Text as Data. Justin Grimmer. Associate Professor Department of Political Science Stanford University. November 20th, 2014
|
|
- Neil Beasley
- 5 years ago
- Views:
Transcription
1 Text as Data Justin Grimmer Associate Professor Department of Political Science Stanford University November 20th, 2014 Justin Grimmer (Stanford University) Text as Data November 20th, / 24
2 Ideological Scaling 1) Task - Measure political actors position in policy space - Low dimensional representation of beliefs 2) Objective function - Linear Discriminant Analysis (ish) Wordscores (today) - Item Response Theory Wordfish - Item Response Theory + Roll Call Votes Issue-specific ideal points (12/2) 3) Optimization - Wordscores straightforward, based on training texts - Wordfish EM, MCMC methods 4) Validation - What is the goal of embedding? - What is the gold standard? Justin Grimmer (Stanford University) Text as Data November 20th, / 24
3 The Spatial Model - Suppose we have actor i (i = 1, 2, 3,..., N) Justin Grimmer (Stanford University) Text as Data November 20th, / 24
4 The Spatial Model - Suppose we have actor i (i = 1, 2, 3,..., N) - Actor has ideal point θ i R M Justin Grimmer (Stanford University) Text as Data November 20th, / 24
5 The Spatial Model - Suppose we have actor i (i = 1, 2, 3,..., N) - Actor has ideal point θ i R M - We describe actor i s utility from proposal p R M with utility function Justin Grimmer (Stanford University) Text as Data November 20th, / 24
6 The Spatial Model - Suppose we have actor i (i = 1, 2, 3,..., N) - Actor has ideal point θ i R M - We describe actor i s utility from proposal p R M with utility function u i (θ i, p) = d(θ i, p) Justin Grimmer (Stanford University) Text as Data November 20th, / 24
7 The Spatial Model - Suppose we have actor i (i = 1, 2, 3,..., N) - Actor has ideal point θ i R M - We describe actor i s utility from proposal p R M with utility function u i (θ i, p) = d(θ i, p) L = ( θ il p }{{} l ) 2 l=1 ideal policy Justin Grimmer (Stanford University) Text as Data November 20th, / 24
8 The Spatial Model - Suppose we have actor i (i = 1, 2, 3,..., N) - Actor has ideal point θ i R M - We describe actor i s utility from proposal p R M with utility function u i (θ i, p) = d(θ i, p) L = ( θ il p }{{} l ) 2 l=1 ideal policy Estimation goal: θ i Justin Grimmer (Stanford University) Text as Data November 20th, / 24
9 The Spatial Model - Suppose we have actor i (i = 1, 2, 3,..., N) - Actor has ideal point θ i R M - We describe actor i s utility from proposal p R M with utility function u i (θ i, p) = d(θ i, p) L = ( θ il p }{{} l ) 2 l=1 ideal policy Estimation goal: θ i Scaling placing actors in low-dimensional space (like principal components!) Justin Grimmer (Stanford University) Text as Data November 20th, / 24
10 Estimating Ideal Points: Roll Call Data and the US Congress US Congress and Roll Call Justin Grimmer (Stanford University) Text as Data November 20th, / 24
11 Estimating Ideal Points: Roll Call Data and the US Congress US Congress and Roll Call - Poole and Rosenthal voteview Justin Grimmer (Stanford University) Text as Data November 20th, / 24
12 Estimating Ideal Points: Roll Call Data and the US Congress US Congress and Roll Call - Poole and Rosenthal voteview - Roll Call Data 1789-Present Justin Grimmer (Stanford University) Text as Data November 20th, / 24
13 Estimating Ideal Points: Roll Call Data and the US Congress US Congress and Roll Call - Poole and Rosenthal voteview - Roll Call Data 1789-Present - NOMINATE methods place legislators on one dimension, estimate of ideology Justin Grimmer (Stanford University) Text as Data November 20th, / 24
14 Estimating Ideal Points: Roll Call Data and the US Congress US Congress and Roll Call - Poole and Rosenthal voteview - Roll Call Data 1789-Present - NOMINATE methods place legislators on one dimension, estimate of ideology - Wildly successful: Justin Grimmer (Stanford University) Text as Data November 20th, / 24
15 Estimating Ideal Points: Roll Call Data and the US Congress US Congress and Roll Call - Poole and Rosenthal voteview - Roll Call Data 1789-Present - NOMINATE methods place legislators on one dimension, estimate of ideology - Wildly successful: - Estimates are accurate: face validity Congressional scholars agree upon Justin Grimmer (Stanford University) Text as Data November 20th, / 24
16 Estimating Ideal Points: Roll Call Data and the US Congress US Congress and Roll Call - Poole and Rosenthal voteview - Roll Call Data 1789-Present - NOMINATE methods place legislators on one dimension, estimate of ideology - Wildly successful: - Estimates are accurate: face validity Congressional scholars agree upon - Insightful unidimensional Congress Justin Grimmer (Stanford University) Text as Data November 20th, / 24
17 Estimating Ideal Points: Roll Call Data and the US Congress US Congress and Roll Call - Poole and Rosenthal voteview - Roll Call Data 1789-Present - NOMINATE methods place legislators on one dimension, estimate of ideology - Wildly successful: - Estimates are accurate: face validity Congressional scholars agree upon - Insightful unidimensional Congress - Extensible: insight of IRT allows model to be embedded in many forms Justin Grimmer (Stanford University) Text as Data November 20th, / 24
18 Estimating Ideal Points: Roll Call Data and the US Congress US Congress and Roll Call - Poole and Rosenthal voteview - Roll Call Data 1789-Present - NOMINATE methods place legislators on one dimension, estimate of ideology - Wildly successful: - Estimates are accurate: face validity Congressional scholars agree upon - Insightful unidimensional Congress - Extensible: insight of IRT allows model to be embedded in many forms - Widely used: hard to write a paper on American political institutions with ideal points Justin Grimmer (Stanford University) Text as Data November 20th, / 24
19 Estimating Ideal Points in General Two Limitations with the NOMINATE project: Justin Grimmer (Stanford University) Text as Data November 20th, / 24
20 Estimating Ideal Points in General Two Limitations with the NOMINATE project: 1) US Congress is distinct Justin Grimmer (Stanford University) Text as Data November 20th, / 24
21 Estimating Ideal Points in General Two Limitations with the NOMINATE project: 1) US Congress is distinct roll call votes fail to measure ideology in other settings Justin Grimmer (Stanford University) Text as Data November 20th, / 24
22 Estimating Ideal Points in General Two Limitations with the NOMINATE project: 1) US Congress is distinct roll call votes fail to measure ideology in other settings - Weak party pressure individual discretion on votes Justin Grimmer (Stanford University) Text as Data November 20th, / 24
23 Estimating Ideal Points in General Two Limitations with the NOMINATE project: 1) US Congress is distinct roll call votes fail to measure ideology in other settings - Weak party pressure individual discretion on votes - Parliamentary systems no discretion, no variation. Justin Grimmer (Stanford University) Text as Data November 20th, / 24
24 Estimating Ideal Points in General Two Limitations with the NOMINATE project: 1) US Congress is distinct roll call votes fail to measure ideology in other settings - Weak party pressure individual discretion on votes - Parliamentary systems no discretion, no variation. - Spirling and Quinn (2011) mixture model like models for blocs in UK Parliament Justin Grimmer (Stanford University) Text as Data November 20th, / 24
25 Estimating Ideal Points in General Two Limitations with the NOMINATE project: 1) US Congress is distinct roll call votes fail to measure ideology in other settings - Weak party pressure individual discretion on votes - Parliamentary systems no discretion, no variation. - Spirling and Quinn (2011) mixture model like models for blocs in UK Parliament 2) Not everyone votes! Justin Grimmer (Stanford University) Text as Data November 20th, / 24
26 Estimating Ideal Points in General Two Limitations with the NOMINATE project: 1) US Congress is distinct roll call votes fail to measure ideology in other settings - Weak party pressure individual discretion on votes - Parliamentary systems no discretion, no variation. - Spirling and Quinn (2011) mixture model like models for blocs in UK Parliament 2) Not everyone votes! - Voters survey responses (but problems with that) Justin Grimmer (Stanford University) Text as Data November 20th, / 24
27 Estimating Ideal Points in General Two Limitations with the NOMINATE project: 1) US Congress is distinct roll call votes fail to measure ideology in other settings - Weak party pressure individual discretion on votes - Parliamentary systems no discretion, no variation. - Spirling and Quinn (2011) mixture model like models for blocs in UK Parliament 2) Not everyone votes! - Voters survey responses (but problems with that) - Challengers NPAT surveys (but they don t fill those out anymore) Justin Grimmer (Stanford University) Text as Data November 20th, / 24
28 Estimating Ideal Points in General Two Limitations with the NOMINATE project: 1) US Congress is distinct roll call votes fail to measure ideology in other settings - Weak party pressure individual discretion on votes - Parliamentary systems no discretion, no variation. - Spirling and Quinn (2011) mixture model like models for blocs in UK Parliament 2) Not everyone votes! - Voters survey responses (but problems with that) - Challengers NPAT surveys (but they don t fill those out anymore) - Bonica (2013, 2014) estimate ideology from donations (but not everyone donates) Justin Grimmer (Stanford University) Text as Data November 20th, / 24
29 Estimating Ideal Points in General But Everyone talks! Justin Grimmer (Stanford University) Text as Data November 20th, / 24
30 Estimating Ideal Points in General But Everyone talks! - If we could scale based on conversation, we can measure ideology anywhere Justin Grimmer (Stanford University) Text as Data November 20th, / 24
31 Estimating Ideal Points in General But Everyone talks! - If we could scale based on conversation, we can measure ideology anywhere - Much of the literature relies upon intuition from US Congress Justin Grimmer (Stanford University) Text as Data November 20th, / 24
32 Estimating Ideal Points in General But Everyone talks! - If we could scale based on conversation, we can measure ideology anywhere - Much of the literature relies upon intuition from US Congress - Hard not to find ideology Justin Grimmer (Stanford University) Text as Data November 20th, / 24
33 Estimating Ideal Points in General But Everyone talks! - If we could scale based on conversation, we can measure ideology anywhere - Much of the literature relies upon intuition from US Congress - Hard not to find ideology - Behavior that is primarily ideological Justin Grimmer (Stanford University) Text as Data November 20th, / 24
34 Estimating Ideal Points in General But Everyone talks! - If we could scale based on conversation, we can measure ideology anywhere - Much of the literature relies upon intuition from US Congress - Hard not to find ideology - Behavior that is primarily ideological - Reality: scaling is much more difficult than roll call voting examples Justin Grimmer (Stanford University) Text as Data November 20th, / 24
35 Estimating Ideal Points in General But Everyone talks! - If we could scale based on conversation, we can measure ideology anywhere - Much of the literature relies upon intuition from US Congress - Hard not to find ideology - Behavior that is primarily ideological - Reality: scaling is much more difficult than roll call voting examples - Hard to find ideology Justin Grimmer (Stanford University) Text as Data November 20th, / 24
36 Estimating Ideal Points in General But Everyone talks! - If we could scale based on conversation, we can measure ideology anywhere - Much of the literature relies upon intuition from US Congress - Hard not to find ideology - Behavior that is primarily ideological - Reality: scaling is much more difficult than roll call voting examples - Hard to find ideology - Much of political speech reveals little about position on ideological spectrum advertising, regional Justin Grimmer (Stanford University) Text as Data November 20th, / 24
37 Estimating Ideal Points in General But Everyone talks! - If we could scale based on conversation, we can measure ideology anywhere - Much of the literature relies upon intuition from US Congress - Hard not to find ideology - Behavior that is primarily ideological - Reality: scaling is much more difficult than roll call voting examples - Hard to find ideology - Much of political speech reveals little about position on ideological spectrum advertising, regional Healthy skepticism! Justin Grimmer (Stanford University) Text as Data November 20th, / 24
38 Our plan Justin Grimmer (Stanford University) Text as Data November 20th, / 24
39 Our plan 1) Wordscores supervised scaling Justin Grimmer (Stanford University) Text as Data November 20th, / 24
40 Our plan 1) Wordscores supervised scaling 2) Wordfish single dimension Justin Grimmer (Stanford University) Text as Data November 20th, / 24
41 Wordscores Big in Europe Justin Grimmer (Stanford University) Text as Data November 20th, / 24
42 Wordscores, Like the Hoff Big in Europe Justin Grimmer (Stanford University) Text as Data November 20th, / 24
43 Wordscores: Objective Function For each legislator i, suppose we observe D i documents. Justin Grimmer (Stanford University) Text as Data November 20th, / 24
44 Wordscores: Objective Function For each legislator i, suppose we observe D i documents. Define: Justin Grimmer (Stanford University) Text as Data November 20th, / 24
45 Wordscores: Objective Function For each legislator i, suppose we observe D i documents. Define: x i = D i l=1 x il Justin Grimmer (Stanford University) Text as Data November 20th, / 24
46 Wordscores: Objective Function For each legislator i, suppose we observe D i documents. Define: x i = = D i l=1 D i x il (x il1, x il2,..., x ilj ) l=1 Justin Grimmer (Stanford University) Text as Data November 20th, / 24
47 Wordscores: Objective Function For each legislator i, suppose we observe D i documents. Define: x i = = D i l=1 D i x il (x il1, x il2,..., x ilj ) l=1 x i aggregation across documents. Justin Grimmer (Stanford University) Text as Data November 20th, / 24
48 Wordscores: Objective Functions Choose two legislators as exemplars Justin Grimmer (Stanford University) Text as Data November 20th, / 24
49 Wordscores: Objective Functions Choose two legislators as exemplars - Legislator L {1, 2,..., N} is liberal. Y L = 1 Justin Grimmer (Stanford University) Text as Data November 20th, / 24
50 Wordscores: Objective Functions Choose two legislators as exemplars - Legislator L {1, 2,..., N} is liberal. Y L = 1 - For example, might select Elizabeth Warren Justin Grimmer (Stanford University) Text as Data November 20th, / 24
51 Wordscores: Objective Functions Choose two legislators as exemplars - Legislator L {1, 2,..., N} is liberal. Y L = 1 - For example, might select Elizabeth Warren - Legislator C {1, 2,..., N} is Conservative. Y C = 1 Justin Grimmer (Stanford University) Text as Data November 20th, / 24
52 Wordscores: Objective Functions Choose two legislators as exemplars - Legislator L {1, 2,..., N} is liberal. Y L = 1 - For example, might select Elizabeth Warren - Legislator C {1, 2,..., N} is Conservative. Y C = 1 - For example, might select Ted Cruz Justin Grimmer (Stanford University) Text as Data November 20th, / 24
53 Wordscores: Objective Functions Choose two legislators as exemplars - Legislator L {1, 2,..., N} is liberal. Y L = 1 - For example, might select Elizabeth Warren - Legislator C {1, 2,..., N} is Conservative. Y C = 1 - For example, might select Ted Cruz For each word j we can define: Justin Grimmer (Stanford University) Text as Data November 20th, / 24
54 Wordscores: Objective Functions Choose two legislators as exemplars - Legislator L {1, 2,..., N} is liberal. Y L = 1 - For example, might select Elizabeth Warren - Legislator C {1, 2,..., N} is Conservative. Y C = 1 - For example, might select Ted Cruz For each word j we can define: P jl = Probability of word from Liberal Justin Grimmer (Stanford University) Text as Data November 20th, / 24
55 Wordscores: Objective Functions Choose two legislators as exemplars - Legislator L {1, 2,..., N} is liberal. Y L = 1 - For example, might select Elizabeth Warren - Legislator C {1, 2,..., N} is Conservative. Y C = 1 - For example, might select Ted Cruz For each word j we can define: P jl = Probability of word from Liberal P jc = Probability of word from Conservative Justin Grimmer (Stanford University) Text as Data November 20th, / 24
56 Wordscores: Objective Functions Choose two legislators as exemplars - Legislator L {1, 2,..., N} is liberal. Y L = 1 - For example, might select Elizabeth Warren - Legislator C {1, 2,..., N} is Conservative. Y C = 1 - For example, might select Ted Cruz For each word j we can define: P jl = Probability of word from Liberal P jc = Probability of word from Conservative Define the score for word j Justin Grimmer (Stanford University) Text as Data November 20th, / 24
57 Wordscores: Objective Functions Choose two legislators as exemplars - Legislator L {1, 2,..., N} is liberal. Y L = 1 - For example, might select Elizabeth Warren - Legislator C {1, 2,..., N} is Conservative. Y C = 1 - For example, might select Ted Cruz For each word j we can define: P jl = Probability of word from Liberal P jc = Probability of word from Conservative Define the score for word j S j = Y C P jc + Y L P jl Justin Grimmer (Stanford University) Text as Data November 20th, / 24
58 Wordscores: Objective Functions Choose two legislators as exemplars - Legislator L {1, 2,..., N} is liberal. Y L = 1 - For example, might select Elizabeth Warren - Legislator C {1, 2,..., N} is Conservative. Y C = 1 - For example, might select Ted Cruz For each word j we can define: P jl = Probability of word from Liberal P jc = Probability of word from Conservative Define the score for word j S j = Y C P jc + Y L P jl = P jc P jl Justin Grimmer (Stanford University) Text as Data November 20th, / 24
59 Wordscores: Objective Functions Scale other legislators: Justin Grimmer (Stanford University) Text as Data November 20th, / 24
60 Wordscores: Objective Functions Scale other legislators: N i = J j=1 x i Justin Grimmer (Stanford University) Text as Data November 20th, / 24
61 Wordscores: Objective Functions Scale other legislators: N i = J j=1 x i ˆθ i is Justin Grimmer (Stanford University) Text as Data November 20th, / 24
62 Wordscores: Objective Functions Scale other legislators: N i = J j=1 x i ˆθ i is ˆθ i = J j=1 ( xij N i ) S j Justin Grimmer (Stanford University) Text as Data November 20th, / 24
63 Wordscores: Objective Functions Scale other legislators: N i = J j=1 x i ˆθ i is ˆθ i = J j=1 ( xij N i ) S j = x i N i S Justin Grimmer (Stanford University) Text as Data November 20th, / 24
64 Wordscores: Optimization N L = N C = J m=1 J m=1 x ml x mc Justin Grimmer (Stanford University) Text as Data November 20th, / 24
65 Wordscores: Optimization N L = N C = J m=1 J m=1 x ml x mc Estimate P jl, P jc, and S j Justin Grimmer (Stanford University) Text as Data November 20th, / 24
66 Wordscores: Optimization N L = N C = J m=1 J m=1 x ml x mc Estimate P jl, P jc, and S j P jl = x jl N L x jl N L + x jc N C Justin Grimmer (Stanford University) Text as Data November 20th, / 24
67 Wordscores: Optimization N L = N C = J m=1 J m=1 x ml x mc Estimate P jl, P jc, and S j P jl = x jl N L x jl N L + x jc N C P jc = 1 P jl = x jc N C x jl N L + x jc N C Justin Grimmer (Stanford University) Text as Data November 20th, / 24
68 Wordscores: Optimization N L = N C = J m=1 J m=1 x ml x mc Estimate P jl, P jc, and S j P jl = x jl N L x jl N L + x jc N C P jc = 1 P jl = S j = P jc P jl x jc N C x jl N L + x jc N C Justin Grimmer (Stanford University) Text as Data November 20th, / 24
69 Applied to the Senate Press Releases L = Ted Kennedy C = Tom Coburn Apply to other senators. Justin Grimmer (Stanford University) Text as Data November 20th, / 24
70 Applying to Senate Press Releases Gold Standard Scaling from NOMINATE Density Ideal Point Estimate Justin Grimmer (Stanford University) Text as Data November 20th, / 24
71 Applying to Senate Press Releases WordScores Density Score Justin Grimmer (Stanford University) Text as Data November 20th, / 24
72 Supervised Scaling Wordscores is one example of supervised embedding: Justin Grimmer (Stanford University) Text as Data November 20th, / 24
73 Supervised Scaling Wordscores is one example of supervised embedding: 1) Linear Discriminant Analysis (LDA) the Federalist Papers Justin Grimmer (Stanford University) Text as Data November 20th, / 24
74 Supervised Scaling Wordscores is one example of supervised embedding: 1) Linear Discriminant Analysis (LDA) the Federalist Papers 2) Multinomial Inverse Regression Justin Grimmer (Stanford University) Text as Data November 20th, / 24
75 Supervised Scaling Wordscores is one example of supervised embedding: 1) Linear Discriminant Analysis (LDA) the Federalist Papers 2) Multinomial Inverse Regression 3) Any method that separates words Justin Grimmer (Stanford University) Text as Data November 20th, / 24
76 Supervised Scaling Wordscores is one example of supervised embedding: 1) Linear Discriminant Analysis (LDA) the Federalist Papers 2) Multinomial Inverse Regression 3) Any method that separates words The findings will depend heavily on supervision Justin Grimmer (Stanford University) Text as Data November 20th, / 24
77 Supervised Scaling Wordscores is one example of supervised embedding: 1) Linear Discriminant Analysis (LDA) the Federalist Papers 2) Multinomial Inverse Regression 3) Any method that separates words The findings will depend heavily on supervision - Sensitive to who is chosen Justin Grimmer (Stanford University) Text as Data November 20th, / 24
78 Supervised Scaling Wordscores is one example of supervised embedding: 1) Linear Discriminant Analysis (LDA) the Federalist Papers 2) Multinomial Inverse Regression 3) Any method that separates words The findings will depend heavily on supervision - Sensitive to who is chosen - False prediction problem speech accomplishes many goals, only some of which are ideological Justin Grimmer (Stanford University) Text as Data November 20th, / 24
79 What Can Go Wrong with Wordscores? Lowe (2008): Discusses potentially problematic wordscores properties Justin Grimmer (Stanford University) Text as Data November 20th, / 24
80 What Can Go Wrong with Wordscores? Lowe (2008): Discusses potentially problematic wordscores properties 1) Each word is weighted equally (fixable with different scoring procedure) Justin Grimmer (Stanford University) Text as Data November 20th, / 24
81 What Can Go Wrong with Wordscores? Lowe (2008): Discusses potentially problematic wordscores properties 1) Each word is weighted equally (fixable with different scoring procedure) 2) Unique words are conflated with centrist (fixable with MCQ fightin words style algorithm) Justin Grimmer (Stanford University) Text as Data November 20th, / 24
82 What Can Go Wrong with Wordscores? Lowe (2008): Discusses potentially problematic wordscores properties 1) Each word is weighted equally (fixable with different scoring procedure) 2) Unique words are conflated with centrist (fixable with MCQ fightin words style algorithm) 3) General problem: hard to interpret lack of a model Justin Grimmer (Stanford University) Text as Data November 20th, / 24
83 What Can Go Wrong with Wordscores? Lowe (2008): Discusses potentially problematic wordscores properties 1) Each word is weighted equally (fixable with different scoring procedure) 2) Unique words are conflated with centrist (fixable with MCQ fightin words style algorithm) 3) General problem: hard to interpret lack of a model To be fair: fast, nonparametric, and novel [trailblazing] method for scoring documents (starts conversation) Justin Grimmer (Stanford University) Text as Data November 20th, / 24
84 Unsupervised Embedding Basic idea: Justin Grimmer (Stanford University) Text as Data November 20th, / 24
85 Unsupervised Embedding Basic idea: - Actors have underlying latent position Justin Grimmer (Stanford University) Text as Data November 20th, / 24
86 Unsupervised Embedding Basic idea: - Actors have underlying latent position - Actors articulate that latent position in their speech Justin Grimmer (Stanford University) Text as Data November 20th, / 24
87 Unsupervised Embedding Basic idea: - Actors have underlying latent position - Actors articulate that latent position in their speech - This is associated with word usage, so high discriminating words correspond to ideological speech Justin Grimmer (Stanford University) Text as Data November 20th, / 24
88 Unsupervised Embedding Basic idea: - Actors have underlying latent position - Actors articulate that latent position in their speech - This is associated with word usage, so high discriminating words correspond to ideological speech - Some words discriminate better than others encode that in our model Justin Grimmer (Stanford University) Text as Data November 20th, / 24
89 Unsupervised Embedding Basic idea: - Actors have underlying latent position - Actors articulate that latent position in their speech - This is associated with word usage, so high discriminating words correspond to ideological speech - Some words discriminate better than others encode that in our model Simplest model: Principal Components Justin Grimmer (Stanford University) Text as Data November 20th, / 24
90 Application of Principal Components in R Consider press releases from 2005 US Senators Justin Grimmer (Stanford University) Text as Data November 20th, / 24
91 Application of Principal Components in R Consider press releases from 2005 US Senators Define x i = (x i1, x i2,..., x ij ) as the rate senator i uses J words. Justin Grimmer (Stanford University) Text as Data November 20th, / 24
92 Application of Principal Components in R Consider press releases from 2005 US Senators Define x i = (x i1, x i2,..., x ij ) as the rate senator i uses J words. x ij = No. Times i uses word j No. words i uses Justin Grimmer (Stanford University) Text as Data November 20th, / 24
93 Application of Principal Components in R Consider press releases from 2005 US Senators Define x i = (x i1, x i2,..., x ij ) as the rate senator i uses J words. x ij = No. Times i uses word j No. words i uses dtm: matrix containing word rates for senators Justin Grimmer (Stanford University) Text as Data November 20th, / 24
94 Application of Principal Components in R Consider press releases from 2005 US Senators Define x i = (x i1, x i2,..., x ij ) as the rate senator i uses J words. x ij = No. Times i uses word j No. words i uses dtm: matrix containing word rates for senators prcomp(dtm) applies principal components Justin Grimmer (Stanford University) Text as Data November 20th, / 24
95 Application of Principal Components in R Projection Dewine2005 Alexander2005 Bunning2005 Shelby2005 Warner2005 Graham2005 Isakson2005 Chafee2005 Specter2005 Allen2005 Feingold2005 Kennedy2005 Hagel2005 Byrd2005 Bayh2005 Obama2005 Cantwell2005 Brownback2005 Stevens2005 Lautenberg2005 Kyl2005 Conrad2005 Feinstein2005 Grassley2005 Rockefeller2005 Thomas2005 Hutchison2005 Craig2005 Murkowski2005 Salazar2005 Landrieu2005 McConnell2005 Lott2005 Akaka2005 Coleman2005 Crapo2005 Jeffords2005 Martinez2005 Demint2005 Corzine2005 Boxer2005 Murray2005 Lieberman2005 Carper2005 Vitter2005 Talent2005 Baucus2005 Sessions2005 Enzi2005 Ensign2005 Roberts2005 Dorgan2005 Wyden2005 Allard2005 BenNelson2005 Pryor2005 Durbin2005 Inhofe2005 Voinovich2005 Domenici2005 Frist2005 Clinton2005 Dodd2005 Snowe2005 Dayton2005 Johnson2005 Burns2005 Smith2005 Harkin2005 Dole2005 Lugar2005 Thune2005 Sununu2005 Schumer2005 Sarbanes2005 Bingaman2005 Chambliss2005 Kohl2005 Burr2005 Lincoln2005 Mikulski2005 Santorum2005 Collins2005 Bennett2005 Bond2005 Gregg2005 Stabenow2005 Inouye2005 Reed2005 Levin2005 Leahy2005 Reid2005 BillNelson2005 Kerry2005 Cornyn2005 Biden2005 McCain2005 Coburn2005 Hatch2005 Cochran Order Justin Grimmer (Stanford University) Text as Data November 20th, / 24
96 Application of Principal Components in R Shelby2005 Dewine20 Difference: Credit Claiming, Position Taking Graham2005 Bond2005 Bunning2005 Sarbanes2005 Specter2005 Bennett2005 Allen2005 Warner2005 Isakson2005 Chafee2005 Allard2005 Dole2005 Gregg2005 Harkin2005 Chambliss2005 Inouye2005 Talent2005 Burr2005 Mikulski2005 Thune2005 Collins2005 Levin2005 Johnson2005 Dodd2005 Bingaman2005 Lugar2005 Lincoln2005 Santorum2005 Stabenow2005 Schumer2005 Durbin2005 Burns2005 Kohl2005 Hutchison2005 Domenici2005 Conrad2005 Roberts2005 Dorgan2005 Clinton2005 Dayton2005 Crapo2005 Lott2005 Carper2005 Enzi2005 Wyden2005 Pryor2005 Snowe2005 Sununu2005 Murray2005 Boxer2005 BenNelson2005 Inhofe2005 Reed2005 Obama2005 Landrieu2005 Sessions2005 Vitter2005 Voinovich2005 Smith2005 Brownback2005 Coleman2005 Hatch2005 Ensign2005 chran2005 Bayh2005 Grassley2005 Thomas2005 Martinez2005 Lieberman2005 Hagel2005 Byrd2005 McConnell2005 Stevens2005 Feinstein2005 Rockefeller2005 Akaka2005 Baucus2005 Murkowski2005 Salazar2005 Craig2005 BillNelson2005 Cornyn2005 Corzine2005 Demint2005 Frist2005 Leahy2005 Feingold2005 Lautenberg2005 Jeffords2005 Coburn2005 Biden2005 Kyl2005 McCain2005 Kerry2005 Kennedy2005 Cantwell2005 Reid2005 Alexander Principal Component Justin Grimmer (Stanford University) Text as Data November 20th, / 24
97 Probabilistic Unsupervised Embeddings Principal components is powerful Justin Grimmer (Stanford University) Text as Data November 20th, / 24
98 Probabilistic Unsupervised Embeddings Principal components is powerful statistical model for unsupervised scaling Justin Grimmer (Stanford University) Text as Data November 20th, / 24
99 Probabilistic Unsupervised Embeddings Principal components is powerful statistical model for unsupervised scaling Item Response Theory (IRT) Justin Grimmer (Stanford University) Text as Data November 20th, / 24
100 Probabilistic Unsupervised Embeddings Principal components is powerful statistical model for unsupervised scaling Item Response Theory (IRT) - Origins: educational testing Justin Grimmer (Stanford University) Text as Data November 20th, / 24
101 Probabilistic Unsupervised Embeddings Principal components is powerful statistical model for unsupervised scaling Item Response Theory (IRT) - Origins: educational testing - Jackman (2002), Clinton, Jackman, and Rivers (2004) apply to roll call voting Justin Grimmer (Stanford University) Text as Data November 20th, / 24
102 Probabilistic Unsupervised Embeddings Principal components is powerful statistical model for unsupervised scaling Item Response Theory (IRT) - Origins: educational testing - Jackman (2002), Clinton, Jackman, and Rivers (2004) apply to roll call voting - Power of IRT: Justin Grimmer (Stanford University) Text as Data November 20th, / 24
103 Probabilistic Unsupervised Embeddings Principal components is powerful statistical model for unsupervised scaling Item Response Theory (IRT) - Origins: educational testing - Jackman (2002), Clinton, Jackman, and Rivers (2004) apply to roll call voting - Power of IRT: a) Estimate ideal points with few observations Justin Grimmer (Stanford University) Text as Data November 20th, / 24
104 Probabilistic Unsupervised Embeddings Principal components is powerful statistical model for unsupervised scaling Item Response Theory (IRT) - Origins: educational testing - Jackman (2002), Clinton, Jackman, and Rivers (2004) apply to roll call voting - Power of IRT: a) Estimate ideal points with few observations b) Makes clear how to extend models Justin Grimmer (Stanford University) Text as Data November 20th, / 24
105 Probabilistic Unsupervised Embeddings Principal components is powerful statistical model for unsupervised scaling Item Response Theory (IRT) - Origins: educational testing - Jackman (2002), Clinton, Jackman, and Rivers (2004) apply to roll call voting - Power of IRT: a) Estimate ideal points with few observations b) Makes clear how to extend models c) Tuesday IRT and topic models to scale with votes + text Justin Grimmer (Stanford University) Text as Data November 20th, / 24
106 Probabilistic Unsupervised Embeddings Principal components is powerful statistical model for unsupervised scaling Item Response Theory (IRT) - Origins: educational testing - Jackman (2002), Clinton, Jackman, and Rivers (2004) apply to roll call voting - Power of IRT: a) Estimate ideal points with few observations b) Makes clear how to extend models c) Tuesday IRT and topic models to scale with votes + text - Clinton, Jackman, and Rivers (2004) intuition about IRT Justin Grimmer (Stanford University) Text as Data November 20th, / 24
107 Probabilistic Unsupervised Embeddings Principal components is powerful statistical model for unsupervised scaling Item Response Theory (IRT) - Origins: educational testing - Jackman (2002), Clinton, Jackman, and Rivers (2004) apply to roll call voting - Power of IRT: a) Estimate ideal points with few observations b) Makes clear how to extend models c) Tuesday IRT and topic models to scale with votes + text - Clinton, Jackman, and Rivers (2004) intuition about IRT - Rivers (2002) Identification conditions Justin Grimmer (Stanford University) Text as Data November 20th, / 24
108 Probabilistic Unsupervised Embeddings Principal components is powerful statistical model for unsupervised scaling Item Response Theory (IRT) - Origins: educational testing - Jackman (2002), Clinton, Jackman, and Rivers (2004) apply to roll call voting - Power of IRT: a) Estimate ideal points with few observations b) Makes clear how to extend models c) Tuesday IRT and topic models to scale with votes + text - Clinton, Jackman, and Rivers (2004) intuition about IRT - Rivers (2002) Identification conditions - Bonica (2014a, 2014b) uses IRT (like the one we re about to use) to scale donors Justin Grimmer (Stanford University) Text as Data November 20th, / 24
109 Probabilistic Unsupervised Embeddings Principal components is powerful statistical model for unsupervised scaling Item Response Theory (IRT) - Origins: educational testing - Jackman (2002), Clinton, Jackman, and Rivers (2004) apply to roll call voting - Power of IRT: a) Estimate ideal points with few observations b) Makes clear how to extend models c) Tuesday IRT and topic models to scale with votes + text - Clinton, Jackman, and Rivers (2004) intuition about IRT - Rivers (2002) Identification conditions - Bonica (2014a, 2014b) uses IRT (like the one we re about to use) to scale donors Monroe and Maeda (2005) and Slopkin and Proksch (2008) develop similar algorithms Justin Grimmer (Stanford University) Text as Data November 20th, / 24
110 WordFish Objective Function Suppose we have legislator i. Justin Grimmer (Stanford University) Text as Data November 20th, / 24
111 WordFish Objective Function Suppose we have legislator i. x ij Poisson(λ ij ) Justin Grimmer (Stanford University) Text as Data November 20th, / 24
112 WordFish Objective Function Suppose we have legislator i. x ij Poisson(λ ij ) λ ij = exp(α i + ψ j + β j θ i ) Justin Grimmer (Stanford University) Text as Data November 20th, / 24
113 WordFish Objective Function Suppose we have legislator i. Where, x ij Poisson(λ ij ) λ ij = exp(α i + ψ j + β j θ i ) Justin Grimmer (Stanford University) Text as Data November 20th, / 24
114 WordFish Objective Function Suppose we have legislator i. x ij Poisson(λ ij ) λ ij = exp(α i + ψ j + β j θ i ) Where, λ ij = Rate individual i uses word j Justin Grimmer (Stanford University) Text as Data November 20th, / 24
115 WordFish Objective Function Suppose we have legislator i. x ij Poisson(λ ij ) λ ij = exp(α i + ψ j + β j θ i ) Where, λ ij = Rate individual i uses word j α i = Individual i loquaciousness Justin Grimmer (Stanford University) Text as Data November 20th, / 24
116 WordFish Objective Function Suppose we have legislator i. x ij Poisson(λ ij ) λ ij = exp(α i + ψ j + β j θ i ) Where, λ ij = Rate individual i uses word j α i = Individual i loquaciousness ψ j = Word j s frequency Justin Grimmer (Stanford University) Text as Data November 20th, / 24
117 WordFish Objective Function Suppose we have legislator i. x ij Poisson(λ ij ) λ ij = exp(α i + ψ j + β j θ i ) Where, λ ij = Rate individual i uses word j α i = Individual i loquaciousness ψ j = Word j s frequency β j = Word j s discrimination Justin Grimmer (Stanford University) Text as Data November 20th, / 24
118 WordFish Objective Function Suppose we have legislator i. x ij Poisson(λ ij ) λ ij = exp(α i + ψ j + β j θ i ) Where, λ ij = Rate individual i uses word j α i = Individual i loquaciousness ψ j = Word j s frequency β j = Word j s discrimination θ i = Legislator i s latent positions Justin Grimmer (Stanford University) Text as Data November 20th, / 24
119 WordFish Objective Function Suppose we have legislator i. x ij Poisson(λ ij ) λ ij = exp(α i + ψ j + β j θ i ) Where, λ ij = Rate individual i uses word j α i = Individual i loquaciousness ψ j = Word j s frequency β j = Word j s discrimination θ i = Legislator i s latent positions Regression of x ij on ideal points θ i, where we have to learn θ i Justin Grimmer (Stanford University) Text as Data November 20th, / 24
120 WordFish Objective Function/Optimization Implies the following posterior distribution: Justin Grimmer (Stanford University) Text as Data November 20th, / 24
121 WordFish Objective Function/Optimization Implies the following posterior distribution: p(θ, α, ψ, β) p(α)p(β)p(ψ)p(θ) N J exp [ (α i + ψ j + β j θ i )] (α i + ψ j + β j θ i ) x ij i=1 j=1 x ij! Justin Grimmer (Stanford University) Text as Data November 20th, / 24
122 WordFish Objective Function/Optimization Implies the following posterior distribution: p(θ, α, ψ, β) p(α)p(β)p(ψ)p(θ) N J exp [ (α i + ψ j + β j θ i )] (α i + ψ j + β j θ i ) x ij Estimate parameters: i=1 j=1 x ij! Justin Grimmer (Stanford University) Text as Data November 20th, / 24
123 WordFish Objective Function/Optimization Implies the following posterior distribution: p(θ, α, ψ, β) p(α)p(β)p(ψ)p(θ) N J exp [ (α i + ψ j + β j θ i )] (α i + ψ j + β j θ i ) x ij Estimate parameters: - EM-algorithm i=1 j=1 x ij! Justin Grimmer (Stanford University) Text as Data November 20th, / 24
124 WordFish Objective Function/Optimization Implies the following posterior distribution: p(θ, α, ψ, β) p(α)p(β)p(ψ)p(θ) N J exp [ (α i + ψ j + β j θ i )] (α i + ψ j + β j θ i ) x ij Estimate parameters: - EM-algorithm - MCMC algorithm i=1 j=1 x ij! Justin Grimmer (Stanford University) Text as Data November 20th, / 24
125 WordFish Objective Function/Optimization Implies the following posterior distribution: p(θ, α, ψ, β) p(α)p(β)p(ψ)p(θ) N J exp [ (α i + ψ j + β j θ i )] (α i + ψ j + β j θ i ) x ij i=1 j=1 Estimate parameters: - EM-algorithm - MCMC algorithm - Variational Approximation x ij! Justin Grimmer (Stanford University) Text as Data November 20th, / 24
126 WordFish Objective Function/Optimization Implies the following posterior distribution: p(θ, α, ψ, β) p(α)p(β)p(ψ)p(θ) N J exp [ (α i + ψ j + β j θ i )] (α i + ψ j + β j θ i ) x ij i=1 j=1 Estimate parameters: - EM-algorithm - MCMC algorithm - Variational Approximation Wordfish package in R x ij! Justin Grimmer (Stanford University) Text as Data November 20th, / 24
127 Applications: German Party Manifestos Wordfish and German Platforms Policy Position PDS Greens SPD CDU/CSU FDP Year Justin Grimmer (Stanford University) Text as Data November 20th, / 24
128 Applications: German Party Manifestos Wordfish Scaling Scaling, Grimmer (2012) Dem GOP Wordfish and Senate Press Releases Wordfish Scaling Density Justin Grimmer (Stanford University) Text as Data November 20th, / 24
129 The Problem with Text-Based Scaling What does validation mean? 1) Replicate NOMINATE, DIME, or other gold standards? 2) Agreement with experts 3) Prediction of other behavior Must answer this to make progress on pure text scaling Justin Grimmer (Stanford University) Text as Data November 20th, / 24
Senators of the 109th Congress
Home > Senators Home Senators of the 109th Congress Sort by: Name State Party a class? What is The Senators page on Statistics & Lists is a great resource for information about current and former Senators.
More informationSenators of the 110th Congress
Find Your Senators Search Home > Senators Home Senators of the 110th Congress Sort by: Name State Party Choose a State Choose a Senator Choose a Class Photos and contact information for the new senators
More informationFAIR s Congressional Voting Report is designed to help you understand a U.S. senator s support for. immigration control during the second session of
L E G E N D FAIR s Congressional Voting Report is designed to help you understand a U.S. senator s support for immigration measures to strengthen or weaken + Introduced Bill FAIR Supports immigration control
More informationCongressional Scorecard. 111th Congress First Session How to Judge a Member s Voting Record
111th Congress First 2009 How to Judge a Member s Record selects a few roll-call votes from the hundreds cast by members of Congress every session. In choosing these votes, attempts to fairly represent
More informationSenators of the 111th Congress
Senators of the 111th Congress Begich, Mark - (D - AK) 144 RUSSELL SENATE OFFICE BUILDING WASHINGTON DC 20510 (202) 224-3004 Web Form: begich.senate.gov/contact/contact.cfm Murkowski, Lisa - (R - AK) 709
More informationWashington, DC Washington, DC 20510
December 21, 2010 The Honorable Harry Reid The Honorable Mitch McConnell Majority Leader Minority Leader United States Senate United States Senate Washington, DC 20510 Washington, DC 20510 The Honorable
More informationOctober 3, United States Senate Washington, DC Dear Senator:
October 3, 2010 United States Senate Washington, DC 20510 Dear Senator: Congratulations and thank you on behalf of the National Association of Manufacturers (NAM), the nationʼs largest industrial trade
More informationSort by: Name State Party. What is a class?
1 of 7 Home > Senators Home Senators of the 111th Congress Sort by: Name State Party What is a class? XML Information about any senator, representative, vice president, or member of the Continental Congress.
More informationNRCAT Action Fund Senate Scorecard
The following scorecard is based on records of Senators actions on major pieces of torture related legislation in the 109th, 110th, 111th and 112th Congresses (2005 2012). = Acted against torture = Failed
More informationCongressional Scorecard. 112th Congress First Session How to Judge a Member s Voting Record
112th Congress First 2011 How to Judge a Member s Record selects a few roll-call votes from the hundreds cast by members of Congress every session. In choosing these votes, attempts to fairly represent
More informationPOLITICAL LAW AND GOVERNMENT ETHICS NEWS
Congress Passes Lobbying and Ethics Legislation...1 Lessons for Lobbyists: Travel...1 Federal Election Commission (FEC) Updates...3 Pointers for PACs (and Individuals): Contributing to Senators...5 Special
More informationTITLE VII PROTECTING TENANTS AT FORECLOSURE ACT
etreine on CSHP91QD1 with PUBLIC LAWS 123 STAT. 1660 PUBLIC LAW 111 22 MAY 20, 2009 Protecting Tenants at Foreclosure Act of 2009. 12 USC 5201 note. 12 USC 5220 note. shall have access, upon request, to
More informationWorking Paper Number 149 July 2008
Working Paper Number 149 July 2008 Why Warner-Lieberman Failed and How to Get America s Working Families behind the Next Cap-and-Trade Bill By David Wheeler Abstract Among partisans of greenhouse gas emissions
More informationRed River Army Depot 100 Main Drive Texarkana, TX
DCN 6846 Red River Army Depot 100 Main Drive Texarkana, TX 30 July, 2005 My name is Cecil Parker and I am a Civilian Employee at Red River Army Depot, Texarkana Texas,. I am a Mason of Lodge #69 and I
More informationS IN THE SENATE OF THE UNITED STATES
II 110TH CONGRESS 1ST SESSION S. 1041 To amend the National Labor Relations Act to establish an efficient system to enable employees to form, join, or assist labor organizations, to provide for mandatory
More informationFOLLOW UP WITH PHONE CALLS!!!!!!!!!
http://www.hr2003.org E-mail: passhr2003@hr2003.org Tel: 323-988-5688 Fax: 323-924-5563 For Immediate Release October 18, 2003 H.R. 2003 Ethiopia Democracy and Accountability Act of 2007 U.S. SENATE ADVOCACY
More informationMeasuring Political Positions from Legislative Speech
Advance Access publication July 20, 2016 Political Analysis (2016) 24:374 394 doi:10.1093/pan/mpw017 Measuring Political Positions from Legislative Speech Benjamin E. Lauderdale Department of Methodology,
More informationRed River Army Depot 100 Main Drive Tm~,ission Texarkana, TX
DCN 6743 Red River Army Depot 100 Main Drive Tm~,ission Texarkana, TX 30 July, 2005 r t.b'env~~, My name is Cecil Parker and I am a Civilian Employee at Red kver Army Depot, Texarkana Texas. I am a Mason
More informationCongressional Scorecard
Congressional Scorecard 114th Congress First 2015 How to Judge a Member s Voting Record AFSCME selects a few roll-call votes from the hundreds cast by members of Congress every session. In choosing these
More informationLEADERSHIP CHANGES IN THE 113 TH CONGRESS
David Leiter, President djleiter@mlstrategies.com ML Strategies, LLC 701 Pennsylvania Avenue, N.W. Washington, DC 20004 USA 202 296 3622 202 434 7400 fax www.mlstrategies.com OCTOBER 18, 2012 LEADERSHIP
More informationElection. A Guide to Changes in Congress. November 2006
Election 2006 A Guide to Changes in Congress November 2006 61103_11 1735 New York Avenue NW, Suite 500 Washington, DC 20006 www.pgerm.com Copyright 2006 Preston Gates Ellis & Rouvelas Meeds LLP Table Of
More informationFIRST SESSION, January to December 2013
SEPTEMBER 2014 FAIR s Congressional Voting Report is designed to help you understand the positions that each U.S. Senator has taken on immigration measures during the 113th Congress in furtherance of a
More informationA Strategy to Eliminate Wasteful Federal Spending
No. 806 Delivered October 16, 2003 October 28, 2003 A Strategy to Eliminate Wasteful Federal Spending The Honorable Sam Brownback Federal waste is a grave disservice to hardworking taxpayers across our
More informationKEY CONGRESSIONAL DISTRICTS FOR HIV/AIDS ADVOCACY
KEY CONGRESSIONAL DISTRICTS FOR HIV/AIDS ADVOCACY The following is a state-by-state list of Legislators who hold seats on Congressional committees which are capable of making decisions that impact the
More informationUnless otherwise noted below, photographs are Official U.S. Government public domain photographs taken from Wikimedia.org under the Creative Commons License 2.0 or 3.0 August 15, 2015 CHOCOLECT Politically
More informationNRCAT Action Fund Senate Scorecard
NRCAT Action Fund Senate card The following scorecard is based on records of Senators actions on major pieces of torture related legislation in votes between 2011 and 2018. = Acted against torture = Failed
More informationPolarization: Implications for Policymaking & Accountability
Polarization: Implications for Policymaking & Accountability Carlos Algara calgara@ucdavis.edu July 26, 2017 Agenda 1 Wrapping up the Pivotal Politics Model 2 Procedural Cartel (Partisan) Model 3 Polarization
More informationS To amend titles 10 and 18, United States Code, to protect unborn victims of violence. IN THE SENATE OF THE UNITED STATES
II TH CONGRESS 1ST SESSION S. 1 Calendar No. To amend titles and 1, United States Code, to protect unborn victims of violence. IN THE SENATE OF THE UNITED STATES MAY, 00 Mr. DEWINE (for himself, Mr. GRAHAM
More information2016 Club for Growth Senate Rating
Motion to stop debate on a bill to require an audit of the Federal eserve Override of veto of bill to disapprove of the EPA's "Waters of the U.S." regulation Amendment to require agencies to repeal one
More information2008 U.S. SENATE CANDIDATE SCORECARD
2008 U.S. SENATE CANDIDATE SCORECARD While education policy falls largely into the hands of state leaders to decide, the role of a U.S. Senator can be instrumental in advancing - or stopping reform in
More informationMeasuring the Political Sophistication of Voters in the Netherlands and the United States
Measuring the Political Sophistication of Voters in the Netherlands and the United States Christopher N. Lawrence Department of Political Science Saint Louis University November 2006 Overview What is political
More informationIs John McCain more conservative than Rand Paul? Using activists pairwise comparisons to measure ideology
Is John McCain more conservative than Rand Paul? Using activists pairwise comparisons to measure ideology Daniel J. Hopkins Associate Professor University of Pennsylvania danhop@sas.upenn.edu Hans Noel
More informationMeasuring the Political Sophistication of Voters in the Netherlands and the United States
Measuring the Political Sophistication of Voters in the Netherlands and the United States Christopher N. Lawrence Department of Political Science Saint Louis University November 2006 Overview What is political
More informationNovember Election 2008 A Guide to Changes in Congress. K&L Gates LLP 1601 K Street Washington, DC
November 2008 Election 2008 A Guide to Changes in Congress K&L Gates LLP 1601 K Street Washington, DC 2006 202.778.9000 Table Of Contents Introduction...2 Members by State...9 Senate Committees...35 House
More informationAssessing differences in legislators revealed preferences: A case study on the 107 th U.S. Senate
Assessing differences in legislators revealed preferences: A case study on the 107 th U.S. Senate arxiv:1601.07617v1 [stat.ap] 28 Jan 2016 Chelsea L. Lofland, Abel Rodríguez and Scott Moser Abstract Roll
More informationCRS Report for Congress Received through the CRS Web
Order Code RL30704 CRS Report for Congress Received through the CRS Web Major Leadership Election Contests In the Senate: A 27-Year Survey Updated November 14, 2001 Mildred Amer Specialist in American
More informationInfluence in Social Networks
CSCI 3210: Computational Game Theory Influence Games Ref: Irfan & Ortiz, AI (2014) Reading: Sections 1 3(up to pg. 86), Sections 4.5, 5 (no proof), 6 bowdoin.edu/~mirfan/papers/irfan_ortiz_influence_games_ai2014.pdf
More informationTHE AMERICAN POWER. All- Stars. Scorecard & Voting Guide
THE AMERICAN POWER All- Stars Scorecard & Voting Guide History Introduction About every two years, when Congress takes up an energy bill, the Big Oil Team and the Clean Energy Team go head to head on the
More informationCampaign 16. A Hawthorn Group visit with Kansas City Chamber June 24, 2016
Campaign 16 A Hawthorn Group visit with Kansas City Chamber June 24, 2016 Agenda I. The Knowns II. The Unknowns III. The Early Appeals IV. The Hard Questions Bring Down Title The Knowns It s a Big Political
More informationRecipients of Tobacco PAC Contributions Among Congressional Leadership PACs January 1, 1999 August 7, 2007 (partial cycle)
APPENDIX D Recipients of Tobacco PAC Contributions Among Congressional Leadership PACs January 1, 1999 August 7, 2007 (partial cycle) RECIPENT MEMBER AFFILIATION (IF APPLICABLE) AMOUNT THE FREEDOM PROJECT
More informationIAALS
FEDERAL JUDICIAL SCREENING COMMITTEES This table lists the judicial screening committees that appear to be in operation as of August, or that stand ready to operate in the event of a vacancy. These committees
More informationElection 2014: Its Impact on Federal Policy-Making in 2015
Election 2014: Its Impact on Federal Policy-Making in 2015 Slade Gorton, Bart Gordon, Jim Walsh, Manny Rouvelas and Rick Valentine Copyright 2014 by K&L Gates LLP. All rights reserved. Former Senator Slade
More informationCommittees Fall 2016
Committees 17.251 Fall 2016 Wilson s Famous Quote Congress in session is Congress on public exhibition, whilst Congress in its committee-rooms is Congress at work. Organization Mechanics Theoretical perspectives
More informationJune 20, The Honorable Richard G. Lugar United States Senate Washington, DC 20510
PHYSICIANS FOR SOCIAL RESPONSIBILITY OCEANA U.S. PUBLIC INTEREST RESEARCH GROUP CENTER FOR INTERNATIONAL ENVIRONMENTAL LAW NATIONAL ENVIRONMENTAL TRUST FRIENDS OF THE EARTH ENVIRONMENTAL DEFENSE SIERRA
More informationSenate Committee Musical Chairs. August 15, 2018
Senate Committee Musical Chairs August 15, 2018 Key Retiring Committee Seniority over Sitting Chair/Ranking Member Viewed as Seat Republicans Will Most Likely Retain Viewed as Potentially At Risk Republican
More informationPartisan Influence in Congress and Institutional Change
Partisan Influence in Congress and Institutional Change Scott de Marchi Duke University demarchi@duke.edu Michael Ensley Indiana University ensley@indiana.edu Michael Tofias UW-Milwaukee tofias@uwm.edu
More informationThe Staff and The Sword Ministry. LETTERS TO THE PRESIDENTS OF THE UNITED STATES OF AMERICA (1979 to 2008)
LETTERS TO THE PRESIDENTS OF THE UNITED STATES OF AMERICA (1979 to 2008) PART 6E DISTRIBUTION LIST FOR LETTERS TO PRESIDENT GEORGE W. BUSH (2006) INTRODUCTION In addition to just sending letters to President
More informationThe Issue-Adjusted Ideal Point Model
The Issue-Adjusted Ideal Point Model arxiv:1209.6004v1 [stat.ml] 26 Sep 2012 Sean Gerrish Princeton University 35 Olden Street Princeton, NJ 08540 sgerrish@cs.princeton.edu David M. Blei Princeton University
More informationEstimating Candidate Positions in a Polarized Congress
Estimating Candidate Positions in a Polarized Congress Chris Tausanovitch Department of Political Science UCLA Christopher Warshaw Department of Political Science Massachusetts Institute of Technology
More informationSenate committee overviews
Senate committee overviews Full committee rosters, oversight and subcommittee list February 21, 2017 Senate Agriculture, Nutrition and Forestry Committee Deals with farming, food programs and forest resources
More information2017 Federal Budget Budget
2017 Federal Budget 2017 Budget By law, President s Budget must be submitted by first Monday of February (Feb. 6.) Past three outgoing administrations submitted transition budgets to Congress to help with
More information27 Additional Votes For Higher Taxes As Of 3/2/04
27 Additional Votes For Higher Taxes As Of 3/2/04 2003 CQ Senate Vote 147 Kerry Did Not Voted But Announced Support Tax Reductions - On-Budget Surplus S 1054 Reid, D-Nev., motion to waive the Budget Act
More informationJANUARY FEBRUARY MARCH MAY JUNE APRIL JULY AUGUST SEPTEMBER OCTOBER NOVEMBER DECEMBER S M T W T F S S M T W T S M T W T F S S M T W T F S
JANUARY FEBRUARY MARCH 5 6 7 8 9 0 4 5 6 7 8 9 0 9 0 4 5 9 0 4 4 5 4 5 6 7 8 6 7 8 6 7 8 9 0 9 0 APRIL MAY JUNE S M T W T 4 F 5 S 6 S M T W T F S 4 5 6 7 8 7 8 9 0 4 5 6 7 8 9 0 9 0 4 5 4 5 6 7 8 9 0 4
More information2003 TWV Congressional Panel
2003 TWV Congressional Panel William A. Daoulas - AM General LTC Scott Gorske - National Guard Bureau COL Gary Profit - OCAR Doug Stewart - Robinson International Mike Chase - Chase Associates 107 th Congress
More informationThe 2016 Election and U.S. Foreign Policy
The 2016 Election and U.S. Foreign Policy Paul Sracic, Ph.D. Professor and Chair Department of Politics and International Relations Youngstown State University Paradox The election will matter for U.S.
More informationEstimating Candidates Political Orientation in a Polarized Congress
Estimating Candidates Political Orientation in a Polarized Congress Chris Tausanovitch Department of Political Science UCLA Christopher Warshaw Department of Political Science Massachusetts Institute of
More informationAdvisory. Client Election Summary.
Client Advisory 2004 Election Summary TABLE OF CONTENTS: 2004 Election Summary... 1 Election Analysis... 2 Senate... 3 House of Representatives... 5 Legislation & Going Forward... 6 The Cabinet... 6 The
More informationElection 2014: The Midterm Results, the ACA and You
Election 2014: The Midterm Results, the ACA and You James Slotnick, JD Sun Life Financial AVP, Broker Education Join the conversation on Twitter using #SLFElection2014 The Midterm Results The Outlook for
More informationOn April 25, the U.S. Senate adopted
NALC disappointed, but determined, as Senate passes S. 1789 Top: NALC President Fredric Rolando explains to a delegation of Pennsylvania letter carriers why S. 1789 is a flawed bill. Above: An advertisement
More informationLearning and Visualizing Political Issues from Voting Records Erik Goldman, Evan Cox, Mikhail Kerzhner. Abstract
Learning and Visualizing Political Issues from Voting Records Erik Goldman, Evan Cox, Mikhail Kerzhner Abstract For our project, we analyze data from US Congress voting records, a dataset that consists
More informationThe State of Congressional Web Sites
The State of Congressional Web Sites Since 1998, the Congressional Management Foundation has assessed the quality of congressional Web sites to determine how Members of Congress can use the Internet to
More informationHolding Up the Senate: Bob Dole and the Politics of Holds in the U.S. Senate
Holding Up the Senate: Bob Dole and the Politics of Holds in the U.S. Senate Nicholas O. Howard Ph.D Candidate University of North Carolina nohoward@email.unc.edu Jason M. Roberts Associate Professor University
More informationDiscovering Influential Members of Congress. Deepank Gupta Travis Skare Simone Wu. Project Overview
Discovering Influential Members of Congress Deepank Gupta Travis Skare Simone Wu Project Overview In the US Congress, a legislator can endorse a bill publicly before the vote to determine whether it is
More informationUsing Text to Scale Legislatures with Uninformative Voting
Using Text to Scale Legislatures with Uninformative Voting Nick Beauchamp NYU Department of Politics August 8, 2012 Abstract This paper shows how legislators written and spoken text can be used to ideologically
More informationLeadership in the 115 th Congress
Leadership in the 115 th Congress A combination of events will spell changes to the leadership of both parties in both Chambers in the 115 th Congress. In the Senate, the retirement of Democratic Leader
More informationCan Ideal Point Estimates be Used as Explanatory Variables?
Can Ideal Point Estimates be Used as Explanatory Variables? Andrew D. Martin Washington University admartin@wustl.edu Kevin M. Quinn Harvard University kevin quinn@harvard.edu October 8, 2005 1 Introduction
More informationParty Current # of Seats Incoming # of Seats Net Gain/Loss Republicans 45 52* +7* Democrats 55 46* + -7*
FEDERAL LEGISLATIVE REPORT: 2014 MIDTERM ELECTION ANALYSIS November 7, 2014 U.S. SENATE Even with election results for three Democratic-held Senate seats still undecided, Republicans secured the necessary
More informationIntroduction to AUCD Advocacy
Introduction to AUCD Advocacy Cornerstone Government Affairs Mark Mioduski Amy Souders March 3, 2007 Role of Appropriations Committee Approve or modify the President s s Budget request for annual discretionary
More informationCRS Report for Congress Received through the CRS Web
Order Code RL30199 CRS Report for Congress Received through the CRS Web Budget FY2000: A Chronology with Internet Access Updated December 1, 1999 Susan E. Watkins Senior Research Librarian Information
More informationThe Gingrich Senators and Their Effect on the U.S. Senate
The Gingrich Senators and Their Effect on the U.S. Senate Sean M. Theriault Government Department 1 University Station A1800 The University of Texas at Austin Austin, TX 78701 seant@mail.utexas.edu 512-232-7279
More informationHierarchical Item Response Models for Analyzing Public Opinion
Hierarchical Item Response Models for Analyzing Public Opinion Xiang Zhou Harvard University July 16, 2017 Xiang Zhou (Harvard University) Hierarchical IRT for Public Opinion July 16, 2017 Page 1 Features
More informationSeptember Board Meeting. September 16, 2015
September Board Meeting September 16, 2015 White, Working Class: non-college educated households still struggling; suffering from the unaffordability of middle class life. Older Young Voters: those 25-32
More informationCommittee on Rules & Administration Committee on Rules & Administration
BARRY M. KAMINS PRESIDENT Phone: (212) 382-6700 Fax: (212) 768-8116 bkamins@nycbar.org September 25, 2006 The Honorable Trent Lott The Honorable Chris Dodd Chairman Ranking Member Committee on Rules &
More informationMeasuring Bias and Uncertainty in Ideal Point Estimates via the Parametric Bootstrap
Political Analysis (2004) 12:105 127 DOI: 10.1093/pan/mph015 Measuring Bias and Uncertainty in Ideal Point Estimates via the Parametric Bootstrap Jeffrey B. Lewis Department of Political Science, University
More information24 th Annual Health Sciences Tax Conference
24 th Annual Health Sciences Tax Conference December 10, 2014 Disclaimer EY refers to the global organization, and may refer to one or more, of the member firms of Ernst & Young Global Limited, each of
More informationTAX REFORM: Prospects for Reform Benefiting Indian Tribes and Indian Country
TAX REFORM: Prospects for Reform Benefiting Indian Tribes and Indian Country Mary J. Streitz NCAI Mid Year Conference Lincoln, Nebraska June 19, 2012 OVERVIEW Backdrop for tax reform Expiring tax and related
More informations e c o n d s e s s i o n t h c o n g r e s s environmental Scorecard League of conservation voters
s e c o n d s e s s i o n 1 1 0 t h c o n g r e s s N a t i o n a l environmental 08 Scorecard League of conservation voters www.lcv.org LCV Board of directors * John H. Adams Founding Director, Natural
More informations e c o n d s e s s i o n t h c o n g r e s s environmental Scorecard League of conservation voters
s e c o n d s e s s i o n 1 1 0 t h c o n g r e s s N a t i o n a l environmental 08 Scorecard League of conservation voters www.lcv.org LCV Board of directors * John H. Adams Founding Director, Natural
More informationS To amend title 18, United States Code, to exempt certain rocket propellants from prohibitions under that title on explosive materials.
II 10TH CONGRESS 1ST SESSION S. Calendar No. 1 To amend title 1, United States Code, to exempt certain rocket propellants from prohibitions under that title on explosive materials. IN THE SENATE OF THE
More informationTrump Wins. the White House. Republicans Hold the Senate and House. What do the Election Results Mean for Housing? DECISION
DECISION 2016 Trump Wins the White House Republicans Hold the Senate and House What do the Election Results Mean for Housing? One of the most divisive election campaigns in our nation s history is finally
More informationIndividual Contributions Arranged By Type, Giver, Then Recipient
Presented by the Federal Election Commission Individual Contributions Arranged By Type, Giver, Then Recipient Non-Federal Receipts "Exempt From Limits" NRSC - NONFEDERAL 12/21/1999 7500.00 20020022311
More informationFederal Policy and Legislative Update. DDAA Board of Directors Meeting January 17, 2017
Federal Policy and Legislative Update DDAA Board of Directors Meeting January 17, 2017 Presentation Overview Trump Administration s Agenda and Congress Role 115 th Congress: What s Changed Trump Cabinet
More informationPolitical Parties and Congressional Leadership /252 Fall 2012
Political Parties and Congressional Leadership 17.251/252 Fall 2012 Lecture Organization Natural history of political parties in the U.S. Why do parties exist? What explains varying levels of party cohesion
More informationCommittees Fall 2016
Committees 17.251 Fall 2016 1 Wilson s Famous Quote Congress in session is Congress on public exhibition, whilst Congress in its committee-rooms is Congress at work. This image is in the public domain.
More informationThe Congressional Debate on Partial-Birth Abortion: Constitutional Gravitas and Moral Passion
B.J.Pol.S. 38, 383 410 Copyright 2008 Cambridge University Press doi:10.1017/s0007123408000203 Printed in the United Kingdom The Congressional Debate on Partial-Birth Abortion: Constitutional Gravitas
More informationThe Congressional Debate on Partial-Birth Abortion: Constitutional Gravitas and Moral Passion
B.J.Pol.S. 38, 383 410 Copyright 2008 Cambridge University Press doi:10.1017/s0007123408000203 Printed in the United Kingdom The Congressional Debate on Partial-Birth Abortion: Constitutional Gravitas
More informationA POST-ELECTION VIEW FROM WASHINGTON: IMPACT OF THE 2016 PRESIDENTIAL AND CONGRESSIONAL CONTESTS
2015 Morgan, 2016 Morgan, Lewis Lewis & Bockius & Bockius LLP LLP A POST-ELECTION VIEW FROM WASHINGTON: IMPACT OF THE 2016 PRESIDENTIAL AND CONGRESSIONAL CONTESTS Matthew Miner, Partner, Washington D.C.
More informationCongressional Gridlock: The Effects of the Master Lever
Congressional Gridlock: The Effects of the Master Lever Olga Gorelkina Max Planck Institute, Bonn Ioanna Grypari Max Planck Institute, Bonn Preliminary & Incomplete February 11, 2015 Abstract This paper
More informationAdam Bonica. RSF: The Russell Sage Foundation Journal of the Social Sciences, Volume 2, Number 7, November 2016, pp.
A Data-Driven Voter Guide for U.S. Elections: Adapting Quantitative Measures of the Preferences and Priorities of Political Elites to Help Voters Learn About Candidates Adam Bonica RSF: The Russell Sage
More informationMarx s Conflict Theory
Marx & Modern Times Marx s Conflict Theory All societies are divided into two groups Owners Workers Our society is capitalist Owners are bourgeoisie Workers are proletarians Owners and Workers Owners exploit
More informationDo Individual Heterogeneity and Spatial Correlation Matter?
Do Individual Heterogeneity and Spatial Correlation Matter? An Innovative Approach to the Characterisation of the European Political Space. Giovanna Iannantuoni, Elena Manzoni and Francesca Rossi EXTENDED
More informationWNBC/Marist Poll Poughkeepsie, NY Phone Fax
WNBC/Marist Poll Poughkeepsie, NY 12601 Phone 845.575.5050 Fax 845.575.5111 www.maristpoll.marist.edu EMBARGOED FOR RELEASE: Tuesday 6:00 p.m. October 3, 2006 All references must be sourced WNBC/Marist
More informationNational Guard Legislative. FY09 Military Construction/ Veterans Affairs Appropriations Act. Current Bill Status
National Guard Legislative FY09 Military Construction/ Veterans Affairs Appropriations Act Summary and Analysis of H.R. 6599 and S. 3301 As of August 1, 2008 Current Bill Status Senate Appropriations Committee
More informationPartition Decomposition for Roll Call Data
Partition Decomposition for Roll Call Data G. Leibon 1,2, S. Pauls 2, D. N. Rockmore 2,3,4, and R. Savell 5 Abstract In this paper we bring to bear some new tools from statistical learning on the analysis
More informationCAPITOL HILL DAY 2014
CAPITOL HILL DAY 2014 Report and Summary National Alliance to End Homelessness August 2014 Prepared by: Kate Seif, Policy Outreach Coordinator Photo Credit: Joan Serviss and Margaret Kilman Executive Summary
More informationSeptember 26, 2013 Robert Moller NOAA Office of Legislative and Intergovernmental Affairs
September 26, 2013 Robert Moller NOAA Office of Legislative and Intergovernmental Affairs Our Work With Congress NOAA works with Congress through the Office of Legislative and Intergovernmental Affairs
More informationInferring Roll Call Scores from Campaign Contributions Using Supervised Machine Learning
Inferring Roll Call Scores from Campaign Contributions Using Supervised Machine Learning Adam Bonica March 24, 2016 Abstract. This paper develops a generalized supervised learning methodology for inferring
More informationnational congresses and show the results from a number of alternate model specifications for
Appendix In this Appendix, we explain how we processed and analyzed the speeches at parties national congresses and show the results from a number of alternate model specifications for the analysis presented
More informationFlorida v. HHS - Amicus Brief of Republican U.S. Senators
Santa Clara Law Santa Clara Law Digital Commons Patient Protection and Affordable Care Act Litigation Research Projects and Empirical Data 1-1-2011 Florida v. HHS - Amicus Brief of Republican U.S. Senators
More informationFacts. A total of 246 women (158 Democrats and 88 Republicans) have served in the U.S. Congress.
Florence Kahn of California (facing camera) and Edith Nourse Rogers of Massachusetts in early 1927, in Congress s first cloakroom for women Members. A House Page (far left) delivers a book to Kahn. Facts
More information