The Effects of Race, Income, and Ideology on People’s Views on State Health Spending
Caleb Butts
Dr. Steve Shaffer
Research Methods
Fall 2007
(NOTE TO READERS: YOU NOW START A NEW PAGE OF THE PAPER)
Introduction
Over the last thirty years, the federal government has increasingly used state governments to accomplish tasks that it wishes to see carried out. These mandates, when properly funded, have become a stalwart on which many states depend upon to carry out necessary services. Because of the federal system within which our governments operate, the federal government can use its huge budget to pay for states to carry out these services for the state’s citizens. These funds can be used a variety in of ways, with two of the main funding destinations being highway upkeep by state Departments of Transportation and healthcare for near poor children and the poor through CHIPs and Medicaid, respectively.
Currently, Mississippi receives 76% of its total Medicaid and SCHIPs budget from the federal government (KFF 2007). This means that for every $2.40 the state of Mississippi dedicates to Medicaid, the federal government will kick in another $7.60—a more than tripling of the state Medicaid budget because of federal funds alone. This funding mechanism has caused a real battle in the state of Mississippi in the upcoming gubernatorial election. Republican Haley Barbour opposes a more expansive SCHIPs plan that would require nearly all children to participate in a child healthcare plan (Chandler 2007) and had already enacted some “Draconian” cuts on the state Medicaid plan earlier in his current term (Salter 2007). Democratic challenger, John Arthur Eaves, proposes that his child health insurance plan would bring health insurance to all Mississippi children, though he has made no mention of plans to change Medicaid enrollment.
In order to further examine the levels of public support for state funding for health care, I have proposed five hypotheses. Based on prior experience, I expect that blacks are more likely to have lower incomes compared to whites, people with lower incomes are more likely support an increase in health care spending compared to people with higher incomes, blacks are more likely to be liberal in self-identification compared to whites, liberal self-identifiers are more likely to support an increase in health care spending compared to conservative self-identifiers, and blacks are more likely to support an increase in health care spending compared to whites.
Model and Hypotheses
Assume that race is the earliest independent variable, income and ideology are intervening variables, and health spending is the dependent variable.
(NOTE TO READERS: INCOME AND IDEOLOGY SHOULD BE PLACED AT SPACES AT HEADS OF ARROWS IN DIAGRAM BELOW)
Income
(low)
Race (black) Health Spending (more)
Ideology (liberal)
The hypotheses are:
Hypothesis 1: Blacks are more likely to have lower incomes compared to whites.
Hypothesis 2: People with lower incomes are more likely to support an increase in state healthcare spending compared to people with higher incomes.
Hypothesis 3: Blacks are more likely to be liberal in self-identification compared to whites.
Hypothesis 4: Liberal self-identifiers are more likely to support an increase in state healthcare spending compared to conservative self-identifiers.
Hypothesis 5: Blacks are more likely to support an increase in state healthcare spending compared to whites.
Literature Review
Blacks are more likely to have lower incomes compared to whites.
The article “Race, Socioeconomic Status, and Health: The Added Effects of Racism and Discrimination” examines the role that racism and discrimination has played in negative health effects (Williams 1999). The author uses national data on race and socioeconomic status (SES) to show the direct and indirect effects that race has on health. Based on the study, there is a causal pathway from race to poor health outcomes via SES, which is supported by data that shows that blacks are more likely to have lower incomes than that of whites. In addition, this also adds an added control variable, that of health status, that can also explain an individual preference for increasing health spending.
Williams and Collins (2004) wrote a follow-up paper to the 1999 examination. This survey also used national data on mortality to illustrate the racial differences between health outcomes. Understanding that SES, and by definition income, is a major causal pathway to positive health outcomes and a driving force between the racial disparities, the authors lobby for reparations to be paid to African-Americans to level the playing field. This further encases race as a predictor of income levels.
A third study that draws conclusions on this relationship is “And How Are We Supposed to Pay for Health Care? Views of the Poor and Near Poor on Welfare Reform” (Schneider 2000). In this study, the author (an anthropologist) uses ethnographic methods and some quantitative analysis to examine the differences in race and income of those that receive means-tested government benefits. The author finds that those on welfare tend to be African American, making up about 37% of those on welfare, but only 12% of the US population as a whole.
People with lower incomes are more likely
to support an increase in state healthcare spending compared to people with
higher incomes.
The article “The Politics of Discretionary Medicaid Spending, 1980-1993” examines state level characteristics in conjunction with Medicaid spending, both optional recipient spending (creating eligible classifications that are not required by the federal government) and total spending (state funds plus federal matching funds) (Kousser 2002). While it is not the most important characteristic in projecting the level of state funding, state poverty rates are positively correlated with the level of Medicaid. This indicates that states with high levels of poverty are more likely to have higher levels of spending on Medicaid, though the causal pathway is not clearly defined. While it is possible that politicians are serving their constituents and representing them by passing beneficial legislation, it is probably just as likely that another reason exists and higher spending on Medicaid is not reflective of the policy preferences of individuals with lower incomes.
In the article “The End of Welfare as We Know It? The Effect of Welfare Reform on Racial Stereotypes and Welfare Attitudes”, Dyck et al examine what changes have occurred post-welfare reform and how this has altered individuals’ perceptions of racial stereotypes (2005). Of particular interest to this study is the role that income plays on welfare attitudes. The authors find that higher levels of income have increased opposition to welfare spending, a program similar to Medicaid, and that the income variable has the largest effect coefficient of any of the SES indicators in the regression model, as well as the third largest effect coefficient in the entire model. The authors also note that the direct effect income plays in opposition to increased spending, while consistently high in 1996 and 2000, had decreased slightly in 2000, producing a negative direct effect. The total effect of income, however, has remained consistently positive.
Most recently, in the article “Political Trust, Ideology, and Public Support for Government Spending,” the authors analyze the roles that an individual’s level of political trust and ideology plays in his support of increased government spending on particular programs (Rudolph and Evans 2005). Using data from the National Annenberg Election Survey, the researchers analyzed the data at both the aggregate and individual levels. Individual level analysis found that low income individuals tended to favor increased government spending on all programs, with Social Security, national health insurance, Medicare, Medicaid, and aid to mothers with children experiencing the highest levels of support.
Blacks are more likely to be liberal in self-identification compared to
whites.
Perhaps because of the overwhelming conclusions on race and political ideology, a review of the literature only produced two articles dealing with blacks likelihood of holding a particular ideology. The first article reviewed was “Race and Ideology: A Research Note Measuring Liberalism and Conservatism in Black America” (Seltzer and Smith 1985). Using data from the 1982 General Social Survey, the authors examined racial differences on self-identified ideology and policy preferences on a host of issues. The authors found that 42.9% of blacks identified themselves as liberals, while only 24.4% of whites identified themselves as liberals. The authors caution reading too much into these results, as self-identified ideology often bears little connection to actual policy preferences.
More recently, in “Ideological Realignment in the U.S. Electorate,” Abramowitz and Saunders examine the outcomes of the 1994 and 1996 elections as a result of a shift of voter bases and relative power of the two major parties (1998). Using data from the American National Election Studies, the authors looked at the percentage of Republican and Democratic Party identifiers and the racial, economic, and ideological composition of the two cohorts. The authors found that the two parties had become increasingly polarized, leaving liberal Republicans and conservative Democrats to switch parties. This indicates that those who identify themselves as Democrats will most likely identify as liberals, and those who identify themselves as Republicans will most likely identify as conservatives. Also, the Democratic Party lost a large share of its supporters, most of them white. African Americans, however, remained a consistent voting base with the party, losing only 4% of all African Americans, while the party lost 13% of its population as a whole between 1976 and 1996. This kept the percentage of black voters who identified themselves as Democrats above 90%. Because of the findings that the parties were increasingly polarized, it can be assumed that this reflects ideology as well.
Liberal self-identifiers are more likely to support an increase in state
healthcare spending compared to conservative self-identifiers.
In a study of the history of Medicare and Medicaid, Grogan and Patashnik argue that the legislation that produced Medicare and Medicaid were intentionally left vague. These legislative indefinites have created a program that has had difficulty in its expansion and is vulnerable in the future. In particular, the authors cite the growing chasm over expansion of Medicaid, with conservatives opposing any additional increases and liberals believing the system can be the foundation for a national health insurance.
Returning to previously mentioned articles, Kousser (2002), using state level data, found that legislatures where one or more houses were controlled by Democrats were more likely to have higher levels of spending on optional recipient groups. Again, though this does not technically mean that these individuals have a liberal ideology, based on party polarization from Abramowitz (1998), it is safe to assume some interchangeability.
Furthermore, Dyck et al (2005) found ideology to be the second largest determinant of overall spending, as liberal identifiers were less likely to oppose welfare spending. Similarly, Rudolph and Evans (2005) found that liberals tended to prefer higher spending on all government spending projects, though Medicaid and aid to mothers with children had the lowest levels of support.
Blacks are more likely to support an increase in state healthcare
spending compared to whites.
Returning again to the Seltzer and Smith (1985) article, using data from the 1982 General Social Survey, the authors also analyzed policy preferences that reflect the liberal-conservative cleavage, in addition self-identified political ideology. This method, analyzing preferences on particular policies, tends to be more reliable than self identified ideological measures. While both race categories found a majority of their members to prefer higher levels of healthcare spending, African Americans were 75.1% in favor of increased spending, compared to 57.0% among whites.
In an article by Deaton, “Health, Inequality, and Economic Development,” the author examines the role that relative income and economic development (particularly the cleavage between the epidemiological transition after the Industrial Revolution) plays on health. The author cites one article that shows that areas with more blacks tend to have higher state spending levels on healthcare.
Additionally, returning again to the Rudolph and Evans (2005) study, the authors found that African Americans preferred higher spending on Medicaid, Medicare, and health insurance for those that cannot afford it. Among blacks, these programs received the most support of any of the government spending programs, including defense and Social Security.
Alternatively, the Kousser (2002) article previously mentioned, used state level data to examine state characteristics on Medicaid spending. The study found that as the percent of the population that are minorities increases in a state, state level spending on Medicaid goes down. While part of this could be a compositional problem (the states with the largest population of African Americans tend to be Southern and very poor), it could also reflect attitudes that were held in the 1980s and early 1990s that those on social programs, such as welfare and Medicaid, tended to be African American and lazy (Dyck et al 2005). This sort of perception could have biased legislators to keep spending low so as to cover fewer people and take a “tough love” approach to those on social assistance.
Methodology
To test my model, I used information drawn from The Mississippi Poll project, a series of statewide public opinion polls conducted by the Survey Research Unit of the Social Science Research Center (SSRC) at Mississippi State University and directed by political science professor Stephen D. Shaffer. In order to maximize my sample size and therefore minimize my sample error, I combined or pooled telephone surveys conducted in two years—2004 and 2006. The 2004 Mississippi Poll surveyed 523 adult Mississippi residents from April 5 to April 21, 2000 and had a response rate of 48%, while the 2006 Mississippi Poll surveyed 574 adult Mississippi residents from April 3 to April 23, 2006 for a response rate of 50%. The two years combined contained 1097 respondents. With 1097 respondents, the sample error is 3.0%, which means that if every Mississippi voter had been interviewed, the results could differ from those reported here by as much as 3.0%. The pooled sample was adjusted or weighted by demographic characteristics to ensure that social groups less likely to answer the surveys or to own telephones were also represented in the sample in rough proportion to their presence in the state population. In both years, a random sampling technique was used to select the households and each individual within the household to be interviewed, and no substitutions were permitted. The SSRC's Computer Assisted Telephone Interviewing System (CATI) was used to collect the data.
I relied on four variables included in both years of the Mississippi Poll. Race was measured by asking the individuals to label themselves as white, black, or other. Income was measured by reported total family income before taxes in the year before each survey. Ideology was a self-identification question, asking respondents the following questions: "What about your political beliefs? Do you consider yourself very liberal, somewhat liberal, moderate or middle of the road, somewhat conservative, or very conservative?" Finally, health spending asked individuals whether they would like to see the state increase, decrease, or spend the same on healthcare.
In order to have enough people to analyze using multivariate tables, I recoded or combined categories of two of the variables. Eight income categories were recoded into three levels—low income was defined as families making less than $20,000 a year, middle income was considered as $20-40,000 per year, and high income included families making over $40,000 annually. Five ideological self-identification categories were combined into three groups—liberals included those considering themselves as "very" or "somewhat" liberal, conservatives were those identifying themselves as "somewhat" or "very" conservative, and the middle category of "moderate/middle of the road" constituted an intermediate "moderate" grouping. Race was a dichotomous variable of black or white. Preferences on state healthcare spending remained a three category grouping.
Findings – Bivariate
Table 1
Race and Income Differences
|
Gamma = -.569
Chi-Square <.001
Note: Percentages total 100% down each column
Source: 2004 and 2006 Mississippi Polls, conducted by Mississippi State University
Hypothesis 1 of my model states that blacks are more likely to have lower incomes compared to whites. In the 2004 and 2006 Mississippi Poll, 48.8% of blacks were in the under $20,000 category, compared to just 19.0% of whites. The magnitude of this relationship is 29.8%, which is the percentage difference between whites and blacks who earn less than $20,000. The magnitude reflected by the gamma value of this relationship is -.569. Furthermore, the chi-square statistic is significant at the .001 level, indicating that the relationship found between race and income in the 2004 and 2006 statewide polls can be generalized to the entire population. Thus, my hypothesis that blacks are more likely to have lower incomes compared to whites is accepted.
Table 2
Income and Preferences on State Healthcare
Spending Differences
|
State Health Spending |
Under $20,000 |
Income $20,000-40,000 |
Over $40,000 |
|
Less |
5.9% |
3.9% |
5.7% |
|
Same |
14.1% |
13.0% |
32.5% |
|
More |
80.1% |
83.1% |
61.8% |
|
N Size |
256 |
254 |
351 |
Gamma = -.315
Chi-Square <.001
Note: Percentages total 100% down each column
Source: 2004 and 2006 Mississippi Polls, conducted by Mississippi State University
Hypothesis 2 of my model is that people with lower incomes are more likely to support an increase in state healthcare spending than those with higher incomes. In the 2004 and 2006 Mississippi Poll, 80.1% of those making under $20,000 favored increased spending, while 61.8% of those earning over $40,000 favored increased spending. The magnitude of this relationship is 18.3%, the percentage difference between those who earn less than $20,000 and those that earn over $40,000 that favor increased state spending. The gamma value reflecting the relationship is -.315. The chi-squared statistic is significant at the .001 level, indicating that the relationship found between income and support for increased state healthcare spending found in the 2004 and 2006 statewide polls can be generalized to the entire population. My hypothesis that those who earn lower incomes are more likely to favor increased state spending on healthcare compared to those with lower incomes is also accepted.
Table 3
Race and Ideology Differences
|
Ideology |
Race White |
Black |
|
Liberal |
13.5% |
25.3% |
|
Moderate |
31.6% |
28.9% |
|
Conservative |
54.9% |
45.8% |
|
N Size |
667 |
332 |
Gamma = -.212
Chi-Square <.001
Note: Percentages total 100% down each column
Source: 2004 and 2006 Mississippi Polls, conducted by Mississippi State University
Hypothesis 3 of my model states that blacks are more likely to be liberal in self-identification compared to whites. Among white respondents, 13.5% identified themselves as holding a liberal ideology, while 25.3% of blacks self-identified themselves as liberal. The magnitude of this relationship is 11.8%, which is the difference between blacks and whites who identified themselves as liberal. The gamma value reflecting this relationship between race and ideology is -.212. In addition, the chi-squared statistic is significant at the .001 level, meaning that the relationship between race and ideology found in the 2004 and 2006 statewide polls can be generalized to the entire population. My hypothesis that blacks are more likely to be liberal in self-identification than whites is also accepted.
Table 4
Ideology and Preferences on State
Healthcare Spending Differences
|
State Health Spending |
Liberal |
Ideology Moderate |
Conservative |
|
Less |
1.7% |
7.1% |
4.5% |
|
Same |
9.7% |
17.6% |
28.1% |
|
More |
88.6% |
75.3% |
67.4% |
|
N Size |
175 |
324 |
509 |
Gamma = -.304
Chi-Square <.001
Note: Percentages total 100% down each column
Source: 2004 and 2006 Mississippi Polls, conducted by Mississippi State University
Hypothesis 4 of my model is that liberal self-identifiers are more likely to support an increase in state healthcare spending compared to conservative self-identifiers. In the 2004 and 2006 Mississippi Poll, 88.6% of liberal self-identifiers favored an increased in state healthcare spending, compared to 67.4% of conservative self-identifiers. The magnitude of this relationship is 21.2%, which is the percentage difference between liberal and conservative self-identifiers who support an increase in state healthcare spending. The gamma value that reflects this relationship is -.304. The chi-squared statistic is significant at the .001 level, meaning that the relationship found between ideology and support for increased state healthcare spending found in the 2004 and 2006 statewide polls can be generalized to the entire population. Thus, my hypothesis that liberal self-identifiers are more likely to support an increase in state healthcare spending compared to conservative self-identifiers is accepted.
Table 5
Race and Preferences on State Healthcare
Spending Differences
|
State Health Spending |
Race White |
Black |
|
Less |
5.9% |
.3% |
|
Same |
28.1% |
6.7% |
|
More |
66.0% |
93.1% |
|
N Size |
683 |
360 |
Gamma = .746
Chi-Square <.001
Note: Percentages total 100% down each column
Source: 2004 and 2006 Mississippi Polls, conducted by Mississippi State University
Hypothesis 5 of my model is blacks are more likely to support an increase in state health spending compared to whites. In the 2004 and 2006 Mississippi Poll, 93.1% of blacks supported an increase in state healthcare spending, compared to 66.0% of whites. The magnitude of this relationship is 27.1%, which is the percentage difference between blacks and whites who supported increased state healthcare spending. The gamma value reflecting this relationship is .746—our largest gamma value yet. The chi-squared statistic is significant at the .001 level, meaning that the relationship found between race and support for state healthcare spending in the statewide polls can be generalized to the entire population. Thus, my hypothesis that blacks are more likely to support increased state healthcare spending compared to white is also accepted.
Finding – Multivariate
Table 6
Race and Preferences on State Healthcare
Spending Differences
(Less than $20,000 Only)
|
State Health Spending |
Race White |
Black |
|
Less |
6.5% |
0.0% |
|
Same |
20.6% |
9.4% |
|
More |
72.9% |
90.6% |
|
N Size |
107 |
139 |
Gamma = .574
Chi-Square <.001
Note: Percentages total 100% down each column
Source: 2004 and 2006 Mississippi Polls, conducted by Mississippi State University
Table 7
Race and Preferences on State Healthcare
Spending Differences
($20,000-40,000 Only)
|
State Health Spending |
Race White |
Black |
|
Less |
6.3% |
0.0% |
|
Same |
18.8% |
2.2% |
|
More |
75.0% |
97.8% |
|
N Size |
160 |
90 |
Gamma = .873
Chi-Square <.001
Note: Percentages total 100% down each column
Source: 2004 and 2006 Mississippi Polls, conducted by Mississippi State University
Table 8
Race and Preferences on State Healthcare
Spending Differences
(Over $40,000 Only)
|
State Health Spending |
Race White |
Black |
|
Less |
6.5% |
0.0% |
|
Same |
37.1% |
8.9% |
|
More |
56.4% |
91.1% |
|
N Size |
291 |
56 |
Gamma = .778
Chi-Square <.001
Note: Percentages total 100% down each column
Source: 2004 and 2006 Mississippi Polls, conducted by Mississippi State University
Tables 6-8 are multivariate tables that control for the intervening variable of income. The findings show that among individuals with incomes less than $20,000, $20,000-40,000, and over $40,000, race is an important factor in determining an individual’s support for increased state healthcare spending. Blacks are consistently and significantly more supportive of increased state healthcare spending compared to whites. Among those with incomes of less than $20,000, 90.6% of blacks favored increased state healthcare spending compared to 72.9% of whites (Table 6). The magnitude of this relationship between race and support for state healthcare spending among those who earn less then $20,000 is 17.7%. The gamma value reflecting this relationship between race and support for state healthcare spending among those who earn less than $20,000 is .574. Furthermore, the chi-squared statistic is significant at the .001 level.
Among those with incomes between $20,000 and $40,000, 97.8% of blacks favored increased state healthcare spending compared to 75.0% of whites (Table 7). The magnitude of this relationship between race and support for state healthcare spending among those who earn between $20,000 and $40,000 is 22.8%. The gamma value reflecting this relationship between race and support for state healthcare spending among those who between $20,000 and $40,000 is .873. Furthermore, the chi-squared statistic is significant at the .001 level.
Among those with incomes over $40,000, 91.1% of blacks supported increased state spending healthcare compared to 56.4% of whites (Table 8). The magnitude of this relationship between race and support for state healthcare spending among those who earn over $40,000 is 34.7%. The gamma value reflecting this relationship between race and support for state healthcare spending among those who earn over $40,000 is .778. Finally, the chi-squared statistic is significant at the .001 level.
The multivariate tables 6-8 show a strong relationship between race and support for increased state healthcare funding, while income plays a role among whites. For example, 72.9% of whites from the lowest income level support increased state funding for healthcare, compared to 56.4% of whites from the highest income level. This is a difference of 16.5%.
At least 90% of blacks at all income levels support increased state spending on healthcare. 90.6% of blacks in the lowest income level support increased state funding for healthcare, compared to 91.1% of blacks in the highest income level. This is a percentage difference of a mere 0.5%—essentially no difference at all.
Table 9
Race and Preferences on State Healthcare
Spending Differences
(Liberals Only)
|
State Health Spending |
Race White |
Black |
|
Less |
1.1% |
1.2% |
|
Same |
14.8% |
3.6% |
|
More |
84.1% |
95.2% |
|
N Size |
88 |
83 |
Gamma = .564
Chi-Square <.05
Note: Percentages total 100% down each column
Source: 2004 and 2006 Mississippi Polls, conducted by Mississippi State University
Table 10
Race and Preferences on State Healthcare
Spending Differences
(Moderates Only)
|
State Health Spending |
Race White |
Black |
|
Less |
6.7% |
0.0% |
|
Same |
21.2% |
6.3% |
|
More |
72.2% |
93.8% |
|
N Size |
209 |
96 |
Gamma = .708
Chi-Square <.001
Note: Percentages total 100% down each column
Source: 2004 and 2006 Mississippi Polls, conducted by Mississippi State University
Table 11
Race and Preferences on State Healthcare
Spending Differences
(Conservatives Only)
|
State Health Spending |
Race White |
Black |
|
Less |
6.5% |
0.0% |
|
Same |
36.2% |
8.6% |
|
More |
57.3% |
91.4% |
|
N Size |
354 |
151 |
Gamma = .778
Chi-Square <.001
Note: Percentages total 100% down each column
Source: 2004 and 2006 Mississippi Polls, conducted by Mississippi State University
Tables 7-9 are multivariate tables that control for the intervening variable of ideology. The findings show that among individuals with liberal, moderate, or conservative ideologies, race is still an important factor in determining an individual’s support for increased state healthcare spending. Blacks are consistently and significantly more supportive of increased state healthcare spending compared to whites. Among those with a liberal ideology, 95.2% of blacks favored increased state healthcare spending compared to 84.1% of whites (Table 9). The magnitude of this relationship between race and support for state healthcare spending among those who are liberal self-identifiers is 11.1%. The gamma value reflecting this relationship between race and support for state healthcare spending among those who are liberal self-identifiers is .564. Furthermore, the chi-squared statistic is significant at the .001 level.
Among those with a moderate ideology, 93.8% of blacks favored increased state healthcare spending compared to 72.2% of whites (Table 10). The magnitude of this relationship between race and support for state healthcare spending among those who are moderate self-identifiers is 21.6%. The gamma value reflecting this relationship between race and support for state healthcare spending among those who are moderate self-identifiers is .708. Furthermore, the chi-squared statistic is significant at the .001 level.
Among those who are conservative self-identifiers, 91.4% of blacks supported increased state spending healthcare compared to 57.3% of whites (Table 11). The magnitude of this relationship between race and support for state healthcare spending among those who are conservative self-identifiers is 34.1%. The gamma value reflecting this relationship between race and support for state healthcare spending among those who are conservative self-identifiers is .778. Finally, the chi-squared statistic is significant at the .001 level.
The multivariate tables 9-11 show a strong relationship between race and support for increased state healthcare funding, while ideology plays a role among whites. For example, 84.1% of whites who self-identified as liberals support increased state funding for healthcare, compared to 57.3% of whites who self-identified as conservatives. This is a difference of 26.8%.
Regardless of ideology, at least 90% of blacks supported increased state spending on healthcare. 95.2% of blacks who self-identified as liberals support increased state funding for healthcare, compared to 91.4% of blacks self-identified as conservatives. This is a percentage difference of 3.8%—again, a small difference that indicates that blacks, as a whole, favor increased state spending on healthcare.
Conclusion
(NOTE TO READERS: INCOME AND IDEOLOGY SHOULD BE PLACED AT SPACES AT HEADS OF ARROWS IN DIAGRAM BELOW)
Income
(low)
Race (black) Health Spending (more)
Ideology (liberal)
Through this critical analysis of how race, income, and ideology affected individuals’ support for increased state healthcare spending, it can be concluded that all of these variables play an important role. After testing each hypothesis, all of the variables were significant. The multivariate tables show that while blacks seem to universally favor increased state healthcare spending, ideology and income play an important role for whites. Liberal whites and whites who earn less than $20,000 favor increased state healthcare spending much more than conservative whites and whites who earn more than $40,000. It is interesting to note the solidarity among blacks, while whites tend to have an opinion gradient.
While literature was not too difficult to find, there were two interesting issues that likely require further research. The article “The Politics of Discretionary Medicaid Spending, 1980-1993” examined state level data to examine the relationship between state characteristics and Medicaid spending. The authors found that there was a slightly negative relationship between Medicaid spending and states that had a high percentage of minorities. Based upon this analysis, if democracy represents the preferences of the people, it would seem that states with higher percentages of minority populations, particularly black population, would have a higher level of Medicaid spending. Closer analysis should examine whether those populations voices are being heard in state legislatures, as should be expected with a representative form of government or if this is simply a compositional problem—that states with a higher percentage of minority population are also the poorest and cannot spend as much as other states with lower levels of minority population.
In addition, the article “Race, Socioeconomic Status, and Health: The Added Effects of Racism and Discrimination” showed that there was a significant relationship between health status and preference for higher levels of spending on social programs, such as Medicaid. Given that blacks and the poor are more likely to have a worse health status, further analysis should examine whether the racial and economic relationships found in regards to support of increased state healthcare spending are actually a spurious relationship with health status.
Furthermore, much of the analysis found African Americans to be supportive of social programs, though not to the extent found in this review. While this could indicate that blacks in Mississippi overwhelmingly prefer increased state healthcare spending, it could also indicate that certain blacks prefer certain healthcare programs (i.e. Medicaid and SCHIPs, or Medicare—which is not controlled by the federal government) and the effect seen in this analysis was an aggregation of the many policy preferences in one loosely defined, vague survey question.
Finally, contrary to articles finding that polarization between the parties has tightened up inner-party ideological variations, we also see that liberal ideology may not definitively indicate a Democratic voter in Mississippi. Mississippians polled in these two survey years showed that conservatives held a at least a plurality in Mississippi, if not a majority. The state, however, remains a battleground state, and blacks, where 45.8% identified themselves as conservative, vote overwhelmingly as Democrats. This indicates that the relationship between ideology and party identification may not be as strong in Mississippi as it is in other parts of the country.
While these relationships appear to be conclusive, further research is always necessary to examine the changes in attitudes over time. In addition to the above recommendations, further research should also take other factors into account and pay particular attention to how the changes in the economy affect policy preferences. If Deaton is correct, and relative wealth is more important than actual wealth, these trends should persist regardless of inflation. With the cost of healthcare constantly on the rise (Catlin et al 2007), the issue of preference regarding state spending levels and what groups prefer these issues will continue to play a large role in the development of health policy.
References
Abramowitz, AI, and KL Saunders. 1998. Ideological realignment in the U.S. electorate. The Journal of Politics 60(3):634-652.
Catlin, A, C Cowan, S Heffler, B Washington, and the National Health Expenditure Accounts Team. 2007. National health spending in 2005: the slowdown continues. Health Affairs 26(1):142-153.
Chandler, Natalie. 2007. No new twists in second Barbrour-Eaves debate. The Clarion-Ledger, October 5, 2007.
Deaton, A. 2003. Health, inequality, and economic development. Journal of Economic Literature 41:113-158.
Dyck, JJ, L Hussey, and LF Williams. 2005. The end of welfare as we know it? The effect of welfare reform on racial stereotypes and welfare attitudes. Paper presented at American Politics Workshop, March 2005, University of Maryland.
Grogan, C, and E Patashnik. 2003. Between welfare medicine and mainstream entitlement: Medicaid at the political crossroads. Journal of Health Politics, Policy, and Law 28(5):821-858.
Kaiser Family Foundation. 2007. State Health Facts. http://www.statehealthfacts.org (accessed October 8, 2007).
Kousser, T. 2002. The politics of discretionary Medicaid spending, 1980-1993. Journal of Health Politics, Policy, and Law 27(4):639-671.
Rudolph, TJ, and J Evans. 2005. Political trust, ideology, and public support for government spending. American Journal of Political Science 49(3):660-671.
Salter, Sid. 2007. Gimmicks won’t solve Medicaid, SCHIP woes. The Clarion-Ledger, September 30, 2007.
Schneider, JA. 2000. And how are we supposed to pay for health care? Views of the poor and near poor on welfare reform. American Anthropologist 101(4):761-782.
Seltzer, R, and RC Smith. 1985. Race and ideology: a research note measuring liberalism and conservatism in Black America. Phylon 46(2):98-105.
Williams, DR. 1999. Race, socioeconomic status, and health: the added effects of racism and discrimination. Annals of the New York Academy of Sciences 896:173-188.
Williams, DR, and C Collins. 2004. Reparations: a viable strategy to address the enigma of African American health. American Behavioral Scientist 47(7):977-1000.