POLS 207 1st Edition Lecture 5 Outline of Last Lecture I State Relationships and Differences Outline of Current Lecture II Correlations A Graphs available on lecture slides III Issues in Empirical Analyses a Measurement b Variable linkage c Is there a relationship IV State Governments Current Lecture Correlations The strength of a relationship is measured on a scale from 1 1 Educational Attainment If you graduate from high school you are more likely to graduate from college Positive relationship with a score of 49 Percentage Metropolitan and Per Capita Income Relationship seems to suggest that people who live in urban areas will also have higher incomes The type of occupations that are available in the city have a higher income than in rural areas More urban a state is the higher the income level Percentage Metropolitan and Abortion Rate r 67 This means that girls who live within cities are more likely to obtain an abortion than those who live in rural areas There is more anonymity than in rural areas and it is also easier to obtain an abortion Therefore there is more opportunity in this respect Also the idea that in the city there is the proximity to people and women are more likely to find themselves in such a situation However correlation does not always mean causation The percentage of abortions also depends on how restrictive the laws are in that state Percentage Metropolitan and Murder Rate r 35 There is a quite weak correlation between urban environments and a high murder rate but it does exist States such as Louisiana and Mississippi have a much higher murder rate than what is predicted These states suffer from violent crime Texas is slightly above the average but very close to the predicted point Per Capita Income and State and Local Taxes There is a very strong relationship r 84 This means the more that people earn the more that they will be taxed However this means that the more you are taxed the more you will gain from this The states that tax the most spend the most on public services Therefore states such as Texas have very little taxes but the state also spends very little on their residents Per Capita State and Local Nontax revenue and Total Revenue Nontax revenue is where the state gains money from nontax sources These are things such as gun licenses college tuition and fees library membership speeding fines They are not taxes but they make money for the local government The relationship between the two is positive r 94 so the more nontax revenue there is the greater the total revenue will be Many states have high nontax revenue Texas is one of these This means that tax rates are actually quite low because the revenue comes from other fees for services that the public wants Per Capita State and Local Revenue and Expenditure Strong positive relationship r 83 This means that the higher a states revenue is the more they will spend on services States with older populations need more spending on services emergency services healthcare etc States with an older population will have a lower population of children There is a negative relationship between these two variables r 75 Texas has about 25 of its population between 5 17 and 10 of it s population over 65 years There is extremely close to the regression line States such as Florida have a high population of people over 65 years and only about 15 of the population is school age children Per Capita Income and Poverty 2010 moderate negative relationship between the two variables r 61 If per capita income increases the amount of those in poverty decreases Connecticut have a high per capita income and a low poverty rate they are on the regression line Texas is an outlier although the per capita income is fairly high the poverty rate is higher than it should be as predicted Mississippi has a low per capita income and an extremely high poverty rate around 25 Utah is a good outlier because it has a lower per capita income but also a low rate of poverty Infant Mortality and Life Expectancy Strong negative relationship r 82 The higher the infant mortality rate the lower the life expectancy States such as Mississippi have a high infant mortality rate and low life expectancy age States that have a low infant mortality rate and higher life expectancy are Florida and Hawaii Texas is below the regression line meaning that people are not living as long as they should According to the infant mortality rate the life expectancy should be 79 5 years when it is actually 79 years Per Capita Income and Poverty 2010 If you divide the graph into quadrants you get a better analysis of the data States such as NY and California have High Income and High Poverty Texas is also included in this High Income and Low poverty states are Connecticut and Massachusetts etc Low Income and High Poverty states are Louisiana and Mississippi etc most of the Southern states are in this category Low income and low poverty states are Utah Maine etc Utah has a low income because of the high number of children that are not earning in their state Violent Crime and Incarceration There is no relationship r 06 Many people are incarcerated for drug related offences but these figures are about to drop in states that have just legalized Marijuana People are not being incarcerated for just violent crime it is a myriad of offences and therefore there is no correlation Business Climate and Quality of Life Ranking positive weak relationship r 32 States that have a strong business climate will also have a better life quality for their residents Issues in Empirical Analyses Measurement Variable Linkage Is there a relationship With correlation you cannot measure cause and effect therefore you must use more sophisticated tools Casual Relationship A relationship in which a causes b There is a clear link that can be defined Such as high levels of education and high levels of income A better education leads to a higher income not true for all cases though Spurious Relationship The correlations are very high but there is no causal relationship Correlation does not mean causation This can be because of an intervening variable or an unknown variable that links the two Divided Partisan Control and Medicare Payments to Disabled All States When you collect the data there seems to be a correlation r 42 So this relationship suggest that when there is divided government the higher the Medicare payments to the disabled If you take out the Southern states from the data there is no