1. Md. Mizanur Rahman
Lecturer
Department of Population Science and Human Resource Development
University of Rajshahi,
Rajshahi-6205
Bangladesh
E-mail: mizanur_rub@yahoo.com

2. Shamima Akter, M. Phil fellow
Department of Population Science and Human Resource Development,
University of Rajshahi,
Rajshahi-6205
Bangladesh
E-mail:samimarub@yahoo.com

Correspondence:
Md. Mizanur Rahman
Lecturer
Department of Population Science and Human Resource Development
University of Rajshahi, Bangladesh
E-mail: mizanur_rub@yahoo.com

Introduction

Population ageing is one of the most distinctive demographic events nowadays. The literal meaning of the term "age" refers to the time during which a person or thing has lived or existed. The United Nations identifies exclusively for the purpose of demographic comparison, populations who have reached the age of 60 years as "older persons". For developing countries, aged is defined as the population aged sixty years and above (UN 2004), and is an emerging issue in Bangladesh. Although in Bangladesh ageing is at its early stage compared with developed countries.

The elderly population (aged 60years and over) is rapidly increasing worldwide and the increase will, both in relative and absolute terms, be much more dramatic in low-income than in high-income regions of the world (UN 2003). Today, world-wide, there are around 690 million persons aged 60 years and over; this total will almost double by 2025 and is expected to reach nearly two billion by 2050. The vast majority of these older persons will reside in the developing world (Huber 2005). Out of a population of almost 140 million in Bangladesh, more than 7 million (5.1%) are aged 60 years and above (WHO 2005). By 2020, Bangladesh will, with a projected 14 million elderly people, be one of the 10 nations with the largest elderly population (Solomos et al. 2001, Kabir et al. 2005).

In the near future virtually all countries will face ageing of their population, although at varying levels of intensity and in different time frames. The prospects that the position of the elderly in Bangladesh may undergo significant changes in the current century and the next, causing a disproportionate accumulation of the very old, and therefore the most vulnerable sector of the population is rather high (Sattar & Rahman 1993). Responding to ageing in advance would enable countries to reap the "demographic dividend" produced by reduced fertility rates (Bloom et al. 2002, Kabir 1995).

Most of the developing countries are experiencing population ageing resulting in a rapid increase the proportion of population 60+. The rapidly increasing elderly population is a new and important group in terms of social economic and changing cultural context. In Bangladesh, the aging issue has only recently begun to emerge as a cause of concern. Interest in the subject of aging of the population is of very recent origin and it still needs research and understanding of the issues and their relationships.

To understand the inherent peculiarities about the growing number of aged population of Bangladesh the study is carried out with the following objectives.

1. To examine trends in some demographic parameters of the aged population;
   
2. To investigate various characteristics of the older persons;
   
3. To assess statistically the components affecting the intensity of ageing, dependency of the aged and some of their indicators.

Data and methods

Sources of data
The information presented here is secondary and we have used mainly the census data from the Bangladesh Population Census, held in 1991 & 2001, published in September & July 1994 & 2003 respectively; Analytical report and Statistical Yearbook of Bangladesh 2001, World Population Prospects 2004. All these data are officially published by Bangladesh Bureau of Statistics (BBS) a sponsored agency of Bangladesh Government.

The definition of the elderly is categorized as young-olds (who are of age 60-69 years), old-olds (who are of age 70-79 years) and extremely olds (who are of age 80 years and above). In case of Bangladesh, like many developing countries population aged 60 and over has been considered to be the aged or the elderly population (also older persons) (UN 2004). The present study is carried out considering the population of age 60 and over as the older person is the target population for analyzing ageing of population.

Methods
The study uses standard procedures for aged analyses: Aged growth rate, Total growth rate, Index of ageing, Aged dependency ratio, Total dependency ratio, Aged support ratio, Total support ratio, Familiar support ratio, Elderly sex ratio and also uses well known statistical techniques such as Collinerity statistics, multiple regression analysis and Stepwise regression analysis.

Stepwise regression technique is used to build the model of multiple regression analysis and is one of the most versatile data analysis procedures. Regression can be used to summarize data as well as to study relations among variables. But including a large number of independent variables in a regression model is never a good strategy, unless there are strong, previous reasons to suggest that they all should be included. The observed increase in R2 (Coefficient of determination) does not necessarily reflect a better fit of the model in the population (Montgomery & Peck, 1982). A model with many variables is often difficult to interpret. So, it is important to exclude potentially relevant independents. Thus, we applied the stepwise selection procedures for selecting variables to be included in the model.

Results and Discussion

Trends of life expectancy, crude death rate and total fertility rate of ageing population:

Table 1 represents the life expectancy at birth, crude death rate and total fertility rate by world, LDC and Bangladesh.

Figure 1& 2

Table1. Life expectancy at birth, crude death rate and total fertility rate by World, Lest Developed countries (LDC) and Bangladesh, 1970 - 2030

Note: LDC= Least Developed countries, CDR= Crude Death Rate and TFR= Total Fertility Rate

Table 1 and Figure 1 & 2 show that life expectancy at birth subsequently increasing for different years in the world, LDC and Bangladesh. The result shows that life expectancy for females is higher than males and their life expectancy is lower for Bangladesh than that of the world population and is comparatively lower for LDC. The older population in developing countries like Bangladesh is rapidly increasing by decline in mortality and fertility rates, mainly due to improved health conditions and family planning respectively. The decline in mortality rates has been the main trigger at the initial stage of the demographic transition (ESCAP 2004). As reflected by Table 1 and Figure 1 & 2, the decline in mortality rate has not only been a common feature in the World but also in LDC and Bangladesh. Fertility has also played a marginal role in the developing countries of the region at its primary stage. However, as reflected by Table 1 & Figure 3, total fertility rate of Bangladesh has substantially declined from 1970-75 to 2000-05 and in 2025-30 total fertility will be 2.22 and it is approximately same as the rest of the world.

Figure 3

Table2. Percentage of aged population by World, Lest Developed countries (LDC) and Bangladesh, 1970 - 2030

Population ageing is a process common to almost all developed and developing countries. The dynamics of population ageing differ from country to country depending on their level of economic and social development as well as policies adopted during the past. Table 2 shows that Bangladesh and LDC is already home for than half of the world's older aged and the share would continue to increase in the next 30 years later. The population aged 65 and over for Bangladesh has increased from 4.8% in year 1970 to 5.7% in 2005 and the latter 2030 years will be 10.6% of the total population. Increasing pattern of aged population during 1970 to 2000 for Bangladesh and LDC are approximately the same but later in 2000 it shows substantial difference up to 2030. After 2000 the percentage of aged population is higher than LDC. Population ageing has taken place at a rapid rate in the Asia-Pacific region during the past fifty years and would continue at a faster pace during the next fifty years (ESCAP 2004).

Indicators of ageing

Some indicators for studying the ageing situation of the country as a whole are provided in Table 3 & 4. These indicators are calculated using data both from 1991 and 2001 censuses.

Table3. Some indicators of ageing population in Bangladesh, 1991

Table4. Some indicators of ageing population in Bangladesh, 2001

Table 3 & 4 indicated that the aged population in 1991 was 5.7 millionamd is set to increase to 7.6 million in 2001, of 4.2 million for male and 3.4 million for female aged population. The rate of linear increment is 3.5% per annum. The temporal effect has largely fallen on the rural-urban elderly population. The rural older persons have 4.8 million in 1991 to rise 6.2 million in 2001 but in urban areas 45 million to 1.3 million aged population during 1991 through to 2001. The result shows that the elderly population is higher in rural areas than in urban areas. The share of women in the 60 years and over age group would increase from 5.08 per cent of the total in1991 to 5.60 per cent by 2001. In contrast, the share of men in that age group rises from 5.86 per cent to 6.58 per cent during the same period. This phenomenon is underlined by the higher life expectancy of women due to better survival chances at old age due to biological reasons. An important dimension of the ageing process is the rising share of women in the older age groups. Ageing has been a progressively gendered experience in the Asia-Pacific region (ESCAP 2004)

Work status of elderly population
The working status of elderly population both for male and female are shown in Table 5.

Table5. Percentage distribution of older persons by working status, sex and residence in Bangladesh, 2001

Table 5 shows that most of the female older are doing household work (58.1%), of which 60.2% are in rural areas and 52.2% in urban areas. Table 5 also shows that 51.3% male elder engaged in agriculture, particularly 57.9% in rural areas and 22.5% in urban areas. Transport, construction, services, hotel & restaurant and industrial work are involved in low percentages. Thus, the largest portion of male work force is engaged in agriculture and females are in household work.

Figure 4

Multiple regression analysis
Therefore, to evaluate the multiple effects of explanatory variables OTGR(X₁₁), RTGR(X₁₃), OIA(_X18), OADR(X₂₁), OTDR(X₂₄), OAPHH(X₃₂), OHHS(X₃₃), OALRAP(X₃₄) on OAGR(X₁₀) for census 2001 the following regression equation is assumed:
X₁₀= ß0+ ß1X₁₁+ ß2X₁₃+ ß3_X18+ ß4X₂₁+ ß5X₂₄+ ß6X₃₂+ ß7X₃₃+ ß8X₃₄+U… … … (1)
X₁₀= 1.579+ 0.0647X₁₁+ 0.289X₁₃+ 0.101_X18+ 1.30X₂₁-0.472X₂₄-1.074X₃₂+
1.141X₃₃+ 0.024X₃₄+U… … …

Stepwise regression analysis
The multiple regression equation (1) of Overall Aged Growth Rate (X₁₀) is assumed on eight explanatory variables. The stepwise regression methods are used and resulted in that only three of eight explanatory variables OTGR(X₁₁), OHHS(X33) and OADR(X₂₁) explain more than 66% of total variation in OAGR(X₁₀).

Table6. Estimates of stepswise regeression analysis, OAGR (X10) as the dependent variables, Bangladesh, 2001

The fitted regression equations by using stepwise regression command for the equation 1 are
X₁₀= 1.83 + 0.74(X₁₁) … … … 1(_i)
(0.142) (0.089)
X₁₀= 4.41+ 0.76 (X₁₁) - 0.53 (X₃₃) … … … 1(_ii)
(0.634) (0.079) (0.129)
X₁₀= 3.88 + 0.94 (X₁₁) - 0.69 (X₃₃) + 0.09 (X₂₁) … … … 1(_iii)
(0.642) (0.103) (0.137) (0.035)

(Note: all the regression coefficient, multiple correlation coefficients R, R² and change in R² are significant at 5% level; value in the braces indicates standard error of the estimate)

At the first step the variable OTGR (X₁₁) is included in the model due to the largest correlation with OAGR (X₁₀) among the eight variables. The value of R2 is 0.52 which indicates that approximately 52% of variation of OAGR is explained by OTGR. In the next two steps the variables OHHS and OADR are included in the model and the value of R² 0.67 with highly significant F ratio (F= 51.72 & 39.82, p= 0.000). Other variables are excluded from the model due to minimum tolerance and insignificant t ratio. From the final model we see that all the coefficients (Table 7) are highly significant including the intercept.

Table7. Collinearity diagnosis for the variables included in the model when OAGR as dependent variables, Bangladesh, 2001

Table8. Estimated coefficients obtained from stepwise regression considering OAGR as dependent variables in Bangladesh, 2001

From Table 7 of collinearity diagnostics, we see that condition index corresponding to second, third and fourth eigen values are 4.60, 21.06 and 29.10 (below 30) indicates that there is no problem of multicolinearity in the data, the variance inflation factor (VIF) corresponding to OTGR, OHHS & OADR are 1.84, 1.25 & 2.02 is far below 5 also favors the conclusion (Table 8). But the variance proportion corresponding to the smallest eigen value, we see that for OTGR, OHHS & OADR the variance proportions are 0.01, 0.89 & 0.02 respectively.

The estimated coefficients for the model are shown in Table 7. From Table 7, we observed that OTGR and OADR is positively influencing and OHHS is negatively influencing to OAGR. Thus, we may come to the conclusion that if OTGR and OADR are decrease and OHHS is increase then OAGR must be increase.

Conclusion

Ageing is the process of life. Population ageing in Bangladesh is becoming a serious concern for the development agendas. From the results, we find that relative ageing of the old aged population itself is becoming a notable feature of the ageing process in the LDC and Bangladesh. A larger fraction of the older population in the world is living in the LDC and Bangladesh and is growing faster. In Bangladesh, the proportion of persons aged 60 and over is expected to double between 2005 and 2030, from 5.7% to 10.6%.

In the regression model (1), OTGR(X11) is the first most influential variable, OHHS(X33) is the second influential variable, and OADR(X21) is the third influential variable, which explains 67% of the total variation in OAGR(X10). By using stepwise regression we concluded from this model if there is a decrease in the aged growth rate then we must decrease total growth rate and aged dependency ratio.

Aged population in the next decades will encounter different socio-economic circumstances and rapid technological change. Social integration is the best way to help develop under the changing situations. Aged populations are the respected senior citizens of our country. So, we all should come forward to meet their basic needs and provide the needed services for their well-being. Some policy could be taken to to develop their mental and socio-economic conditions are as follows:

1. The elderly people need security in their entire remaining life. This can be done through the social security system and by taking community based activities.
   
2. Elderly peoples' problem is assuming importance in Bangladesh because there will be a steep rise in its elderly population in coming years. So, it needs to change this negative attitude and Government, NGOs, communities and families should come forward.
   
3. For their standard living status, we need to provide more or less financial support and should be given to jobs like small business.
   

References

Bangladesh Bureau of Statistic (BBS). (1991). Population Census: 1991, Preliminary Report, Government of the People's Republic of Bangladesh, Dhaka.
………. (1994). Statistical Year Book of Bangladesh 1994.
.............(2001). Population Census: 2001, Preliminary Report, Government of the People's Republic of Bangladesh, Dhaka.

Bloom, D., David Cunning and Jaypee Sevilla (2002). The Demographic Dividend: A
New Perspective on the Economic Consequences of Population Change, RAND Publication available at http://www.rand.org/publications/MR/MR1274/ ,(Accessed on 30 November 2004).

Huber, B. (2005). Implementing the Madrid Plan of Action on Ageing. Chief,
generational issues and integration section, United nations Department of Economic and Social Affairs. Mexico, D.F., 31 August 2005.

Kabir, M.M. (1995). The Age and Aged in Bangladesh. Paper presented at the Fifth
National Statistics Conference of the Bangladesh Association, 27-28 January, Dhaka.

Kabir, M.M., Haque, M. and Chaklader, H. (2005). Mainstreaming ageing in health: Will
it be possible. Paper presented in the international conference on Mainstreaming ageing in health system and rural development, Held in Dhaka, November, 29-30, 2005.

Economic and Social Commission for Asia and the Pacific (ESCAP). (2004). Dynamics
of Population Ageing: How can Asia-Pacific Respond?. Expert Group Meeting on Development Issues and Policies, Bangkok.

Montgomery, Douglas, C. & Peck, Elizabeth A. (1982). Introduction to linear Regression
Analysis. New York: John Wiley and Sons.

Sattar., A.M. and Rahman, M.M. (1993). The Elderly in Bangladesh 1911-2025. Asian
Profile. Vol. 6.

Solomons, NW. Health and Ageing. In: Flores R. Gillepsie S. eds. Health and Nutrition:
Emerging and Remerging issues in Developing countries. Washington. DC. International Food Policy Research Institute, 2001.

United Nations (UN). (2003). World Population Prospects: The 2002 Revision.
Highlights. New York: UN, 2003.
……… (2004). World population ageing 1950-2050, United Nations, New York

……… (2005). World Population Prospects: The 2004 Revision, vol. 1:
Comprehensive Tables (Sales No. E.05.XIII.5); vol. II: Sex and Age Distribution of Population (Sales No. E.05.XIII.6).

World Health Organization (WHO). (2005). The World Health Report 2004 changing
History, 2004. Available at http//www.who.int/whr/2004, (Accessed 4 January 2005).

End note: OAGR(X₁₀) = Over all aged growth rate, OTGR(X₁₁) = Over all total growth rate, OHHS(X₃₃) = Over all household size and OADR(X₂₁) = Over all dependency ratio, RTGR(X₁₃) = Rural total growth rate, OIA(X₁₈) = Overall index of ageing, OTDR(X₃₂) = Overall total dependency ratio, OALRAP(X₃₄) = Overall literacy rate among aged population