Enright 1
    
    Michael F. Enright The Effect of Income on Economically Related Bill Sponsorship michael.enright@mansfield.oxon.org +01 (408) 564-1900 21 May 2009
    
    Enright 2
    
    Abstract:
    
 In a time of economic turmoil for both the middle and working classes, it can be
    
    assumed that the issue of economic progress would be of great saliency to legislators and voters. 2008 saw one of the most challenging years for the global financial industry. These challenges continue into 2009 and the term of the 111th United States Congress. 
 This paper examines if there is a relationship between economic status of a
    
    representativeʼs district and economic legislation activity of its representative. This is done by estimating the effect of a districtʼs average constituent income on the number of bills the representative sponsors that are aimed to: stimulate the local economy, restructure government spending and taxation, make infrastructure improvements, protect American industry, improve the financial industry, maximize employment, or improve healthcare access. No effect of income on the number of economically related bills sponsored is found. It is concluded that both high-income and low-income districts are equally likely to have a representative that sponsors economically related bills.
    
    Enright 3
    
    Introduction:
    
 On January 28, 2009, the U.S. House of Representatives passed the American
    
    Recovery and Reinvestment Act of 2009 (ARRA). The bill aimed to make “... supplemental appropriations for job preservation and creation, infrastructure investment, energy efficiency and science, assistance to the unemployed, and State and local fiscal stabilization...”1 This piece of legislation was the first bill to be voted on by the 111th Congress. It was a much awaited response to the financial crisis that reached relatively severe levels in 2008. 
 The Democrats strongly favored the bill while the Republicans unanimously
    
    opposed it.2 In the final roll call vote, only 11 Democrats out of 255 in the House voted against the bill, while all 177 voting Republicans opposed ARRA. The bill passed with 244 to 188. At first glance, the vote results may seem to obey party lines. The objectives of this bill raise another question. The question of whether the income of constituents motivates their representative to pursue such types of legislation. 
 ARRA was passed as a response to financial crisis. One would expect those
    
    most likely to be affected by economic cycles to be most in favor of this type of legislation. That is, those who have a lower income as would indicate workers of less-
    
    1 2
    
    Thomas Thomas
    
    Enright 4
    
    0
    
    20
    
    Frequency 40
    
    60
    
    )DSU( stcirtsid larotcele niergnniCgnrena fo ubiirtsitsidemmqenI )DSU( stcirtsid larotcele niergnniCgnrena fo ubiirtsitsidemmqenI ss emoco e a i v oit no tub r D yc1eer0grn ss emoco e a i v oit no tubiir D yc1eer0grnii : n0o0ci0 : n0o0ci0 00u 06 00u 06 00004 00004 0u 002 0u 002 e 0 o02 e 0 o02 0 c04 0 c04 08 08 06 06 F F F F 20000 40000
    
    Income Distribution in Congress
    
    60000 income Figure 1: distribution of average income in electoral districts (USD)
    
    80000
    
    valued and easily replaced skills. Figure 1 shows the distribution of of the average income of constituents in electoral districts of the House. The mean personal income is $43,433 and the standard deviation is $10,975. 
 To deduce if income correlates with ʻyeaʼ or ʻnayʼ votes on the bill, one can divide
    
    the district average incomes into quantiles and compare the number of ʻyeaʼ and ʻnayʼ
    
    Enright 5
    
    15
    
    10
    
    Vote Split 5
    
    0
    
    -5
    
    aeelstnvitisop eelitoaunIeytnni wTAyRAstilstilps etov i:l2SRug6aeelstnvitisop eelitoaunIeytnni wTAyRAstilstilps etov i:l2SRug6s y i e auQ , m n c q moc e yb R b no pS etoV p er R4V s y i ii e auQ , m n c q moc e yb R b no pS etoV p er R4iiV t A et01t A et0108 08 97 97 86 86 75 75 64 64 53 53 42 42 31 31 21 21 11 11 05 05 50 50 o9 o9 21 21 5F 5F A A 2 2 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 2 5 5 4 4 0 0 1 1 0 0 5 5 2 2 2 2 1 1 2 2 1 1 6 6 1 1 2 2 3 3 1 1 1 1 14 12 6 5 3 2 1 1 2 1 2 2 0 1 -6
    
    ARRA Vote Splits by Twenty Income Quantiles
    
    4
    
    5 2
    
    0 -1
    
    1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Figure 2: vote splits on ARRA by income quantile, positive is yea
    
    votes in each. Figure 2 presents the difference between ʻyeaʼ and ʻnayʼ votes on ARRA organized by 20 quantiles of average income. Quantiles 1 and 2 represent the lowestincome districts. They demonstrate a strong split in favor of the bill, having 14 and 12 more ʻyeaʼ than ʻnayʼ votes respectively. The split in quantiles 3 through 20 is far less predictable. It varies from a vote split of 6 in favor of the bill to a split of -6 opposing the bill in quantile 7.
    
    Enright 6
    
    Vote Split 20
    
    30
    
    10
    
    0
    
    aey si evitisop ,lettniiuq emocni cnIAyRAstilstilps etov i:l3SRugiV aey si evitisop ,lettniiuq emocni cnIAyRAstilstilps etov i:l3SRugiV e i n u Q emo yb R b no pS etoV p er R10 e i llii n u Q emo yb R b no pS etoV p er R10 t A et01 t A et01 03 03 02 02 o1 o1 02 02 A A 5 5 4 4 3 3 F F 1 1 6 6 7 7 2 2 3 3
    31 7 6 2
    
    ARRA Vote Splits by Income Quintile
    
    10
    
    1 2 3 4 Figure 3: vote splits on ARRA by income quintile, positive is yea
    
    5
    
    If one reduces the the highly varied quantiles in Figure 2 to quintiles of average constituent income, the trend appears more smooth. Figure 3 presents vote splits between ʻyeaʼ and ʻnayʼ votes on ARRA as defined by income quintiles. Once again, the strong tendency for lower quintiles to favor the bill is observed in the lowest quintile. 31 more representatives supported the bill than opposed it. All else held constant, it appears reasonable to assume low income of a district is a strong indicator of the districtʼs representative pursuing legislation for economic stimulation. 
 This paper tests the hypothesis that average district income influences the
    
    number of economically related bills the districtʼs representative sponsors. Data is drawn from Brone and Ujifusaʼs Almanac of American Politics (2008). The total number
    
    Enright 7
    
    of economically related bills for a sample of 86 representatives in the 111th Congress is taken and the number of economically related bills tabulated. Little correlation is found and it is concluded that the hypothesis that economic bill sponsorship is dependent on district income requires further testing if it is to be proven.
    
    Theory and Model:
    
 It is commonly thought that the preferences of the median voter are a good
    
    predictor for what sorts of policy will be supported by the constituentʼs representative.3 Meanwhile, it is not commonly agreed upon that economic fluctuations influence the short-run opinions of voters.4 However, it has often been observed that negative economic fluctuations will lower voterʼs opinions of a policy-maker. This occurs if the policy makerʼs economic competence, or lack thereof is perceived as somewhat responsible for the economic damage.5 Financial crises leave congressmen responsible for acting or not acting through fiscal policy. Thus, it is reasonable to assume the saliency of the financial crisis of 2008 would influence the preferences of the median voter against a representative who did not act against the conditions. 
 To define the median voter, this analysis uses the income of the average voter,
    
    the average personal income of the constituency. The ARRA was by far the highest profile economic stimulus bill. Therefore, it could have been tread differently than others. We wish to see if the relationship between very low district income and support for the ARRA can be applied to economically related legislation in a broader sense. To
    
    3 4 5
    
    Mathis 404 Kramer 133 Norpoth 953
    
    Enright 8
    
    find out if this trend is isolated to the ARRA or is a general tendency, one must first count the economically related bills of representatives. In this study, bills are classified as economically related if they aim to meet any of the following objectives: • stimulating the local economy • restructuring government spending • making infrastructure improvements • protecting American industry • protecting home ownership • changing regulations of the financial industry • changing taxation policy • maximizing employment • increasing access to healthcare Some of the above objectives may be appealing to the lower income groups and some of them may be appealing to the higher income groups. Both types of legislation are included in our count to ensure both effects from high and low-income groups will be counted. Though the voting results of the ARRA suggest a correlation between low income groups and economically related bills, it is possible that the opposite correlation is true. That is, higher income groups could demand economic protection. Therefore, bills addressing issues such as reducing government spending and decreasing taxes, things which normally appeal to a wealthier and more conservative constituency, must fit into the economic bill count. 
 To properly investigate the correlation between income and the total bill count, a
    
    number of control variables are required. It seems likely that party affiliation is strongly
    
    Enright 9
    
    associated with sponsorship, given the strong partisan voting lines on ARRA. A variable for whether or not the representative is a Democrat is included in the model. Population could also lead to higher income and more economically concerned citizens in a city, so population is included. Likewise, race and occupational category could be correlated with income and economic legislation demand. Variables noting ethnicity and if the respondent identified as white collar are included. Finally, a variable for what percentage of the constituency is elderly is included as age may correlate with income and political demands. It may have been desirable to include regional dummies but given the limited sample size of 86 representatives6 for investigation, this is not possible. Total Economic Bills 0 1 2 3 4 5 6 7 13 Frequency 27 16 16 13 5 2 5 1 1
    
    Table 1: frequency of economic bill sponsorship 
 The mean number of economic bills sponsored is 1.942, as shown in Table 1.
    
    The mode of economic bill sponsorship is zero but most representatives sponsored
    6
    
    See Analysis for explanation of sampling method.
    
    Enright 10
    
    more than that. Because of a nonzero mean, it is appropriate to use a standard OLS regression instead of a censored model such as a Tobit regression.7 The relationship to be tested in the analysis is shown below:
    
    econ _ billsi = β 0 + β incomeincome + βX i + ui
    (1) The term βXi represents the control variables discussed above while the coefficient βincome is the variable to be analyzed. If there is a relationship between the districtʼs average income and the number of economic bills sponsored, βincome ≠ 0.
    
    Existing Literature:
    
 Current literature on the ARRA is sparse because it is so recent. Though many
    
    studies that suggest a link between income and legislative behavior have been conducted. 
 Barret and Cook (1991) offer a model most closely resembling the one presented
    
    herein. Barret and Cook sampled 58 members of the House of Representatives and examined the relation between preferences for social welfare and voting behavior for social welfare legislation. Like the model in this paper, Barret and Cook equate socioeconomic status of the district with the family income of a district. They also use the Almanac of American Politics. The conclusion is that district factors do have an effect on voting behavior on social welfare issues. The most significant effect is that of party affiliation. However, I incorporate variables for party affiliation in the manner that Barret and Cook do in an attempt to test the significance of income effects.
    7
    
    The results of various regression types are available in the source data and coding for this project.
    
    Enright 11
    
    
    
    Jackson and Kingdon (1992) offer support for the use of average factors of the
    
    constituency influencing policy action. Jackson and Kingdon attempt to define the ideology of a legislator and how it contributes to voting factors. The most significant figure found is that of mass opinion. 
 Both the Barret and Jackson models are supported by the work of Mathis and
    
    Zech (1986), which aims to support the model of the median voter. Mathis and Zech attempt to model the median voter and compare the importance of a number of factors in determining what makes an accurate picture of the median voter. When predicting the median voter, they find median income of a district outperforms other measures. With such an emphasis on the importance of income, one could infer that income determines policy support among representatives. The analysis of this paper takes a specific angle on Mathis and Zechʼs analysis by testing if income levels can predict economic legislation patterns. 
 Lewis-Beck and Paldam (2000) not only use mass opinion as a tool to explain
    
    legislative behavior but also support the idea that issue salience leads to such opinion being reflected in legislation. The authors affirm the salience hypothesis that voters punish party X if it is in power when x goes wrong. Of particular importance is that they confirm voterʼs awareness of economic issues spiking close to election times. 
 The findings of Lewis-Beck and Paldam suggest issues of salience and issues of
    
    an economic nature must be of particular concern to a representative who wants to stay in office. The ARRA and subsequent legislation has been made against an issue of salience, so the results of this analysis will serve as an evaluation of Lewis-Beck and Paldamʼs findings.
    
    Enright 12
    
    
    
    Norpoth (1987) also finds the trends discussed in Lewis-Beck and Paldam.
    
    Norpoth examines levels of public support for Thatcher in response to two variables, her contractionary of monetary policy and the Falkland Islands War. It is found that voters usually pay more attention to negative results of policy than positive ones, supporting a concept similar to the saliency hypothesis. When monetary contraction was implemented in the UK in the early 1980s, it was estimated that a vote cost of about 10 points to the incumbent in the following election resulted. Meanwhile electoral gains from economic stabilization were minimal. Norpoth adds credibility to the concept of saliency. It follows that representatives of districts with negative economic experiences would be sponsor economically stimulating legislation. By doing so they would lower the cost of allowing negative economic cycles to affect their constituents. 
 Norpothʼs findings are in agreement with Kramer (1971), who suggests that while
    
    changes in unemployment do not affect electoral outcomes, changes in per capita income do. Kramer finds that a 10% decrease in per capital real personal income would cost incumbent administrations 4 to 5% of the congressional vote. Kramerʼs conclusion is contested by Stigler (1973) who not only argues that unemployment is insignificant but also that economic fluctuations have no effect on the congressional vote. 
 The work of Erikson (1990) is in agreement with Stigler and finds no link between
    
    economic downturn and congressional elections. Erikson concludes that while many believe economic conditions determine electoral outcomes, there is little evidence to support this claim. The paper contends that with examination of macroeconomic data, there is no statistically significant link between the performance of the economy and congressional vote. If there is no link between congressional elections and the
    
    Enright 13
    
    economy, and those running for election are aware of this, Eriksonʼs findings suggest βincome will not be significantly different from zero. 
 Aside from investigating if economic conditions transfer to the congressional
    
    vote, it is also important to examine the sensitivity of public opinion to economic changes. If the public is reasonably aware of economic changes and reacts to them, our eventual results of a zero value of βincome are meaningful. If the public opinion reacts to economic fluctuation, the question of whether or not this reaction is catered to in congressional bill sponsorship leads to insight of the strategies used by legislators to stay in office. If the public never cared about economic downturns, then any null result in the above discussed literature or this project is not useful. If the public never cared about economic downturn, there is no reason to expect a rational policy maker would react to income levels. Most literature on the topic of public opinion reaction suggests a strong reaction to economic conditions, making the question of whether or not this transfers into congressional action a useful topic. 
 Shapiro (1989) presents a literature review that suggests the public will regret
    
    policy changes that are of negative economic consequence. The main piece of evidence is referred to as a public opinion ʻbacklashʼ in the 1980s after the monetary experimentation of Reagan. At first public support for monetary contraction was high but after unemployment was observed, the public opinion polls reflected a reversion to egalitarian goals. In such a situation, average income of a constituency would go down, so if this finding were true and there was a link between public reactions to economic downturn, one would expect βincome to have a nonzero solution.
    
    Enright 14
    
    
    
    In concurrence with Shapiro is the work of Listhaug and Alaberg (1999).
    
    Listhaug investigates public opinions regarding income inequality in post-communist transitional states of Eastern Europe. They test to see if public opinion reacted against the democratization process by putting a renewed emphasis on egalitarian goals after the income gaps of the post-communist era developed. They conclude that the lower classes are more sensitive to economic consequences of policy change and desire increased government intervention. If Listhaug and Alabergʼs findings are able to be applied to the U.S., one might expect to find an inverse relationship between income and economic bill sponsorship. 
 Another relevant study is that of Kelly (2004) where income effects on public
    
    opinion are investigated. Kelly aims to test the hypothesis that a transition from a socialist state to a free market economy changes support for income inequality. The identification strategy used focused on asking individuals how much they thought was a legitimate income for various jobs. He finds that income inequality support in the postcommunist countries of Eastern Europe is considerably higher than in countries of the Western world. Since Kelly argues Western democracies have a stronger affinity for egalitarian goals, his argument lends support to the argument that income levels may cause members of the House who serve lower income districts to more actively pursue economic legislation to satisfy the constituency. 
 This paper fits into the existing literature by testing the hypothesis that income
    
    relates to economically related legislation but tests this trend against at a time of economic disturbance. Most of the literature from Eastern European transitions suggests a strong causal effect between economic disturbance and public opinion
    
    Enright 15
    
    change, making the necessity of studying legislative effects under economic stresses clear. 
 Most of the literature on American politics that studies links between income and
    
    electoral outcomes suggests whether or not it would be advantageous for a representative to cater to a desire for economic improvement in order to garner votes in the next election. 
 By examining the effect of district income on economic bill sponsorship at a time
    
    when economic downturn is especially concerning to the public, the gap between economic voting at normal times and at times of economic change illustrated by the two groups of literature is bridged. A nonzero value of βincome would illustrate that the public opinion over economic change was mounting and that goals of the public were changing while the representative aimed to cater to this opinion shift. We know the changes discussed in most of the literature of economic transitions are happening, albeit in a less extreme form, suggesting that public opinion is changing but we do not know if this shift translates to electoral changes that would ultimately determine the behavior of a competitive representative. 
 Bridging the gap between opinion change resulting from issue salience and
    
    behavior in legislation is important because it assesses the link between electoral votes and behavior required of representatives in order to remain elected. If there is no relationship between income levels of a district and the number of economically related bills sponsored, one can infer legislators donʼt cater to mass opinion at times of unusually high issue salience. Therefore they do not perceive a failure to do so as an electoral risk.
    
    Enright 16
    
    Analysis:
    
 To preform the economic bill tabulation, a sample of 86 representatives is
    
    selected, consisting of either 17 or 18 representatives from each quintile of district income levels. The mean income of the sample is $43,744 with a standard deviation of $11,287. The mean constituent income of the entire House of Representatives is
    )DSU( elpmas ni eopubiaSid emocni uneiutitsinocemmqenI )DSU( elpmas ni eopubiaSid emocni uneiutitsinocemmqenI n l it m rts ni noit t b rts D yc4eer0gr0 n l it m rts ni noit t b rts D yc4eer0gr0 : n0o0cni : n0o0cni 00u 01 00u 01 00051 00051 0u 00i5 0u 00i5 e 0 o02 e 0 o02 0 c02 0 c02 06 06 04 04 8 8 F F F F 0 5 Frequency 10 15 20 20000 40000
    
    Income Distribution in Sample
    
    60000 income Figure 4: constituent income distribution in sample (USD)
    
    80000
    
    $43,433 with a standard deviation of $10,975. When Figure 4 is compared to Figure 1, it is clear that the distribution of incomes in the sample closely approximates that of the House as a whole. The extreme left end of the curve is omitted in the sample but this should not obscure the relationship between very low income and economic bill sponsorship as there were only four districts with an average income less than $26,672,
    
    Enright 17
    
    the lowest income represented in our sample. This omission could eliminate the lowest quantile of income in a 20 quantile division but it leaves the second lowest quantile that still had unusually high support for the HRRA, as seen in Figure 2. 
 The party distribution in the sample differs from that of the House but this is not
    
    relevant to the results. The sample is composed of 53.49% Democrats and 46.51% Republicans. The actual house composition is 59.03% Democrats and 40.97% Republicans. While the difference in gap between the sample and the actual population is 11.08%, the party variable is not significantly related to the number of economically related bills sponsored as seen in Table 2. Therefore the difference in gap size should not be of concern. Coefficients Yielded From Equation 1 Variable income democratic party district population % elderly in district % white ethnicity in district % white collar residents in district Constant Coefficient .0000295 (.0000369) .5017 (.5309) 5.67x10-06(9.15x10-06) 20.25 (10.29)* -3.0408 (1.5138)** -3.8296 (5.0550) -1.441 (6.325) 0.427 0.348 0.537 0.053 0.048 0.451 0.82 P-Value
    
    R2=0.1034, Adjusted R2=0.0353, N=86, *=significant at 10% level, **=significant at 5% level Table 2: Results form equation 1. 
 Table 2 presents the results of the regression described by Equation 1. Average
    
    income of a district is not a statistically significant factor in predicting the number of
    
    Enright 18
    
    economically related bills sponsored by a representative. It does seem, however, that the percentage of those who are elderly in the district has a strong correlation on economic bill sponsorship. For each tenth of a precent the proportion of elderly in the district increases, one can expect 2 additional economically related bills. Perhaps this is because bills of economic objectives related to healthcare were included in the tally. It could also be due to increased political activism among the elderly community. Also significant is percentage of voters in a district who are identified as being of the ʻwhiteʼ ethnicity. For each additional percentage of whites in the district population, one can expect 3 to 4 fewer economically related bills. While it appears that the elderly and ethnicity variables explain most of the variation in economic bill sponsorship, it is important to note the very low R2 value. It is only 0.1034 and .353 after being adjusted for the inclusion of all explanatory variables. The large amount of insignificant explanatory variables clearly explains the gap between the R2 and adjusted R2 values. 
 Given the low number of control variables possible in this model due to a small
    
    sample size, it is necessary to do further testing to ensure that βincome is not significantly different from zero. Figure 5 shows the scatter plot of district income on the number of
    
    Enright 19
    
    total number of economic bills sponsored 0 5 10 15
    
    pihsrosnops llib cimonoce dna eresUopcnlIcb ictmrtnohe garsvbop0malgB pihsrosnops llib cimonoce dna eresUopcnlIcb ictmrtnohe garsvbop0malgB d )DS n(o s s li n nci opiid e ro ee am5eer0tioF d )DS n(o s s li n nci opiid e ro ee am5eer0tioF o m emo d i a s c s f o r n : u000cni o m emo d i a s c s f o r n : u000cni 00055 00055 0000i0 0000i0 n0u 01 n0u 01 S o01 Sll o01 l 02 l 02 04 04 8 8 6 6 t t
    
    Bill Sponsorship and Income
    
    20000
    
    60000 80000 income Figure 5: average district income (USD) and economic bill sponsorship
    
    40000
    
    economically related bills sponsored by a representative in the sample. Visually, one can see a large concentration of incomes around the mean of $43,433 that is centered around 2 bills sponsored. Among the higher half of income districts, there seems to be a few more observations with higher bill tallies. Therefore, it is appropriate to run a difference of means test between the upper and lower half of incomes. 
 The sample was divided into two quantiles of income and the mean number of
    
    economically related bills compared. A difference in the predicted number of economic bills was only 0.163 between the two groups and had an insignificant p-value of 0.7303. This test was repeated comparing only the uppermost and lowermost quintile of income. This test returns an expected difference of 0.905 bills but the p-value is still insignificant
    
    Enright 20
    
    at 0.3339. Even when only looking at districts with extremely high and low incomes, the trend observed under the ARRA vote does not hold and low income does not cause high economic bill sponsorship. 
 Since the sample size is small due to limited research resources, it is necessary
    
    to do yet more comparison before dismissing the hypothesis that bill sponsorship of economically related bills is related to a districtʼs average income. Table 3 shows a cross-tabulation of the number of economically related bills sponsored by representatives, organized by the 20 quantiles of income.
    Number of Economically Related Bills By Income Quantile Quantile
    Count
    
    1 0 0 2 1 0 0 1 0 0 4
    
    2 0 3 0 0 0 0 0 0 1 4
    
    3 2 1 1 1 0 0 0 0 0 5
    
    4 0 1 1 1 0 0 1 0 0 4
    
    5 2 0 0 1 1 0 0 0 0 4
    
    6 2 0 0 3 0 0 0 0 0 5
    
    7 1 1 1 0 1 0 0 0 0 4
    
    8 1 1 1 1 0 0 0 0 0 4
    
    9 2 1 1 1 0 0 0 0 0 5
    
    10 11 12 13 14 15 16 17 18 19 20 2 1 0 0 1 0 0 0 0 4 0 1 1 2 0 0 0 0 0 4 2 1 0 0 0 1 0 0 0 4 2 2 0 0 0 0 0 0 0 4 2 1 1 0 0 1 0 0 0 5 0 0 3 0 0 0 1 0 0 4 0 0 2 0 1 0 1 0 0 4 4 0 0 0 0 0 1 0 0 5 2 0 1 0 0 0 0 1 0 4 1 1 1 1 0 0 0 0 0 4 2 1 0 1 1 0 0 0 0 5
    
    0 1 2 3 4 5 6 7 13 Total
    
    Table 3: Tabulation of economic bill count over 20 quantiles of income.
    
    Enright 21
    
    As one can see, the number of bills only varies between 4 and 5 in each quantile. Even with a sample of just 86 observations, this trend is very strong and not likely to be obtained in such a consistent pattern if the trend is not true for the population. Especially since a sample of 86 observations covers 20% of the House. 
 Having ensured that the sample is not only representative of the House as a
    
    whole but also having ensured that a trend exists to the contrary of βincome ≠ 0, we dismiss the hypothesis that income is a determinant of economically related bill sponsorship.
    
    Conclusion:
    
 We have concluded that no relationship between income and economically
    
    related bill sponsorship exists but it is important to analyze why this relationship does not exist, especially since much of the literature suggests otherwise. 
 The percentage of elderly and white constituents in a district appears to be a
    
    significant determinant of economic bill sponsorship. It is hardly disputed that older people have higher political efficacy than younger ones. Perhaps elderly people who are on a fixed income are concerned about economic cycles causing damage to the purchasing power of a fixed income. This concern could be reflected in their political demands, causing representatives to pursue economic improvement for this reason instead of the income levels of the district. 
 As for the race variable, it is possible that white people have an easier time
    
    gaining political efficacy than minorities do and their concern over economic issues is therefore more likely to be heard in the House.
    
    Enright 22
    
    
    
    It would be beneficial to add an interaction term to Equation 1 that interacts race
    
    and age variables with political efficacy. This may capture a more reliable predictor of economic bill sponsorship if it is the political efficacy of groups that causes these two group-oriented variables to become significant. The income variable addresses all constituents, not just those of particular efficacy. Further supporting the idea of efficacy is the insignificance of the white collar variable. Aside from labor unions, most occupational statuses, especially white collar ones, may not have as consolidated of political efforts as groups defined by age and race do.8 The absence of efficacy in the occupational group and its insignificance may warrant further investigation of a way to integrate efficacy into the model. 
 Finally, albeit inevitably, there is some vulnerability for measurement error that
    
    could have resulted in a false rejection of the hypothesis that βincome ≠ 0. Inclusion of economic bills related to health care could have caused a spurious correlation between the elderly variable and the total bill count. Further investigation with a refined search criteria could create differing results. 
 It does seem counterintuitive that no relationship between income and economic
    
    bill sponsorship exists but this confusion could also be due to an inappropriate indicator. It is possible that the nominal income of a district does not determine the economic legislation demands of the district. Perhaps the constituents are accustomed to lower incomes and donʼt find having a below-average income troubling because their income does not differ form those in close proximity. If the test were run with the change in average income before elections replacing the income variable, a completely different
    8
    
    The alternative regressions in the analysis files available with this paper suggest a blue collar variable also has no significant effect.
    
    Enright 23
    
    story could emerge9 as this would express not an absolute measure but a measure of loss or gain. It is highly possible that an incorporation of change into the model instead of the static variable used would capture a different emotional effect on the constituency. Though it is hard to believe one could live in one of poorest districts in the country and not notice or care that they were so much worse off than their others.
    
    9
    
    Jacobson 160
    
    Enright 24
    
    References Barrett, Edith J. Fay Cook. Congressional Attitudes and Voting Behavior: An Examination of Support for Social Welfare. Legislative Studies Quarterly, 1991. Vol 16. No.3, pp. 375-392. Barone, Michael; Grant Ujifusa (1998). The Almanac of American Politics. Washington, D.C.: National Journal Group. (ed. James Cottrill, Santa Clara University) Edward J. Mathis, Charles Zech. An Examination into the Relevance of the Median Voter Model: Empirical Evidence Offers Support for the Model of Certain Uses. American Journal of Economics and Sociology, 1986 Vol. 45, No. 4, pp. 403-412. Erikson, Robert S. Economic Conditions and the Congressional Vote. American Journal of Political Science, 1990; Vol. 34, No. 2, pp. 373-399. Jacobson, Gary G. The Politics of Congressional Elections (7th). San Francisco: Pearson, 2009. Kelly, Jonathan. Economic Change and the Legitimation of Inequality: The Transition From Socialism to the Free Market In Central-East Europe. Research in Social Stratification and Mobility. 2004. Vol. 22, pp. 319-364. Kramer, Gerald. Sort-Term Fluctuations in U.S. Voting Behavior, 1896-1964. the American Political Science Review. Vol. 65, No. 1 (1971), pp. 131-143. Listhaug, Ola. Toril Alaberg. Comparative Public Opinion on Distributive Justice: A Study of Equality Ideals and Attitudes toward Current Policies. International Journal of Comparative Sociology, 1999; 40; 117. Library of Congress, "HR 1 Bill Information." Thomas. 21 May 2009 .
    
    Enright 25
    
    Norpoth, Helmut. Guns and Butter and Government Popularity in Briain. The American Political Science Review, Vol. 81, No. 3 (Sep., 1987), pp. 949-959. Shapiro, Robert Y. John T. Young. Public Opinion and the Welfare State: The United States in Comparative Perspective. Political Science Quarterly, Vol. 104, No. 1 (Spring, 1989), pp. 59-89. Stigler, George.General Economic Conditions and National Elections. The American Economic Review, 1973; Vol. 63, No. 2, pp. 160-167.

Add Comment

To add a comment, you need to login or signup (you can signup without leaving this page)