Dependence of Demographic Factors on GDP


Published on

  • Be the first to comment

  • Be the first to like this

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide

Dependence of Demographic Factors on GDP

  1. 1. qwertyuiopasdfghjklzxcvbnmqwertyuiopasdfghjklzxcvbnmqwertyuiopasdfghjklzxcvbnmqwer Dependence of Demographic Factors on GDPtyuiopasdfghjklzxcvbnmqwerty Anant Damle (PGP-11-102)uiopasdfghjklzxcvbnmqwertyuiopasdfghjklzxcvbnmqwertyuiopasdfghjklzxcvbnmqwertyuiopasdfghjklzxcvbnmqwertyuiopasdfghjklzxcvbnmqwertyuiopasdfghjklzxcvbnmqwertyuiopasdfghjklzxcvbnmqwertyuiopasdfghjklzxcvbnmqwertyuiopasdfghjklzxcv
  2. 2. Introduction Total Population and Labour ParticipationThis paper looks at the impact of birth rate on Rate for a particular year is a long term effectlong term GDP of countries throughout the of historical Birth The GDP is also affected by Health andThis study aims to establish the effect of birth Education which has actually created thatrate on GDP. It also tries to capture it due to Demographic dividend.Health, Education, Urbanisation,Infrastructure and Labour parameters. Literature The media and academic circle is gung-hoBackground about the growth story of India and China andTraditionally, GDP is expressed as the contribution of demographics. There was no conclusive study as to how and what is the exact impact of this to GDP of countries.Where, Various parameters have been chosen from aGDP Gross Domestic Product list of 175 available indicators. The details ofC Private Consumption each indicator can be found in the MetadataI Gross Investments sheet of the Excel file. Table 1 describes theG Government Spending/Expenditure variable names and Labels in brief.X Exports Rationale of Variable SelectionM Imports The indicators chosen are representative of the sectors the effect of which the study aimsThe effect of demographics, Infrastructure, to investigate.Health, Education and Labour parameters is Dependent Variablecaptured in the equation indirectly. Like, due In GDP in terms of constant 2000 USD. Thisto better Labour, the production will be high, has been done to remove the effect ofresulting in more Exports; or due to larger Exchange rate fluctuations of country’snumber of working population, the Private currency.consumption and Gross Investments will behigher and likewise. Health Health Expenditure per capita, PPP (atDemographic Dividend is a parameter which constant 2005 USD) is an indicative of thecan be defined as a combination of Total general health of the population. Good healthPopulation and Labour Participation Rate1.1 Expressed as % of total population aged 15+ 2years Birth Rate (crude, births per 1000 persons)
  3. 3. of population is pre-requisite for high Other variables initially considered wereproductivity and hence it is considered. Number of Physicians (per 1000), number of beds available etc. but were later dropped asTable 1 Variable Code and Indicator Names they didn’t make sense.Variable Code Year Indicator NameNY.GDP.MKTP.KD.2008 2008 GDP (constant Manufacturing 2000 US$) The manufacturing has a large impact on GDP Birth rate, crudeSP.DYN.CBRT.IN.1983 1983 (per 1,000 and has already been explored. The indicators people) Urban of manufacturing are Goods transported by populationSP.URB.GROW.2008 2008 rail, Overall Logistic Index etc. growth (annual %)SL.TLF.TOTL.IN.2008 2008 Labour force, Consumption total Electricity Consumed (kWh) is a gross Unemployment,SL.UEM.TOTL.ZS.2008 2008 total (% of total indicator of private consumption. labour force) Labour Demographics participationSL.TLF.CACT.ZS.2008 2008 rate, total (% of A human being comes into the labour force total population ages 15+) after being educated / trained around the age Health between 15 – 25 years. Hence, Birth Rate with expenditure perSH.XPD.PCAP.PP.KD.2008 2008 capita, PPP a lag of 15 – 25 years is more relevant. (constant 2005 international $) Median Age of population etc. Electric powerEG.USE.ELEC.KH.2008 2008 consumption The Birth-Rate lag of 25 years was finalised (kWh) Railways, goods after 6 iterations, of regression between GDPIS.RRS.GOOD.MT.K6.2008 2008 transported (million ton-km) and Birth-Rate with lags of 1, 5, 10, 15, 20 and Logistics 25 Yrs. Highest explanatory power was found performanceLP.LPI.OVRL.XQ.2009 2009 index: Overall in a lag of 25 Yrs., and hence it used in the (1=low to 5=high) study. Median Age ofSP.DYN.MEDIAN.AGE 2010 the Population (in Years) Data Collection Data was collected from Two (2) differentHealth SourcesHealth Expenditure per capita, PPP (atconstant 2005 USD) is an indicative of the 1. World Bankgeneral health of the population. Good health World Development Indicators andof population is pre-requisite for high Global Development Financeproductivity and hence it is considered. Databank 2. The Fact Book by CIA
  4. 4. Methodology Factor AnalysisThe theory / objective of study is to identify On running the factor analysis, the KMO (high)the relation between factors described and and Bartlett’s test (Sig .000) supported thetheir effect on GDP of a country. notion that data is good for factor analysis.Initially, Only Birth Rate was considered and All MSA values in the Anti-Image Correlationregression was run for varying amount of lag were found to be > 0.5 and also theperiods. The lag period of 25 was best suited communalities for all variables was higherfor exploration. Hence birth rate data is for than 0.5.year 1983 Variable were factored into Three (3) factorsOther indicators provided logical relation only as per Table 2. The factors seem logical andto the current value and hence their current hence no variable was dropped. There is novalues for year 2008 was considered independent variable left after factor formation. The factors were then constituted based on the scaled (to 1) co-efficient matrix.Model BuildingA data table was constituted consisting of all Cluster Analysisindicators in Excel. It was found that out of A hierarchical cluster analysis was performed208 countries under observation; only 85 had to identify any groups in countries. The clusterall the required data points. These 85 analysis has resulted in primarily 3 clusters,observations were transferred to usable USA being an exception. The clusters weresheet. logical and hence considered. Table 2 Factor Components Factor # Factor Name Variables 1 Demographic Birth rate (crude, per 1000) Factor Urban population growth (annual %) Health expenditure per capita, PPP (constant 2005 international $) Overall Logistic Index Median Age of the Population 2 Industrial Factor Labour force, total Electric power consumption (kWh) Railways, goods transported (million ton-km) 3 Labour Factor Unemployment, total (%) Labour Participation Rate (%)
  5. 5. Table 3 Country Clusters from Cluster Analysis ConclusionCluster-3 Underdeveloped but fast It is observed that Demographic Dividend growing countries with an does play an important role in the GDP. exception of CHINA. The point of concern is absence of LabourCluster-2 Developed Nations mainly Factor for GDP explanation. Europe, including Australia and Japan It seems to the author that some variablesCluster -1 Developing countries including which should have been considered were India. 𝐺𝐷𝑃 𝑈𝑆𝐷 𝑐𝑜𝑛𝑠𝑡𝑎𝑛𝑡 2000 1.8 × 1011Regression Analysis 6.057 × 𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑖𝑎𝑙 𝐹𝑎𝑐𝑡𝑜𝑟Multiple step-wise regression was performed 2.7 × 108 × 𝐷𝑒𝑚𝑜𝑔𝑟𝑎𝑝ℎ𝑖𝑐 𝐹𝑎𝑐𝑡𝑜𝑟 absent and hence Labour Factor didn’t havebetween GDP and Identified factors. The final sufficient explanatory power.regression model didn’t contain LabourFactor. The explanatory power wasacceptable at 80.7% (adjusted R2) References 1. catalog/world-development- indicators 1) Model-2 incorporates two factors 2. Industrial and Demographic and has viewer/pdfviewer?sid=da9ebe64- an R2 of 0.80 which is extremely eb95-40c1-825c- good. As the model is able to explain 087fb2505a47%40sessionmgr4&vid=2 80.7% of variance of the dependent &hid=24 variable. 3. 2) The p-value of the constant (0.049) /articleshow/11918847.cms which is marginal hence considering 4. significant. Using the Unstandardized ountries_by_median_age Coefficients (B), to form the relation. 5. 3) Clusters did not have better ons/the-world- explanatory power and hence factbook/fields/2177.html individual regression models were not constructed.