the influence is much sharper. In Americas+Oceania, influence is a little less. Asia reflects the global pattern with the same threshold of around $15k. Europe has low mortality throughout with almost all countries above the GDP threshold.

Indicator: Maternal mortality ratio (deaths per 100,000 live births)

Maternal Mortality is low in central parts of Africa. It is on the high side in most parts of the world.

This indicator has some association with HDI and tapers off at high levels of mortality.
All continents except Europe show the same pattern. In Europe, this indicator is almost a low constant.

GDP per capita has an influence on maternal mortality up to the threshold of $15k.
Africa and Asia show the same pattern. In Americas+Oceania the influence is weaker. In Europe, there is almost no influence.

Parameter: Environment

Environment factors like pollution, sanitation, hygiene play an important role in the health of an individual. For this parameter, we have these indicators — “Mortality rate attributed to unsafe water, sanitation and hygiene services (per 100,000 population)”, “Mortality rate attributed to household and ambient air pollution (per 100,000 population, age-standardized)”, “Population using at least basic drinking-water services (%)”, “Population using at least basic sanitation services (%)”.
We will examine two out of these viz — “Mortality rate attributed to unsafe water, sanitation and hygiene services (per 100,000 population)” and “Population using at least basic drinking-water services (%)”.

Indicator: Mortality rate attributed to unsafe water, sanitation, and hygiene services (per 100,000 population)

Inadequate sanitation leading to death is a problem in central parts of Africa. While the rest of the world is in a much better position.

There is an association of this indicator with HDI. The association is stronger when the mortality rate is lower and if it goes up, its effect on HDI diminishes.
America+Oceania and Asia have a stronger association as their mortality rates are lower, while Africa has a weaker association as mortality rates are higher. Europe has almost a constant mortality rate across nations.

GDP per capita influences mortality rate up to a threshold of about $15k.
Africa shows the sharpest influence as it has low GDP per capita. GDP does not affect this indicator in Europe.

Indicator: Population using at least basic drinking-water services (%)

Basic drinking water services are low in central parts of Africa as well as some countries in Asia. In the rest of the world, this is much less of a problem.

This indicator has an association with HDI.
In Africa, once this crosses 60%, the increase in HDI is sharper. In Amecicas+Oceania and Asia, the sharpness in association becomes stronger after around 85%. In Europe, the association is stronger after about 90%.

GDP per capita has an influence on this indicator up to the threshold of about $15k.
GDP per capita has an influence on this indicator with minor changes in GDP leading to better water services. Americas+Oceania and Asia influence up to the GDP threshold. Europe already mostly above the GDP threshold, does not influence water services.

Parameter: Health Infrastructure

The infrastructure and access to health services would play a role in the wellbeing of the population. This includes government expenditure. Under this we have — Current health expenditure (% of GDP), Medical doctors (per 10,000), Dentists (per 10,000), Nursing and midwifery personnel (per 10,000), Pharmacists (per 10,000).
We will examine two out of these viz — Current health expenditure (% of GDP), Medical doctors (per 10,000)

Indicator: Current health expenditure (% of GDP)

The government percentage spending on health is low across the world. Except for the USA where the percentage is the highest.

There is some association of this indicator on HD but not much.
Africa and Asia display almost no association, while the other two display a weak positive association.

GDP per capita does not affect the government health expenditure proportion.
In Americas+Oceania there is a weak positive association, but in the other three, there is almost no association.

Indicator: Medical doctors (per 10,000)

Except for a few small countries in Europe and one in Asia, the proportion of doctors in most countries is medium to low.
Africa, the Indian Subcontinent, South-East Asia have the lowest.

Medical doctor proportion has a weak association with HDI up to a threshold of about 20 doctors per 10k people.
In Africa, this threshold is 10. In Americas+Oceania the association remains without a threshold. In Asia, the threshold is closer to the world level of 20. In Europe, there is almost no association.

GDP per capita influences doctor proportion up to a threshold.
In Africa, the GDP influences without any threshold. In Americas+Oceania and Asia, the threshold is around $15k to $20k. In Europe, there is almost no influence.

Parameter: Disease

A population with a high proportion of diseases is not a healthy population. In this parameter, we will explore the diseases and deaths caused by these. The indicators we have here are “Age-standardized mortality rate attributed to non-communicable diseases, female”, “Age-standardized mortality rate attributed to non-communicable diseases, male”, “Prob both sexes dying age 30–70 from cardiovascular, cancer, diabetes, or chronic respiratory disease”, “Prob female dying age 30–70 from cardiovascular, cancer, diabetes, or chronic respiratory disease”, “Prob male dying age 30–70 from cardiovascular, cancer, diabetes, or chronic respiratory disease”, “Tuberculosis incidence (per 100k people)”.
We will choose two out of these viz — “Tuberculosis incidence (per 100k people)” and “Prob both sexes dying age 30–70 from cardiovascular, cancer, diabetes, or chronic respiratory disease”.

Indicator: Tuberculosis incidence (per 100k people)

Central to Southern Africa is the most affected. Indian Subcontinent and South-East Asia are also affected but to a lesser degree.
Most of the rest of the world is less affected by Tuberculosis.

Tuberculosis incidence has a much lesser association with HDI.
In Africa, there is almost no association. In Americas+Oceania and Asia up to a threshold of 200 per 100k, it shows a negative association. Europe has a stronger negative association, although it has the least Tuberculosis case proportions.

GDP per capita affects this indicator up to the same threshold of around $15k.
Africa and Asia show this effect, but not the other two to this extent.

Indicator: Prob both sexes dying age 30–70 from cardiovascular, cancer, diabetes, or chronic respiratory disease

Except for Americas+Oceania and Western Europe, most of the rest of the world has a high incidence of death due to these diseases mentioned in the indicator.

There is a weak association of this indicator with HDI.
In Africa and Asia, the association is up to a threshold of a little over 20% of the indicator. Americas+Oceania has a weal association. Europe has a stronger association.

GDP per cap has a weak effect on this indicator.
Africa has almost no effect. Americas+Oceania has a weak effect. Asia and Europe have a little stronger effect but up to a threshold of about $35k and $50k respectively for GDP per cap.

Parameter: Vaccination

When infants and children are vaccinated, they are more likely to remain healthy as adults. In this parameter, we have two indicators — “Infants lacking immunization, DTP (% of one-year-olds)” and “Infants lacking immunization, measles (% of one-year-olds)”
We will choose “Infants lacking immunization, DTP (% of one-year-olds)”.

Indicator: Infants lacking immunization, DTP (% of one-year-olds)

Except for parts of central Africa and some in LatAm and Asia, most of the world does not lack immunizing infants for DTP.

The association of this indicator with HDI is quite weak.
This can be seen on all continents. However, there is some association in Americas+Oceania and Asia.

Overall, GDP per cap does not seem to affect this indicator.
However, in Asia, there is some effect up to a threshold of about $15k, and in Europe a very weak up to a threshold of about $35k.

Parameter: Suicide

The main cause of suicide is some sort of mental stress. So, this parameter would indicate the mental health of a population. Here we have two indicators — “Suicide rate, female (per 100k people, age-standardized)” and “Suicide rate, male (per 100k people, age-standardized)”.
We will choose “Suicide rate, female (per 100k people, age-standardized)”.

Indicator: Suicide rate, female (per 100k people, age-standardized)

On this indicator, Africa generally does a better job, so does LatAm and the Middle East.
India and a few central African countries are worse on this.

This indicator has a weak association with HDI.
In Africa, there is some association up to a threshold of 5 per 100k. The other continents have a much weaker association.

GDP per cap does not seem to affect female suicides at all.
All continents show a scatter with almost no pattern.

Summary of Section 1

Firstly, it is quite clear that Human Development Index does not completely cover the health aspect as we see weak to no association of many indicators with HDI. Therefore using this indicator for health could be misleading. For the Health of Nations, a Health Index using various health indicators is needed.
The other thing we see is that GDP per capita of US$15,000 at 2017 prices is an inflection point for many indicators up to which GDP affects the indicator. This must be used by countries as a target to achieve.
Although Africa comes out with worse cases, it is not all of Africa and some Asian, as well as LatAm countries too, show up in worse case scenarios. Therefore, the continents do not completely give the health picture. We need to cluster countries into different sets.

Section 2: Country Clusters

We intend to do country clusters at two levels. First, we form main clusters, and then within each, we will form sub-clusters. Since we desire to form a two-level hierarchy, we will use a hierarchical clustering algorithm. For linkage, we will use the “ward” method as it minimizes within-cluster variance which will help in forming more homogenous groups of countries.

Based on the indicators (scale standardized to give equal importance to each indicator), we have the following four clusters of countries. Central Africa and a few for Asia make the lowest cluster c1 (in red). Indian Subcontinent, South-East Asia, and some other countries make the next cluster c3 (in purple). Most of LatAm, many from Asia, Eastern Europe, some from North Africa make the next cluster c0 (in blue). Northern America, Western Europe, and Oceania make the highest cluster c2 (in green).

Countries of Cluster c1

There are 39 of 166 countries in this cluster. Out of these 36 of 49 African, 2 of 38 Americas+Oceania, and 1 of 42 Asian countries are in this cluster. No European country is in this cluster.
This splits into four sub-clusters with African countries spread in all sub-clusters.

Countries of Cluster c3

There are 35 of 166 countries in this cluster. Out of these 18 of 42 Asian, 10 of 38 Americas+Oceania, and 7 of 49 African countries are in this cluster. There are no European countries in this cluster.
This splits into four sub-clusters with Asian countries spread into all four, and African and Americas+Oceania spread into three of four sub-clusters.

Countries of Cluster c0

There are 59 of 166 countries in this cluster. Out of these 21 of 38 Americas+Oceania, 19 of 42 Asian, 13 of 37 European, and 6 of 49 African countries are in this cluster.
This splits into six sub-clusters with Americas+Oceania and Asian countries spread into five, African spread into four, and European spread into three of six sub-clusters.

Countries of Cluster c2

There are 33 of 166 countries in this cluster. Out of these 24 of 37 European, 5 of 38 Americas+Oceania, and 4 of 42 Asian countries are in this cluster. There are no African countries in this cluster.
This splits into three sub-clusters with European and Asian countries spread into all three and Americas+Oceania spread into two of three sub-clusters.

Summary of Section 2

The following inferences we draw from the clustering analysis:

• Continents do not define healthier vs not so healthy countries.
• The four clusters represent the most healthy (c2), less healthy but not critical (c0), the less healthy and a little critical (c3), and the least healthy most critical (c1) nations.
• European countries never go into the lowest two clusters, and African countries never go into the highest cluster.

With these clusters, we can now analyze countries in these homogenous groups.

Section 3: Analysis of Indicator Patterns and Trends by Clusters

Having formed these clusters, we would now like to see how countries fare and compare within and across clusters. Since we have 166 countries, we will not look at all of these but take a subset. We have decided to take countries with a sizable population (in this case over 50 million) to be representative of their clusters. Thus, we have 31 countries to analyze in the section.
We will first look at patterns and trends with HDI and GDP and then further with one indicator from each of the eight parameters we have analyzed in section 1.

Indicator: Human Development Index

The pattern of HDI is very straight forward. As we go down the clusters the HDI steadily decreases.

Over the years, the clusters have witnessed an almost parallel progression. This means the gap between clusters on HDI has not been narrowing.
Cluster 2, cluster 0, and cluster 1 countries have been progressing in a narrow band of HDI, while there is a much larger band in cluster 3.

Indicator: GDP per capita (2017 PPP $)

There is indeed heavy discrimination on GDP per capita. The highest cluster countries have a much higher level than the other three. However, the progressive decrease is still there.

Over the years the GDP gap appears to have increased. The highest cluster (c2) countries have had better growth, than the other three clusters. Even the next cluster (c0) has better growth than the bottom two. The bottom two (c3 and c1) have been on the verge of being stagnant.

Indicator: Inequality in life expectancy (%)

The difference between clusters on Inequality in life expectancy is vast.

Over time the decline in the lowest cluster c1 is more than the higher clusters, which is a good sign as the gap is getting reduced.
We can see most countries in c1 improving on this indicator. However, the next cluster c3 does not improve at the same pace, which should be a matter of worry for the countries in this cluster. The better cluster c0 shows some improvement and the best c2 is almost flat.

Indicator: Population ages 65 and older (per 1000 people)

This indicator discriminates the clusters quite well. The range of old people proportion is wide and gets lower as we move across clusters.

The trend shows that the better clusters c2 and c0 have shown growth, while the worse clusters have moved little.
Across countries, we see a higher upward trend in the best cluster c2, followed by the next cluster c0. In the third cluster c3 some countries have moved up a bit and some remained static. In the worst cluster c1, all countries have been almost static.

Indicator: Mortality rate, infant (per 1,000 live births)

This is yet another indicator that discriminates between clusters a lot.

This indicator has seen a good improvement in all three less than best clusters.
The worst cluster c1 has seen the sharpest decline across countries. Clusters c3 and c0 also have narrowed the gap with the best c2. This is a good sign.

Indicator: Population using at least basic drinking-water services (%)

This indicator is less discriminatory with the top three clusters. However, countries in the worst cluster have quite low proportions.

The trends also do not show a narrowing of the gap for c3.
Most countries across clusters other than the best have shown improvement, but not to the extent to narrow the gap. This indicator is an area of concern.

Indicator: Medical doctors (per 10,000)

This indicator also discriminates the clusters quite well. The lowest two clusters have quite low proportions compared to the highest two.

Although there is some improvement in the top three clusters over time, the lowest cluster has remained almost stagnant with low proportions.
This indicator too is a matter of concern for lower cluster countries.

Indicator: Prob both sexes dying age 30–70 from cardiovascular, cancer, diabetes, or chronic respiratory disease

This indicator discriminates to some extent bot does not discriminate well between the lowest two clusters. The lowest cluster c3 is a little better than c1.

Over the years, the decline of this indicator has been slow.
Across clusters we see a decline in most countries, however, in cluster c3 we do not see any decline across countries.

Indicator: Infants lacking immunization, DTP (% of one-year-olds)

This indicator is not too much discriminatory across clusters. However, the best cluster c2 has a low percentage for this indicator across countries.

The trend has shown some decline for both c1 and c3 the worse clusters.
Countries across clusters have been erratically moving with ups and downs without consistency.

Indicator: Suicide rate, female (per 100k people, age-standardized)

This indicator does not appear to discriminate the clusters at all.

There appears to be a very slight downward trend, but not much across clusters.
At the country level, some countries have declined while many have remained stagnant.

Summary of Section 3

The pattern and trend analysis has shown that many indicators discriminate the clusters far more than the HDI or the GDP. We can also see that for several indicators many countries or even the cluster of countries have not improved much which should be a cause of worry.

Final Takeouts

In summary, we make the following points:

Per capita GDP on purchasing power parity of $15,000 at 2017 prices should be made a global benchmark or target for countries.

There must be a global effort to raise countries to this target level for those that are at a lower level.

To be able to assess the Health of Nations we need to move beyond the Human Development Index and use many others than just the Life Expectancy at Birth as a complete health indicator.

The various indicators that discriminate the clusters and those which have not seen much improvement should be targeted for improvement in countries that are weak on those indicators.

Credits

The sources of data for this report are:

• UNDP Human Development Report at (http://hdr.undp.org/en/data). UNDP has shared their data under Creative Commons Attribution 3.0 IGO which allows free use if credit is given (as we have done here). [No changes to the data were made].
• Kaggle at https://www.kaggle.com/utkarshxy/who-worldhealth-statistics-2020-complete. The Kaggle user has shared their data under the “Creative Commons Zero (CC0) — No rights reserved” license which allows free use.

This report has been prepared by gralovis insights private limited, Mumbai — 400078, India.

Originally published at http://gralovis.com.