Lab7
David Montalvo
10/24/2019
Polarization of Congress Members by State
This report will investigate the polarization of the states in regards to the Economic Liberalism-Conservatism policies of congress members based on the different states through the country. I believe this is important due to how the study will show the different levels of polarization based on the state from where congress members are, providing important information on how defined the ideologies in different states are or not.
Trends
StateFilteredRep <- CongressData %>% filter(statenm == "SOUTH D" | statenm == "NORTH D" | statenm == "TENNESS"|statenm == "WYOMING" |statenm == "ALABAMA" | statenm == "KANSAS" | statenm == "TEXAS" | statenm == "IDAHO" | statenm == "NEBRASK" | statenm == "MONTANA", ptycode =="200")
head(StateFilteredRep)## # A tibble: 6 x 150
## # Groups: year, chamber, rep_or_dem [1]
## cong idno sc cd statenm ptycode name dw1 dw2 votescast
## <dbl> <dbl> <dbl> <dbl> <chr> <dbl> <chr> <dbl> <dbl> <dbl>
## 1 80 3644 63 1 IDAHO 200 GOFF… 0.241 -0.253 138
## 2 80 8193 63 2 IDAHO 200 SANB… 0.386 -0.283 139
## 3 80 1920 32 1 KANSAS 200 COLE… 0.244 -0.19 132
## 4 80 8318 32 2 KANSAS 200 SCRI… 0.428 -0.224 139
## 5 80 6449 32 3 KANSAS 200 MEYE… 0.333 -0.0690 144
## 6 80 7799 32 4 KANSAS 200 REES… 0.297 -0.181 136
## # … with 140 more variables: errors <dbl>, gmp <dbl>, fips <dbl>,
## # postal <chr>, statenum <dbl>, senate <dbl>, ptyunity <dbl>,
## # wptycount <dbl>, ptyvotescast <dbl>, s_days <dbl>, h_days <dbl>,
## # s_hrs <dbl>, h_hrs <dbl>, s_pls <dbl>, h_pls <dbl>, s_pass <dbl>,
## # h_pass <dbl>, s_report <dbl>, h_report <dbl>, conf_repts <dbl>,
## # s_intro <dbl>, h_intro <dbl>, tot_pls <dbl>, tot_pass <dbl>,
## # tot_report <dbl>, tot_intro <dbl>, tot_hrs <dbl>, htos_hrs <dbl>,
## # htos_pls <dbl>, htos_pass <dbl>, htos_report <dbl>, htos_intro <dbl>,
## # h_pctpass <dbl>, s_pctpass <dbl>, tot_pctpass <dbl>, enactments <dbl>,
## # pages <dbl>, pgsperbill <dbl>, vetoes <dbl>, regveotes <dbl>,
## # pocketvetoes <dbl>, cloturevotes <dbl>, cloturewins <dbl>,
## # landsqmi <dbl>, age65 <lgl>, black <dbl>, farmer <dbl>, finance <lgl>,
## # forborn <dbl>, gvtwrkr <dbl>, manuf <dbl>, populatn <dbl>,
## # urban <dbl>, veterans <dbl>, popsqmi <dbl>, blucllr <dbl>, city <dbl>,
## # coast <dbl>, dc <dbl>, enroll <dbl>, port <dbl>, rurlfarm <dbl>,
## # union <dbl>, vabeds <dbl>, unemplyd <dbl>, constrct <dbl>,
## # whlretl <dbl>, statwrkr <lgl>, fedwrkr <lgl>, loclwrkr <lgl>,
## # mdnincm <lgl>, civlab <lgl>, transprt <dbl>, emplyd <dbl>,
## # latino <lgl>, forbornpct <lgl>, stcd <lgl>, highsch <lgl>,
## # pcthighsch <lgl>, pctage65 <dbl>, pctblack <dbl>, pctfarmer <dbl>,
## # pctfinance <lgl>, pctforborn <dbl>, pctgvtwrkr <dbl>, pctmanuf <dbl>,
## # pcturban <dbl>, pctveterans <dbl>, pctunemplyd <dbl>,
## # pcttransprt <dbl>, pctlatino <lgl>, pctmiltpop <lgl>, abstenpct <dbl>,
## # misspct <dbl>, dpres <lgl>, dv <dbl>, chairrm <dbl>, leadership <dbl>,
## # speaker <dbl>, nrsc <lgl>, …ggplot(StateFilteredRep, aes(statenm, dw1)) + geom_boxplot() + labs(title = "Republican Polarization by State", x= "Economic Liberalism-Conservatism", y= "Count") +theme(plot.title = element_text(hjust = 0.5))
In the graph above, the polarization of the Republican Party in Congress is demonstrated based on different states where the Republican party is the favorite. AS it can be seen, there is a case of polarization in the congress members of the states as they are closer to 1 and additionally a few atypical values that are close to zero or negative values on the economic liberalism-conservatism ideology. The state which presents more polarization on the republican party is Texas as it shows a median value of around 0.65, along with high values of some congress members that even reach or surpass 1. Additionally, it can be seen how North Dakota is the state with the least polarization in regards to members of the congress ideologies, with a mean of around 0.1, demonstrating very low values in comparison to the rest of the states.
StateFilteredDem <- CongressData %>% filter(statenm == "RHODE I" | statenm == "DELAWAR" | statenm == "NEW YOR"|statenm == "MARYLAN" |statenm == "NEW JER" | statenm == "ILLINOI" | statenm == "CALIFOR" | statenm == "CONNECT" | statenm == "MASSACH"| statenm == "VERMONT", ptycode =="100")
head(StateFilteredDem)## # A tibble: 6 x 150
## # Groups: year, chamber, rep_or_dem [1]
## cong idno sc cd statenm ptycode name dw1 dw2 votescast
## <dbl> <dbl> <dbl> <dbl> <chr> <dbl> <chr> <dbl> <dbl> <dbl>
## 1 80 5531 71 1 CALIFOR 100 LEA … -0.0380 0.41 126
## 2 80 2965 71 2 CALIFOR 100 ENGL… -0.0740 0.119 126
## 3 80 4201 71 4 CALIFOR 100 HAVE… -0.344 -0.207 146
## 4 80 6483 71 6 CALIFOR 100 MILL… -0.377 0.167 129
## 5 80 2908 71 10 CALIFOR 100 ELLI… 0.142 0.396 123
## 6 80 2689 71 14 CALIFOR 100 DOUG… -0.518 -0.216 135
## # … with 140 more variables: errors <dbl>, gmp <dbl>, fips <dbl>,
## # postal <chr>, statenum <dbl>, senate <dbl>, ptyunity <dbl>,
## # wptycount <dbl>, ptyvotescast <dbl>, s_days <dbl>, h_days <dbl>,
## # s_hrs <dbl>, h_hrs <dbl>, s_pls <dbl>, h_pls <dbl>, s_pass <dbl>,
## # h_pass <dbl>, s_report <dbl>, h_report <dbl>, conf_repts <dbl>,
## # s_intro <dbl>, h_intro <dbl>, tot_pls <dbl>, tot_pass <dbl>,
## # tot_report <dbl>, tot_intro <dbl>, tot_hrs <dbl>, htos_hrs <dbl>,
## # htos_pls <dbl>, htos_pass <dbl>, htos_report <dbl>, htos_intro <dbl>,
## # h_pctpass <dbl>, s_pctpass <dbl>, tot_pctpass <dbl>, enactments <dbl>,
## # pages <dbl>, pgsperbill <dbl>, vetoes <dbl>, regveotes <dbl>,
## # pocketvetoes <dbl>, cloturevotes <dbl>, cloturewins <dbl>,
## # landsqmi <dbl>, age65 <lgl>, black <dbl>, farmer <dbl>, finance <lgl>,
## # forborn <dbl>, gvtwrkr <dbl>, manuf <dbl>, populatn <dbl>,
## # urban <dbl>, veterans <dbl>, popsqmi <dbl>, blucllr <dbl>, city <dbl>,
## # coast <dbl>, dc <dbl>, enroll <dbl>, port <dbl>, rurlfarm <dbl>,
## # union <dbl>, vabeds <dbl>, unemplyd <dbl>, constrct <dbl>,
## # whlretl <dbl>, statwrkr <lgl>, fedwrkr <lgl>, loclwrkr <lgl>,
## # mdnincm <lgl>, civlab <lgl>, transprt <dbl>, emplyd <dbl>,
## # latino <lgl>, forbornpct <lgl>, stcd <lgl>, highsch <lgl>,
## # pcthighsch <lgl>, pctage65 <dbl>, pctblack <dbl>, pctfarmer <dbl>,
## # pctfinance <lgl>, pctforborn <dbl>, pctgvtwrkr <dbl>, pctmanuf <dbl>,
## # pcturban <dbl>, pctveterans <dbl>, pctunemplyd <dbl>,
## # pcttransprt <dbl>, pctlatino <lgl>, pctmiltpop <lgl>, abstenpct <dbl>,
## # misspct <dbl>, dpres <lgl>, dv <dbl>, chairrm <dbl>, leadership <dbl>,
## # speaker <dbl>, nrsc <lgl>, …ggplot(StateFilteredDem, aes(statenm, dw1)) + geom_boxplot() + labs(title = "Democratic Polarization by State", x= "Economic Liberalism-Conservatism", y= "Count") +theme(plot.title = element_text(hjust = 0.5))
The graph above, illustrating the polarization of the states where the Democrat party is favorite states, demonstrates a polarization contrary to the one showing the Republican polarization. In comparison to the previous graph, the members of congress for the Democrat party show polarization levels of values lower than 0, with some atypical values being very close to it. Even though, polarization in the democrat part is not as extreme as it was demonstrated in the Republican one. As it can be seen, the state which represents the most polarization, is California with a economic liberalism-conservatism ideology mean value of about 0.42. On the contrary, the state with the least value for this index is Delaware, which has about 0.3 in its median polarization value based on the ideology. It can also be seen how there is not much variation in the polarization of the different states as all mean values surround the 0.4 mark, whilst the Republican graph demonstrates how a greater variance.
StateFilteredDem %>% ggplot(aes(dw1)) + geom_histogram() + labs(title = "Democratic Polarization", x="Economic Liberalism-Conservatism")
As it is shown in the histogram regarding the polarization of the Democratic party in the selected states, it supports the previous claims in which the most of the polarization values in the Economic Liberalism-Conservatism are around the 0.4 value, demonstrating polarization, yet not in an extreme manner, rather moderately, as it is shown how it is rather closer to a neutral policy rather than a very marked ideology.
StateFilteredRep %>% ggplot(aes(dw1)) + geom_histogram() + labs(title = "Republican Polarization", x="Economic Liberalism-Conservatism")
In this histogram, which illustrates the polarization of the republican congress members in the selected states it is shown more variatio in comparison to the Democratic party one. At a first glance, it can be seen how there is not such a bell form as it can be observed in the “Democratic Polarization” table. Instead, there is a variation in the polarization values from about 0.1 to 0.7 where most of the congress members range in terms of their economic ideology.
Causes
For the analysis of the causes of the polarization of the members of congress, I decided to further research the states in which the Republic and Democratic parties are the most popular. Additionally, in order to reduce the bias when picked the states, I randomized the pick by creating an excel spreasheet and having it select 10 states for each of the parties. The hope of the study is to prove the role that the state of origin of the congress members has a big role in their ideologies and biases, furthermore in the polarization of the liberalist-conservative ideologies.
ModelDem<-summary(lm(StateFilteredDem$dw1 ~ StateFilteredDem$ptyunity + StateFilteredDem$sc + StateFilteredDem$wptycount))
ModelDem##
## Call:
## lm(formula = StateFilteredDem$dw1 ~ StateFilteredDem$ptyunity +
## StateFilteredDem$sc + StateFilteredDem$wptycount)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.87937 -0.05568 0.01178 0.06729 0.28182
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 4.867e-01 2.014e-02 24.168 < 2e-16 ***
## StateFilteredDem$ptyunity -9.691e-03 2.294e-04 -42.242 < 2e-16 ***
## StateFilteredDem$sc -3.270e-04 6.577e-05 -4.973 6.94e-07 ***
## StateFilteredDem$wptycount 4.018e-05 7.196e-06 5.583 2.56e-08 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.1018 on 3285 degrees of freedom
## Multiple R-squared: 0.3709, Adjusted R-squared: 0.3704
## F-statistic: 645.7 on 3 and 3285 DF, p-value: < 2.2e-16Based on the linear model demonstrated above in which; the congress member’s ideology, the percentage of votes casted in cogress, the numerical code for the states and the number of votes from the members it can be seen how for the Democrat party, these variables do not play a big role in their polarization for their ideologies. As it can be seen, the adjusted R^2 value is 0.3704 meaning that a very low percentage of the member’s ideologies is affected by the analyzed variables. This goes against the purpose of the study, due to the hope that there is a strong correlation in which the state from where the congress member is from plays an important role in their ideologies.
ModelRep<-summary(lm(StateFilteredRep$dw1 ~ StateFilteredRep$ptyunity + StateFilteredRep$sc + StateFilteredRep$wptycount))
ModelRep##
## Call:
## lm(formula = StateFilteredRep$dw1 ~ StateFilteredRep$ptyunity +
## StateFilteredRep$sc + StateFilteredRep$wptycount)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.38818 -0.09209 -0.01714 0.08081 0.89464
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) -5.656e-01 3.586e-02 -15.772 < 2e-16 ***
## StateFilteredRep$ptyunity 8.562e-03 4.323e-04 19.805 < 2e-16 ***
## StateFilteredRep$sc 2.490e-03 3.560e-04 6.996 4.24e-12 ***
## StateFilteredRep$wptycount 3.447e-04 1.613e-05 21.368 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.1406 on 1275 degrees of freedom
## Multiple R-squared: 0.5861, Adjusted R-squared: 0.5851
## F-statistic: 601.8 on 3 and 1275 DF, p-value: < 2.2e-16The linear model in which the Republican party’s ideology is analyzed throgh the same variables as the Democrat model, proves distinct results. At first, it can be seen how the R^2 value is significantly higher than the Democrat one, at 0.5851 (0.21496 higher than the previous model), which is not yet highly significant, but does demonstrate a big role of the variables in the Republican ideology. It would seem like the state of origin from the congress members is an important in the ideology the party members take, and combined with their votes statsitics, it can be seen how these variables do play a bigger role in the decisions and apporaches they take in their politics.
t.test(StateFilteredRep$dw1)##
## One Sample t-test
##
## data: StateFilteredRep$dw1
## t = 72.379, df = 1278, p-value < 2.2e-16
## alternative hypothesis: true mean is not equal to 0
## 95 percent confidence interval:
## 0.4297313 0.4536760
## sample estimates:
## mean of x
## 0.4417037t.test(StateFilteredRep$dw1)##
## One Sample t-test
##
## data: StateFilteredRep$dw1
## t = 72.379, df = 1278, p-value < 2.2e-16
## alternative hypothesis: true mean is not equal to 0
## 95 percent confidence interval:
## 0.4297313 0.4536760
## sample estimates:
## mean of x
## 0.4417037Conclusion
In conclusion, what was expected to be proved in the study was not done so in completion due to how the state of origin did not play such a big role as it was expected. For instance, the Democratic party did not prove to be influenced much by where its members are from as their R^2 value was very low regarding the percentage of the role the state had on the ideologies of the Democrats ideologies. Whilst in the case of the Republicans, a higher R^2 percentage value was obtained, possibly indicating a higher influence on the ideologies, the results did not prove to be very strong, thus proving the hypothesis to be proven wrong to some extent. Additionally, through the analysis of the polarization data for both parties, it can be seen how both parties in the selected states present different variances in the polarization values. At first, it can be seen how the Democrats do not present very extreme values in the scale from 0 to -1 as their polarization mainly varies around -0.4, demonstrating to be closer to a neutral stance than to an extreme of the ideology they follow. On the other hand, the Republican Party did demonstrate more variances as they had a great number o values in different numbr in the polarization scale on the other side, being from 0 to 1. Primarily the high values of their polarization are in between 0.1 and 0.7, demonstrating more extreme ideologies and stronger beliefs in what they follow. Thus, it can be concluded to how the support from the voters in different states does not have such a strong relation as it was believed to have in the polarization of the members of congress.