Chris Belch
11-15-00
Political Analysis (210)

 A rising topic of conversation on the world news scene is the topic of life expectancy.  “Life expectancy refers to the average number of years a population age category might live after reaching a specific age” (Seltzer 115).  Determining how long a typical person in a specific region might live is a very vital statistic to many different observers.  Knowing life expectancy, and what effects it, can be used to analyze and determine what can be done to help people live longer in particular areas of the world.  Everyone would agree that people want to live longer, provided that they are living healthy lives, but in order to do so, we must determine what shortens life expectancy.  By knowing what affects life expectancy, we can better compare the living conditions in several different countries, and how they effect a person in the long run.  The data we will be using, from Student MicroCase, measures life expectancy from the ages of 32.91 to 79.20 years of age.  With this given, life expectancy will be compared taking into account other factors such as unemployment, murder, inflation, and population growth of a country.  While research has shown that there are a number of other significant factors that reduce life expectancy, such as alcoholism and aids (www.bmj.com), MicroCase does not provide data for these factors, so they will not be analyzed.

 There have been some previous studies on the effects of unemployment on life expectancy.  Experts have found that unemployment, and the resulting economic effects, will alter the life expectancy of a country.  Though he is speaking towards the African American population, one expert’s research can be applied to the world in general.  He argues that “Economic depravation…apparent throughout the life course of blacks, have caused poor nutritional status, inadequate housing, and poor health conditions combined with deficient access to adequate health care” (Racial Differences in Aging 9).  This study tells us that if a person is unemployed, they can not sustain a decent standard of living, and therefore, as a result, the expected time that they will live will be reduced.  Since a lack of a job means a lack of income, a person will not be able to afford to purchase food for himself or his family, which will result in the deterioration of a healthy body, which will reduce life expectancy in that region. One of the biggest factors effecting unemployment and the resulting economic crash, specifically in third world countries, such as Algeria, is “The restructuring of public companies and the resulting job compression [has] brought about a further growth in unemployment” (www.washtimes.com).  This is one of the reasons that countries like Algeria are experiencing lower life expectancy rates.

 Judging by previous studies, it is reasonable to hypothesize that the relationship between unemployment and life expectancy is a significant, negative relationship; that is, as unemployment rises, life expectancy will decrease.  Unemployment is measured according to the percentage of the population that is currently unemployed, in MicroCase, during 1996, taken from the World Fact Book.    The range is from 0.4 percent to 60 percent, and includes data from 121 countries, while data is excluded or not available from 54 countries across the world.    From the data given, the measurement should be fairly accurate, as a large number of the countries are included in the data set.  The largest possibility for error is that there are fifty-four countries that are not included in the data set.  A large majority of these countries are third world countries, countries with a lower economic standing, and a lower life expectancy.  It is fair to assume that if data were available for these countries, then the effects would be even more significant than they already are.  After all, it stands to reason that a country with a lower unemployment rate would have a higher life expectancy, and there must be some reason why the statistics regarding unemployment are not available—most likely, the government does not want to reveal the actual state of the nation.

 Another area, which could possible effect life expectancy, is population growth.  Researchers have found that as population grows, the life expectancy of a country will decrease, for several reasons.  The main reason is that as a population increases, the available resources decrease at a rapid rate.  A study by the Population Council suggested that, “The expected addition of several billion more people will hamper ongoing efforts to reduce poverty and achieve sustainable development” (www.sdearthtimes.com).  They believe that as the population continues to grow, the amount of resources consumed by those people will be further reduced.  As unemployment rises, the availability of adequate food, housing, health care, and other basic needs will fall.  The Council believes that, “Declines in mortality, historically the main cause of population growth, will almost certainly continue.  Higher standards of living, better nutrition, expanded health services, and greater investment in public health measures have increased life expectancy by fifty percent…” (www.sdearthtimes.com).  While the world trend has shown a reduction in population growth, there have been significant increases in certain countries, typically with lower living conditions.  As this new subset of the population emerges, many jobs, food, and health care will be lost, causing a reduction in life expectancy.

 The relationship between population growth and life expectancy is likely to be a negative one; as population growth increases, life expectancy will decrease.  MicroCase measures the current population growth as provided by The World Factbook, in 1996.  The statistics measure how much a country’s population has increased over the past year by percentages.  The range in growth is from –2.75 to 16.55, a fairly large variation.  The data set will be fairly accurate, as it does account for all 174 countries, with none missing.  The only question regarding accuracy is how well each country measures its population growth, and how often this measurement is taken.  As with any measurement in the social sciences, it is hard to be sure that the information is completely accurate, although this should not be too much of a problem in this case.

 One of the most obvious factors, which will effect life expectancy, is the murder rate of a country.  It is safe to say that murder is a significant cause of reduction in life expectancy of a country.  As one study states, “The underlying pattern of mortality established under transition indicates not only an upsurge in adult (working age) mortality, but also a significant rise in socially induced, non-medical causes of death.  Non-medical reasons have become the principal cause of death among young adult males, accounting for roughly eighty percent of all deaths…Between 1990 and 1996, in Lithuania, 40,467 people died from accidents, poisonings, and injuries…seven percent were from murder” (www.un.lt/).  Murder causes a premature death, that could be avoided, and therefore, significantly reduces the length of time a person could possibly live.  If a county has a high murder rate, it means many people are dying earlier that they should, causing a reduction in the average life expectancy of a country.

 The relationship between murder and life expectancy is likely to be a negative one; as murder goes up, life expectancy will go down.  MicroCase measures murder by the number of homicides in 1990 per 100,000 people, as provided by United Nations statistics.  The number is very inaccurate, as the data was only available for 53 cases, leaving out 121 countries.  Without this large number of cases, the data will not be accurate.  However, if you analyze the countries that are included in the statistics, a majority of them are first world countries, that tend to be more industrialized.  It is likely that the countries that are not included would have an extremely high murder rate, which would greatly increase the significance of murder on life expectancy.  If you were able to analyze these counties, it is likely that they would have a lower life expectancy and a high murder rate.  Countries with strong military dictators, and those with little democracy, are probably hesitant to release information regarding their murder rates, since they would be extremely high, and would hurt the legitimacy of the country.  Another possible problem with the data, is of course, that it only takes into account reported murders.  There are probably many murders that go unreported in each country that would also increase this statistic.  However, we can still use the trends that we are given to demonstrate the relationship of murder to life expectancy.

 One final factor that could possibly effect life expectancy is inflation.  Inflation is the concept that the purchasing power of the dollar decreases over time.  Simply put, one dollar today will buy much less goods than it would ten years ago.  Researchers estimate that inflation has several effects on life expectancy.  They say that countries with high levels of inflation also have lower life expectancies.  This is due much in part to the instability of the economies of these nations.  As an economy grows unstable, the value of its currency is worth less on the world market, forcing many people into poverty, since they cannot afford decent living conditions.  It also has an effect on pension plans that do not adjust for inflation.  Many elderly workers who have retired are forced into poverty because their pension is based on the value of the dollar at the time, but that money cannot buy as much in today’s market (www.urban.org).  As inflation increases, the standard of living is reduced; people can no longer afford health care, insurance, food, and other basic necessities, that will help to increase the length of time that they are likely to live.

 The relationship between inflation and life expectancy is likely to be negative; as inflation goes up, life expectancy will be reduced.  MicroCase measures inflation by percentage, according to data from The World Factbook in 1996.  The data should be relatively accurate since there are a large number of cases.  Data is provided for 167 countries, with only seven cases missing.  This should provide an accurate picture of how inflation effects countries worldwide.  The range of inflation is measured from –0.2 percent to 243.3 percent inflation, providing a great deal of variance among the countries of the world.

 Now that we have formed some theories as to what factors will effect life expectancy, we must analyze the data.  The first comparison that we will make is through the correlation coefficients, or the effects of the relationship between the dependent variable (life expectancy) and all other variables that we will be analyzing.  While it is important to remember that correlation does not necessarily imply causation, it does show a relation and a possible cause.  Below is a table of the correlation between our variables as provided by MicroCase:

Correlation Coefficients

 Life Expectancy
Unemployment -0.583 **
Population Growth -0.455 **
Murder -0.276 *
Inflation -0.215 **
 

 When analyzing the correlation between life expectancy and unemployment, the data shows a correlation coefficient of –0.583.  By using this data, we find that there is a significant correlation between the two variables.  They are related to each other, and it is highly likely that they are significant, as indicated by the two stars.  A rating of two stars indicates a significance level of 0.01, so we are able to reject the null hypothesis that there is no relationship between unemployment and life expectancy.  Since the number is negative, that symbolizes a negative relationship between the two, meaning as one (unemployment) goes up, the other (life expectancy) goes down.  This correlation is also very strong at –0.583, as any correlation between 0.3 and 1 (absolute value) can be considered strong.  This data backs up our original hypothesis that life expectancy should be effected by unemployment.

 If we analyze the correlation between population growth and unemployment, we get a coefficient of –0.455, indicating that there is a strong correlation with a high level of significance (indicated by two stars), and a negative correlation.  We are again able to reject the null hypothesis, and deduce that as population growth goes up, life expectancy will go down.  Again, our original hypotheses, that there is a relationship between life expectancy and population growth, is correct.

 The correlation coefficient between murder and life expectancy is –0.276.  There is only one star after the number, so therefore the relationship is not as strong as the previous two, but it is still statistically significant.  We are able to reject the null, that there is no relationship between life expectancy and murder, since the star indicates a significant level of 0.05.  However, we can not be as sure about the relationship as in other cases.  Any cases that have a correlation between 0.1 and 0.3 can be considered to have some (though not a strong) relationship, so with -0.276, we can say that there is some relationship.  Once again, our original hypothesis has been proven correct, that life expectancy will be effected by murder, although not as strongly as one might have guessed.

 The final correlation, between life expectancy and inflation, yielded a result of –0.215, indicating a negative correlation.  This indicates that there is some relationship, and that it is statistically significant, so we are able to reject the null hypothesis, and deduce that as inflation goes up, life expectancy will go down in that given country, proving once again that our original hypothesis is correct.

Next, we must analyze regression, or the impact that a variable has, controlling for all other variables.  The following is a table of regression:

Regression (ANOVA)

 Un-standard b t
Unemployment -0.142 -1.245
Population Growth -0.848 -1.290
Murder -0.132 -3.192 **
Inflation -0.107 -2.278 *
 

The numbers given for regression are the “b” term, which means for every time you increase “X” by one, then the variable will increase (or in this case, since the variables are negative, decrease) by one.  If we regress unemployment, then we are given –0.142, meaning that if unemployment goes up by one, the life expectancy will be reduced by –0.142.  The “t” value for unemployment is –0.1245, with no stars, meaning that while unemployment will obviously have some effect on life expectancy, it is unlikely that that impact will be significant.  The “b” value for population growth turns out to be –0.848, with a  “t” value of  -1.290 (no stars), meaning that it, too, will have some effect on life expectancy, but not quite a significant effect.

When the regression test in done for murder, we are given a “b” term of –0.132, and a “t” value of –3.192, which is significant because it has two stars.  This data provides us with a strong relationship that shows that as the murder rate rises in a given country, the life expectancy will fall.  Similar regression data can be found for inflation.  Inflation provided us with a “b” term of –0.107, and a “t” term of –2.278, with one star.  Inflation will also have a strong negative effect on life expectancy, although with a little less significance than that of murder.

Now that we have analyzed the data for both correlation and regression, we must determine what the results mean.  It is interesting that when running the correlation numbers, unemployment and inflation both had very significant and strong relationships, but once the they were regressed unemployment became weak, and inflation was only marginally related.  It seems as though, being that inflation and unemployment are both similarly related economic problems, that if one had an effect, then so, too, should the other.  To test to see if they are similar to each other, we ran a correlation test between the two, and found that unemployment and inflation do not correlate too well, with a reading of 0.034.  Next we ran a regression model excluding inflation to see if unemployment would have more of an effect, but unfortunately the number only dropped, from –0.142 to –0.110.  Similar results can be found when dropping unemployment to measure for inflation; the “b” term increased slightly from –0.107 to –0.118, but the significant level dropped slightly from –2.278* (t) to –2.013*.  One possible answer that would explain why inflation and unemployment did not have the impact that we expected was that there was an extremely low number of cases that were used in the data sets.  The regression results only took into account 45 cases, while excluding data that was not available for 129 nations.  Given the low number of cases, it is very likely that if more cases were included, that unemployment and inflation would have much greater effects.  Most of the cases that were included were more industrialized nations, which have large welfare states.  In these nations, unemployment and inflation seems to be significantly lower than in the countries that are not included, thereby, creating less of an impact.  Also, because of the large welfare states, unemployed people are able to sustain a high level of life quality, expanding life expectancy, due to the fact that their governments will help defer the costs of food, shelter, and health costs, while the people remain unemployed.

Not surprisingly, the correlation and regression number for murder demonstrated that there is a strong and significant impact of murder on life expectancy.  Both tests produced two stars, showing high correlation and relation.  This is most likely due to the fact that as more people are killed due to unnatural causes, the expected time for them to live will obviously decrease.

The strength of our argument about factors that effect like expectancy was somewhat weakened by the regression data, as it was reduced from two stars in the correlation to one in the regression.  While this is still a fairly strong argument, it does not effect it quite as much as we would have thought.  After running a correlation between population growth and unemployment, it seems fairly easy to deduce that the two correlate heavily, with a coefficient of 0.378**, which makes it very hard to unpack which one is actually effecting life expectancy.  This is a very logical conclusion, since as a population increases, the rate of unemployment typically grows, due to the availability of fewer jobs.  To determine if this is in fact the case, we ran the regression tests by omitting population growth and came up with a “b” score of –0.167 and a “t score of –1.475 for unemployment, a slightly higher score, but still not significant.  However, when unemployment was taken away, population growth had a huge impact, as the “b” score jumped from –0.848 to –1.774, and the “t” score from –1.290 to –4.633, a dramatic increase in significance.  This proves that perhaps unemployment is heavily effected by population growth, and that was causing our data to be somewhat inaccurate as to what was really effecting life expectancy.

After analyzing the data, we have found that murder and population growth tend to be the largest factors negatively effecting the life expectancy in a given country.  Inflation also seems to be somewhat of a factor in its effects, and unemployment is indirectly effected by the population growth.  This means that, hypothetically speaking, if you wanted to increase your life expectancy, you should try to find a country that has lower, or at least stable, population growth and a low murder rate.  Finding a country with high employment and a lower inflation rate will also help increase your chances for increased life expectancy.  These seem to be four main factors that have an impact on life expectancy, but perhaps there are other factors that could be used as well.  Alcoholism, aids, and infant mortality seem to draw down the life expectancy of a country, but no data was available on MicroCase to test for these important factors.  Perhaps if one wanted to further explain life expectancy, they could investigate further into other economic factors, such as per capita income or the import/export ratio.  Other health factors could also be taken into account, such as abortion, availability of doctors and hospitals, or the quality of food.  Upon further analysis, one could perhaps find thousands of factors that directly effect life expectancy, so many so that you would run into the problem of multicolleniarity, or the inability to unpack which factor has the impact.  To achieve better results, one might also try to find better data measurements that include a larger variety of nations to accurately assess.  Perhaps finding the exact factors that will significantly influence life expectancy is an unreachable task, but it is fairly safe to say that our hypothesis that unemployment, inflation, and especially murder and population growth, will have a definite impact on life expectancy.

 Works Cited

Internet Sites:
http://www.bmj.com/cgi/content/full/319/7208/468/a
http://www.urban.org/retirement/st/straight10.html
http://www.un.lt/hdr/1997/chapter1/ch13.htm
http://www.sdearthtimes.com/et1198/et1198s9.html
http://www.washtimes.com/internatlads/algeria/25.html

Books:
Seltzer, Michael M.  The Impact of Increased Life Expectancy Beyond the Gray
Horizon.  New York, Springer Publishing Company, 1995.

Shrestha, Laura B.  Racial Differences Life Expectancy Among Elderly African
Americans and Whites.  New York: Garland Publishing, 1997.