In this thesis, first of all I have tried to describe what is demography and different ways to
collect demographic data. Then, I have applied some of the demographic techniques to the
population of Pakistan. Here are my findings:
First of all, I have considered the infant mortality in Pakistan and applied the test of
hypothesis along with 2 x 2 table to show that there is a difference of
facilities/services given by the government to the urban and rural area’s population
and find out the results of z and chi-square tests with p-value. The results indicate
that there is really a huge difference of policies between urban and rural areas of
Pakistan and I have found the p-value 0.00001 which show our hypothesis is highly
significant. I have noticed that since only the 35% of the population is residing in the
urban areas but still urban areas are under consideration all the times while the rest
65% areas having the less attention by the government institutions.
Secondly by using the data given by Federal Bureau of Statistics, Pakistan I have set
up different life tables for the total population, urban and rural population and for the
male and females population of Pakistan. The results show that the life expectancy at
birth in urban area (68.7 years) is 6% higher than the rural areas (64.3 years).
Similarly, the probability of dying at the first age interval is also 10% smaller in the
urban area then the rural one (i.e. 0.06444 & 0.07197 for urban and rural
respectively). Moreover, the female life expectancy at birth (68.4 years) is found to be
7% higher than the male life expectancy (64.3 years).
Third, I have applied decomposition technique introduced Kitawaga (1955) to see
how much of the difference between death rates in urban and rural population is
attributable to differences in their age distributions. The results shows that the
original difference between the urban and rural population is -0.00210 (by equation
7.2) while the contribution of age compositional differences and contributions of age
specific rate differences are -0.00052807 & -0.00157492 respectively (by equation
7.3). Further, the proportion of difference attributable to differences in age
composition is found to be 25% whereas the proportion of difference attributable to
differences in rate schedules is 75% which shows that both parts are contributing in
the same direction to the difference.
Page 7 of 103
Lastly I have tried to make a population forecasting for Pakistan. For this, a few
methods has been discussed and have made a forecast by using the compound rate of
growth method and cohort component method. According to the first method, it
shows that there might be 294.96 million population in the year 2032 (equation 8.13)
whereas the second method states that it might be 258.09 million in the year 2031
(equation 8.14). It seems reasonable to say that the estimates found by the cohort
component method are more reliable than the any other method as the cohort
component is now the only method on which demographers are relying much.