3. THE COMBINED EFFECTS OF BIRTH RATE AND INCOME INEQUALITY ON ECONOMIC
3.1 Data, Operationalization, and Measurement
國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
3. THE COMBINED EFFECTS OF BIRTH RATE AND INCOME INEQUALITY ON ECONOMIC DEVELOPMENT
3.1 Data, Operationalization, and Measurement
In this investigation, the researcher seeks to provide explanations on the variation in poverty in LAC by focusing on the interactive effects of income inequality and birth rates. The researcher will conduct empirical analysis using quantitative data from 1960 to 2015 in LAC. The researcher will also conduct a case study of Honduras to show the importance of income inequality and birth rate on economic development. In the case study of Honduras, the researcher will perform interviews to review the opinion of the elites regarding the researcher’s hypothesis.
The 33 LAC countries used this investigations as sample are the following:
Antigua and Barbuda, Argentina, Bahamas, Barbados, Belize, Bolivia, Brazil, Chile, Colombia, Costa Rica, Cuba, Dominica, Dominican Republic, Ecuador, El Salvador, Grenada, Guatemala, Guyana, Haiti, Honduras, Jamaica, Mexico, Nicaragua, Panama, Paraguay, Peru, Saint Kitts and Nevis, Saint Lucia, Saint Vincent and the Grenadines, Suriname, Trinidad and Tobago, Uruguay, and Venezuela.
The researcher purposefully does not include the overseas territories located in the LAC region such as Anguilla, Aruba, British Virgin Islands, Cayman Islands, Curaçao, Montserrat, and Sint Maarten. Data from the LAC countries was retrieved from the World Development Indicators provided by the World Bank. These data consist of the most actual global development time series data.
‧
researcher used to test her hypothesis.The table is divided into three different kinds of variables. The first one: dependent variable, the second one: independent variable, and the third one: control variables. In each cell of the columns, we can find the indicator used by the researcher to measure the type of variable to which it corresponds.
Following this table, a more detailed description of each of the variables and its indicators can be found.
Table 3.1 Variables
Dependent Variable Independent Variables Control Variable
Economic Development
Employment in agriculture sector as a percentage of total employment.
Income Inequality
The indicator used: Gini index of inequality in equivalized household disposable income using Luxembourg Income Study data as the standard
Institutions
The Indicator used: polity2
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Interaction of Birth Rate and Income Inequality
Education
Government expenditure on education, total (% of GDP)
Corruption
The Indicator used: Political risk section from the International Country Risk Guide (ICRG).
Remittances
The Indicator used: Received personal remittances as a percentage of total GDP.
Population
Dependent Variable
The dependent variable for this research is economic development. The Cambridge dictionary defines economic development as “the process in which an economy grows or changes and becomes more advanced, especially when both economic and social conditions are improved.”
The indicator used to measure economic development in this research is the Log of GDP per capita (constant 2010 US$). The variable is operationalized by dividing gross domestic product by midyear population, resulting in GDP per capita. Data are in constant 2010 U.S. dollars. These
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
constant series show the data for each year in the value of 2010 as the base year. Data for GDP per capita constant at 2010 US$ was retrieved from the World Development Indicators in the World Bank.
Independent Variables
My research will have three independent variables: birth rate, income inequality and the interaction of birth rate and income inequality. The first independent variable, birth rate was used by Weintraub (1962), in a study of birth rate and economic development. Subsequent studies about this topic changed their variable birth rate for fertility. The Handbook of Vital Statistics Systems and methods (1991), defined the crude birth rate as “the number of live births occurring among the population of a given geographical area during a given year, per 1,000 mid-year total population of the given geographical area during the same year”, the same reference also defined fertility as
“the number of children ever born alive during the entire reproductive period of the woman.”
I chose birth rate instead of fertility because birth rate provides us with an actual number of births in the population of a determined country in a given year instead of providing us a number of how many births women could have during their reproductive age. This differentiation will allow us to test more accurately if there is any correlation between birth rate and economic development in the LAC region. The indicator for birth rate is crude birth rate. Crude birth rate tells us the number of live births in a given year, per 1,000 population estimated midyear. Data for crude birth rate was collected from the World Development Indicators from the World Bank.
My second independent variable is income inequality; it is operationalized as the estimate of Gini index of inequality in equivalized household disposable income using Luxembourg Income Study data as the standard. Data on the Gini index of inequality was gathered from the Standardized
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
World Income Inequality Database (SWIID) which provides the most-comparable data available on income inequality for cross-national research. From the interaction of birth rate and income inequality, we get our third independent variable.
Control Variables
In this research, I use three main control variables which are the following: education, industrial structure, and institutions. For the education variable, I use the indicator government expenditure on education which is operationalized as the government expenditure on education as a percentage of total GDP. Data for government expenditure on education also includes expenditure funded by international sources to the government. The data was collected from the World Development Indicators of the World Bank.
Regarding the second control variable which is industrial structure, I use employment in agriculture sector as a percentage of total employment as an indicator. The data is operationalized by including persons of working age engaged in the agriculture sector which consists of hunting, forestry, fishing, and of course agriculture. These productions of goods or services should always be in exchange for pay or profit. This data was also retrieved from the World Development Indicators of the World Bank.
The third control variable is the institutional variable. I use the variable polity2, which consists of a modified version of the polity variable for easier use of polity regime measures in time series analyses. This variable is operationalized by modifying the combined annual polity scores (the subtraction of autocracy scores to democracy scores, which results would be from strongly autocratic -10 to strongly democratic +10) and applying a fix to convert instances of standardized authority scores to polity scores. So in a simpler manner, transforming standardized authority
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
scores like the following: -66, -77, and -88, to conventional polity scores within the range, -10 to +10. Polity2 data was collected from the Polity IV project.
This research will use another set of two complementary control variables including corruption, and remittances. The fourth control variable corruption, is withdrawn from the political risk section from the International Country Risk Guide (ICRG). The variables include financial corruption, excessive patronage, nepotism, job reservations, favor-for-favors, close ties between politicians and businesses, and secret party funding. The corruption variable in the ICRG ranges from 0-6, higher values indicate lower levels of corruption. But to help on the better comprehension of my research and provide a more intuitive sense, the variable was transformed by subtracting it from 6.
Therefore, higher values now indicate higher levels of corruption. The data for corruption was collected from the International Country Risk Guide (ICRG) database.
Lastly, our fifth control variable is the received personal remittances as a percentage of total GDP.
The variable is operationalized as the sum of personal transfers and compensation of employees.
Personal transfers refer to all transfers made or received by resident households to or from nonresident households, in cash or in kind. On the other hand, compensation of employees refers to the income of border, seasonal, and other short-term workers employed in an economy where they are non-resident, and of residents employed by non-resident entities. Data on personal remittances as a percentage of total GDP was also retrieved from World Development Indicators of the World Bank.
Originally, the researcher wanted to cover 1,848 observations, which consisted of 33 countries with 55 years from the year 1960 to the year 2015. However, due to missing data for some of the variables included in the model, the research ends up with 255 observations.
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
A limitation experienced by the researcher in this study was the unavailability of data. This data unavailability is what reduced the observations to only 255. In order to aid the reader to have a clearer view about the observations available for each country per every variable, a table is provided below. Therefore, the table shows: in the first column the LAC country being observed and in the second column: the years the data is available for observation. Despite this big limitation, the researcher believes important to observe the results this data can provide. If we were to ignore the LAC’s region problematic every time due to data insufficiency, then no study that could serve as a starting guide to help the region would be developed.
Table 3.2 Observations included in the empirical analysis
Country Years
Argentina 1996-2013
Bolivia 1994-2003, 2006, 2008-2013
Brazil 1995, 1998 -2013
Chile 2000, 2002-2013
Colombia 1998-2013
Costa Rica 1995-1996, 1999-2004, 2006-2013 Dominican Republic 1993-1996, 2000-2003, 2007
Ecuador 1995, 1998-2000, 2009-2013
Guatemala 1993-1996, 2006-2013
Guyana 1994-1995, 1999-2007
Honduras 1994-1995, 2013
Jamaica 1991-1997, 2000-2004
Mexico 1991-1992, 1994-1995, 1998-2013 Nicaragua 1998-2000, 2002-2003, 2010
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Panama 1994-1997, 1999-2004, 2008, 2011
Peru 1993, 1995-2012
Paraguay 1998-2004, 2007, 2010-2012 El Salvador 1993-1995, 1997-1998, 2000-2010
Trinidad y Tobago 1994-2003
Uruguay 2001-2006, 2011
Venezuela 1992-1994, 2006-2007, 2009