Climate Attitudes and State Climate Policies in the United States

Author

Jillian Reynolds

Modified

June 9, 2025

1 Theory

Research Question

Do states’ numbers of climate policies reflect the public opinion in their states?

Hypothesis

There is a positive correlation between pro-climate attitudes and the number of climate policies within a state.

2 Methods

2.1 Analysis

CI = 0.95

alpha = 0.05

Means for individuals’ attitudes: survey-weighted

Correlation: Pearson’s r

Regression: generalized linear models and linear models

3 Results

3.1 Sample for Analysis

The final sample for analysis is 6,318 respondents, 84.82 % of the original post-election sample. Respondents completed both the pre-election and post-election interview. Because DC is not a state and was not included in the State Climate Policy Dashboard, the sample for analysis excluded respondents from DC.

3.1.1 By attitude variable

Variable n
WARMUS 6,311  
WARMIMP 6,312  
WARMDO 6,272  
REGGREEN 6,308  
FSENV 6,310  

3.1.2 By state

Ten states had fewer than 30 respondents, and 15 had fewer than 50 respondents. Since some states had fewer respondents, all conclusions drawn on the bases of these smaller samples must be ___. The means, correlations, and regression using state attitude means may have computed differently were those state samples larger than they were.

State no_SD_vars removed final_n
CA 644   66   578  
TX 521   44   477  
FL 432   44   388  
NY 348   38   310  
PA 304   28   276  
IL 294   21   273  
OH 280   24   256  
MI 254   28   226  
NC 234   21   213  
GA 204   25   179  
WI 175   13   162  
IN 177   16   161  
VA 172   12   160  
WA 182   22   160  
MA 171   18   153  
NJ 167   14   153  
MN 157   11   146  
AZ 151   10   141  
MD 160   20   140  
CO 153   14   139  
TN 151   17   134  
MO 142   11   131  
SC 116   13   103  
KY 111   13    98  
AL 101    6    95  
OR  98   11    87  
OK  93    8    85  
KS  92    9    83  
LA  86    6    80  
UT  74    6    68  
CT  73    7    66  
MS  68    6    62  
ID  64    8    56  
AR  55    4    51  
IA  57    6    51  
NE  46    1    45  
NV  52    7    45  
NM  48    5    43  
WV  42    3    39  
NH  41    6    35  
ME  29    2    27  
HI  27    3    24  
ND  26    4    22  
MT  20    0    20  
RI  19    1    18  
SD  18    1    17  
VT  21    6    15  
DE  13    2    11  
WY  12    1    11  
AK   6    1     5  

3.2 Policy Counts

3.2.1 Enacted and Partially Enacted Policies

New York has been cited, alongside California, as a leader, but it is only the fourth state with the most policies.

I did not anticipate Maryland having so many policies, nor did I realize Colorado has so many.

Hawaii is a very liberal/Democratic state and its flora and fauna seem particularly susceptible to adverse impacts of climate change. I was expecting it to have more policies.

More than half the states (26) have fewer than 14 policies. Three states have only one enacted policy.

Figure 1: Bar Chart of State Policies by Type
state enacted partially_enacted total
CA 47   3   50  
CO 46   4   50  
MA 45   3   48  
NY 44   3   47  
MD 38   3   41  
state enacted partially_enacted total
KS 1   4   5  
MS 1   4   5  
WY 1   4   5  
ID 2   3   5  
SD 3   2   5  
state enacted partially_enacted total in_progress not_enacted
CA 47   3   50   0   12  
CO 46   4   50   0   12  
MA 45   3   48   2   12  
NY 44   3   47   0   15  
MD 38   3   41   3   18  
WA 38   2   40   0   22  
OR 36   4   40   1   21  
MN 34   4   38   1   23  
NJ 32   3   35   1   26  
RI 31   4   35   1   26  
CT 28   4   32   2   28  
ME 26   5   31   0   31  
VT 26   4   30   2   30  
NM 25   4   29   1   32  
DE 23   5   28   1   33  
NC 23   5   28   0   34  
IL 22   5   27   0   35  
PA 21   5   26   0   36  
HI 20   4   24   2   36  
VA 20   4   24   1   37  
MI 21   2   23   1   38  
NV 18   5   23   0   39  
NH 16   3   19   0   43  
WI 13   4   17   3   42  
LA 10   3   13   0   49  
IA  9   4   13   0   49  
TX 10   2   12   0   50  
UT  8   4   12   0   50  
AZ  7   4   11   0   51  
MT  6   4   10   0   52  
OH  6   4   10   0   52  
OK  6   4   10   0   52  
MO  5   5   10   0   52  
FL  6   3    9   0   53  
SC  6   3    9   0   53  
IN  5   4    9   0   53  
KY  5   4    9   0   53  
AR  5   3    8   0   54  
AK  4   4    8   0   54  
GA  4   4    8   0   54  
NE  4   4    8   0   54  
TN  4   3    7   0   55  
WV  3   4    7   0   55  
AL  4   2    6   0   56  
ND  2   4    6   0   56  
SD  3   2    5   0   57  
ID  2   3    5   1   56  
KS  1   4    5   0   57  
MS  1   4    5   0   57  
WY  1   4    5   0   57  

It seems like there are two clusters of states, states with less than 14 policies (26) and states with 23 to 32 policies (12). There are four states with 47 or more policies.

The relative maximum around 27-28 policies indicates a handful of states had around that many policies.

The (unweighted) distribution of states by number of climate policies is positively skewed. There are more states with lower numbers of policies than states with higher numbers of policies.

quantiles_q25 quantiles_q50 quantiles_q75
10 23 40

3.3 Distributions of Individuals’ Attitudes

variable quantile_q25 quantile_q50 quantile_q75
WARMUS 3   4    5  
WARMIMP 2   3    5  
WARMDO 3       4.333 5  
REGGREEN 3       4.333 5  
FSENV 3   4    5  

For four of the five variables, 75% of people had attitudes at or over the midpoint. This means more people than not hold pro-environmental attitudes.

All five variables are negatively skewed. This means more people hold pro-environmental attitudes than not so pro-environmental attitudes.

3.4 Means

3.4.1 Individuals’ Attitudes

The means of all five variables are greater than the midpoint of their respective scale. The lower bounds for the variables’ 95% confidence intervals are all greater than the scale midpoints.

variable mean mean_se mean_low mean_upp
WARMUS 3.526 0.027 3.472 3.58 
WARMIMP 3.313 0.029 3.256 3.37 
WARMDO 3.841 0.024 3.795 3.888
REGGREEN 3.763 0.026 3.712 3.814
FSENV 3.913 0.023 3.868 3.958

3.4.2 States’ Attitudes

3.4.2.1 Maps

state WARMUS_mean WARMIMP_mean WARMDO_mean REGGREEN_mean FSENV_mean
AK 2.722 3.155 3.809 2.846 3.789
AL 3.04  2.895 3.67  3.294 3.64 
AR 3.157 3.067 3.548 3.675 3.697
AZ 3.245 3.097 3.522 3.526 3.736
CA 3.869 3.661 4.048 4.027 4.145
CO 3.653 3.634 4.013 4.025 4.045
CT 3.542 3.335 3.803 4.02  3.821
DE 3.968 3.98  4.535 4.566 4.467
FL 3.398 3.253 3.614 3.748 3.893
GA 3.448 3.141 3.758 3.561 3.662
HI 3.988 3.664 3.891 4.305 4.056
IA 3.641 3.3   4.013 3.814 4.008
ID 2.808 2.877 3.462 3.367 3.617
IL 3.755 3.377 3.922 3.837 4.071
IN 3.34  3.234 3.777 3.626 3.833
KS 3.426 3.114 3.839 3.719 3.686
KY 3.294 3.014 3.339 3.405 3.48 
LA 3.14  2.816 3.533 3.394 3.743
MA 3.885 3.634 4.089 4.2   4.148
MD 3.889 3.704 4.181 4.04  4.188
ME 3.715 3.548 3.868 4.111 3.827
MI 3.257 3.245 3.753 3.595 3.973
MN 3.448 3.221 3.904 3.864 3.938
MO 3.408 3.193 3.803 3.693 3.858
MS 3.572 3.145 3.616 3.501 3.824
MT 2.523 2.22  2.765 2.651 2.681
NC 3.198 3.003 3.558 3.566 3.686
ND 2.879 2.905 3.404 3.567 3.249
NE 3.248 2.75  3.583 3.724 3.621
NH 3.744 3.492 3.99  4.082 4.309
NJ 3.681 3.509 4.083 3.752 4.011
NM 3.765 3.534 3.996 3.767 3.995
NV 4.05  3.718 4.242 3.94  4.234
NY 3.829 3.593 4.052 3.978 4.057
OH 3.321 2.993 3.604 3.632 3.608
OK 3.235 3.091 3.447 3.519 3.739
OR 3.937 3.808 4.097 4.07  4.024
PA 3.612 3.413 3.963 3.766 3.955
RI 2.805 2.502 3.554 3.231 3.546
SC 3.749 3.524 3.906 3.91  4.127
SD 3.271 3.432 3.241 3.332 3.684
TN 3.272 3.077 3.672 3.613 3.692
TX 3.495 3.253 3.856 3.637 3.947
UT 3.439 3.25  3.949 3.896 3.754
VA 3.485 3.205 3.883 3.674 3.921
VT 4.432 3.387 4.294 3.826 4.255
WA 3.765 3.597 4.33  4.091 4.143
WI 3.051 2.788 3.598 3.521 3.729
WV 2.656 2.553 3.096 3.082 3.353
WY 2.735 3.011 3.283 3.601 3.302

3.4.2.2 At least one mean less than the midpoint

Most of these states are generally conservative/Republican. Except as detailed below, these states voted Republican in the Electoral College for the four most recent elections. Ohio voted blue in 2012 but red in three most recent elections. The Democratic candidate has won in all four elections in Rhode Island. Wisconsin voted blue in 2012 and 2020 but red in 2016 and 2024.

state WARMUS_mean WARMIMP_mean WARMDO_mean REGGREEN_mean FSENV_mean
AK 2.722 3.155 3.809 2.846 3.789
AL 3.04  2.895 3.67  3.294 3.64 
ID 2.808 2.877 3.462 3.367 3.617
LA 3.14  2.816 3.533 3.394 3.743
MT 2.523 2.22  2.765 2.651 2.681
ND 2.879 2.905 3.404 3.567 3.249
NE 3.248 2.75  3.583 3.724 3.621
OH 3.321 2.993 3.604 3.632 3.608
RI 2.805 2.502 3.554 3.231 3.546
WI 3.051 2.788 3.598 3.521 3.729
WV 2.656 2.553 3.096 3.082 3.353
WY 2.735 3.011 3.283 3.601 3.302

3.4.2.3 Top 5 highest attitudes

rank WARMUS_mean WARMIMP_mean WARMDO_mean REGGREEN_mean FSENV_mean
1   VT DE DE DE DE
2   NV OR WA HI NH
3   HI NV VT MA VT
4   DE MD NV ME NV
5   OR HI MD WA MD

FIX CHART !!

Three states—Delaware, Massachusetts, and Nevada—have four attitude means among the top 5.

3.4.2.4 Top 10 highest attitudes

rank WARMUS_mean WARMIMP_mean WARMDO_mean REGGREEN_mean FSENV_mean
 1   VT DE DE DE DE
 2   NV OR WA HI NH
 3   HI NV VT MA VT
 4   DE MD NV ME NV
 5   OR HI MD WA MD
 6   MD CA OR NH MA
 7   MA CO MA OR CA
 8   CA MA NJ MD WA
 9   NY WA NY CA SC
10   WA NY CA CO IL

___ states have attitude means in the top 10. California and Colorado, the states with the most and second-most policies only rank in the top 10 for three of the five variables. Neither state has an attitude mean in the top 5.

3.4.3 Policies

variable svymean svymean_se svymean_low svymean_upp
enacted 20.975 0.305 20.377 21.573
in_progress  0.359 0.014  0.332  0.387
partially_enacted  3.454 0.015  3.425  3.483
not_enacted 37.212 0.305 36.614 37.81 
policies 24.429 0.304 23.834 25.024

3.5 Correlations

3.5.1 Individuals’ Attitudes

The correlations between individuals’ attitudes and state policies are weak for all five climate attitudes.

variable corr corr_se corr_low corr_upp
WARMUS 0.154 0.017 0.121 0.187
WARMIMP 0.157 0.017 0.124 0.19 
WARMDO 0.149 0.018 0.114 0.185
REGGREEN 0.145 0.017 0.113 0.178
FSENV 0.128 0.016 0.096 0.159

As I expected, all the coefficients are positive. As one attitude gets stronger across individuals, the other attitudes also get stronger.

The associations between most pairs of attitude variables are moderate. This suggests that the five climate attitudes are related yet distinct constructs. Hence, I created individual regression models in addition to combined ones.

The correlation between WARMUS and WARMIMP stands out. At 0.802, the relationship between WARMUS and WARMIMP is strong. As belief climate change is affecting weather/temperature increases, the importance of climate change generally increases.

variable WARMUS WARMIMP WARMDO REGGREEN FSENV
WARMUS 1        0.802     0.634     0.619     0.552
WARMIMP     0.802 1        0.624     0.622     0.576
WARMDO     0.634     0.624 1        0.599     0.63 
REGGREEN     0.619     0.622     0.599 1        0.521
FSENV     0.552     0.576     0.63      0.521 1   

3.5.2 States’ Attitudes

The correlations between state attitude means and number of policies are moderate (and stronger than those between individuals’ attitudes and number of policies). When aggregating climate attitudes by state, the relationship between attitudes and policies is stronger.

variable corr
WARMUS_mean 0.59 
WARMIMP_mean 0.583
WARMDO_mean 0.626
REGGREEN_mean 0.583
FSENV_mean 0.562

All pairwise correlations are strong (r > 0.7) and positive. As state public opinion for one belief increases, environmental attitudes increase for other beliefs.

The strongest correlation is between WARMDO and FSENV. As public opinion supporting government action increases, public opinion supporting increased federal spending also increases. This is intuitive; government action is costly.

variable WARMUS_mean WARMIMP_mean WARMDO_mean REGGREEN_mean FSENV_mean
WARMUS_mean 1        0.861     0.849     0.831     0.857
WARMIMP_mean     0.861 1        0.837     0.836     0.863
WARMDO_mean     0.849     0.837 1        0.795     0.911
REGGREEN_mean     0.831     0.836     0.795 1        0.781
FSENV_mean     0.857     0.863     0.911     0.781 1   

3.5.3 Maps

Correlations between individuals’ attitudes, grouped by state, and number of policies

3.6 Regression Models

The models that aggregated attitudes by state had higher R2 values than the models based on individuals’ attitudes.

Party identification and left-right ideology were negative predictors in all models, individual and combined, and were both significant at p < 0.001 and p < 0.01, respectively. Stronger Republicans and more conservative people came from states with fewer policies. This relationship was significant for all five climate attitudes. People identifying as (stronger) Republicans and placing themselves farther right on the left-right ideology scale tend to come from states with fewer climate policies.

Individual Models

All variables were significant at p < 0.1, but WARMUS, WARMIMP, and REGGREEN were also significant at p < 0.05.

Despite all variables being significant, all five models had very, very small R2 values. This means each attitude explains almost no variation in the outcome, number of state policies.

Combined Model

None of the attitude variables were significant.

Sociodemographic Control Variables

Income was a significant predictor of individuals’ states having more climate policies. Age was a positive predictor but not significant in individual or combined models. The association between education and policies varied by education level. Some were associated negatively, some were associated positively; none were significant. Not all race/ethnicity categories were significant predictors. There may be race/ethnicity-related reasons people live in their respective state, and this would also relate to number of climate policies.

Research Design

A regression method specifically for ordinal predictors would produce better R2 values. The values for attitudes, 1 to 5, mean there are many, many individuals with the same attitudes but different numbers of policy (see Figure 2).

model_combined_IN model_WARMUS_IN model_WARMIMP_IN model_WARMDO_IN model_REGGREEN_IN model_FSENV_IN
(Intercept) 22.275**** 24.320**** 24.233**** 24.388**** 24.500**** 24.511****
WARMUS 0.240 0.610**
WARMIMP 0.300 0.649**
WARMDO 0.165 0.613*
REGGREEN 0.248 0.599**
FSENV 0.104 0.538*
PTYID_CV -0.660**** -0.762**** -0.757**** -0.762**** -0.805**** -0.814****
LRSELF_CV -0.344*** -0.373*** -0.364*** -0.398*** -0.375*** -0.389***
age_CV 0.005 0.006 0.007 0.005 0.007 0.009
education_CV2 -2.347 -2.309 -2.371 -2.374 -2.368 -2.396
education_CV3 -1.138 -1.098 -1.151 -1.158 -1.185 -1.142
education_CV4 -0.002 0.038 -0.054 -0.030 -0.040 -0.016
education_CV5 0.442 0.458 0.390 0.461 0.396 0.452
income_CV 0.251**** 0.259**** 0.261**** 0.252**** 0.252**** 0.262****
race_CV2 -2.283** -2.414** -2.404** -2.510** -2.307** -2.525**
race_CV3 3.718*** 3.699*** 3.646*** 3.804**** 3.829**** 3.728***
race_CV4 7.596**** 7.635**** 7.639**** 7.731**** 7.619**** 7.680****
race_CV5 -0.791 -0.768 -0.863 -0.801 -0.920 -0.909
race_CV6 1.693 1.727 1.713 1.716 1.677 1.764
sex_CV2 -0.677 -0.652 -0.623 -0.597 -0.624 -0.602
Num.Obs. 6252 6311 6312 6272 6308 6310
R2 0.080 0.079 0.079 0.078 0.078 0.078
R2 Adj. -178.788 -160.491 -160.484 -159.534 -160.561 -160.656
AIC 51916.0 52345.2 52353.0 52059.1 52318.2 52389.5
BIC 60166.5 60250.0 60277.7 59833.5 60219.0 60215.8
RMSE 15.13 15.13 15.12 15.14 15.13 15.13
* p < 0.1, ** p < 0.05, *** p < 0.01, **** p < 0.001

Individual Models

WARMDO explained the most variation in number of policies, and FSENV explained the least amount of variation in number of policies. All five variables were significant predictors at p < 0.001.

Combined Model

The combined model explained about 51% of the variation in number of policies, outperforming all five individual models. In the combined model, only WARMDO was significant.

model_combined_ST model_WARMUS_ST model_WARMIMP_ST model_WARMDO_ST model_REGGREEN_ST model_FSENV_ST
(Intercept) -74.004*** -50.023**** -54.668**** -80.368**** -66.794**** -78.534****
WARMUS_mean 4.333 20.410****
WARMIMP_mean 4.873 23.089****
WARMDO_mean 21.373 26.613****
REGGREEN_mean 6.348 23.461****
FSENV_mean -10.629 25.717****
Num.Obs. 50 50 50 50 50 50
R2 0.421 0.348 0.340 0.392 0.340 0.316
R2 Adj. 0.356 0.334 0.326 0.379 0.326 0.302
AIC 391.8 389.7 390.4 386.3 390.4 392.1
BIC 405.1 395.5 396.1 392.0 396.1 397.9
RMSE 10.58 11.23 11.30 10.85 11.30 11.50
* p < 0.1, ** p < 0.05, *** p < 0.01, **** p < 0.001
Figure 2: Scatterplot for Individuals’ WARMUS Attitudes

3.6.1 Plots: State Attitudes and Policies

Aggregating attitudes at the state level transforms attitudes from ordinal to continuous variables. For linear regression, this means better R2 values.

Of states with fewer than 15 policies, WARMIMP values were the lowest.

4 Conclusion

5 Appendices

5.1 Policy Terms

From the Glossary of the State Climate Policy Dashboard :

Enacted

Enacted policies have been passed or established in a state by a governing body via legislation, executive orders, rules, regulations, and/or other program creation, and remain in effect.

In progress

In progress policies have been established in a state, but final regulations, rules, or plans are pending final approval. This also includes legislation and executive orders that require regulations to be put into effect.

Partially enacted

Partially enacted policies have been enacted in the state, but are missing one or more policy components. Dashboard policies cannot be considered partially enacted unless policy components are available.

Not enacted

Not enacted policies have not been passed or established in the state or are no longer in effect.

Policy opportunity

Policy opportunities are either partially enacted or not enacted.

5.2 Regression Models: Details

5.2.1 Individuals’ attitudes and policies

Stratified 1 - level Cluster Sampling design (with replacement)
With (101) clusters.
Called via srvyr
Sampling variables:
  - ids: wgts_psu 
  - strata: wgts_ST_norata 
  - weights: Weight 

Call:  svyglm(formula = policies ~ WARMUS + PTYID_CV + LRSELF_CV + age_CV + 
    education_CV + income_CV + race_CV + sex_CV, design = .)

Coefficients:
  (Intercept)         WARMUS       PTYID_CV      LRSELF_CV         age_CV  
    24.320376       0.610431      -0.762057      -0.372751       0.005705  
education_CV2  education_CV3  education_CV4  education_CV5      income_CV  
    -2.308514      -1.097786       0.038422       0.457755       0.259235  
     race_CV2       race_CV3       race_CV4       race_CV5       race_CV6  
    -2.413570       3.699071       7.634935      -0.768305       1.726738  
      sex_CV2  
    -0.651886  

Degrees of Freedom: 6310 Total (i.e. Null);  36 Residual
  (7 observations deleted due to missingness)
Null Deviance:      1588000 
Residual Deviance: 1464000  AIC: 54790
MODEL INFO:
Observations: 6311
Dependent Variable: policies
Type: Survey-weighted linear regression 

MODEL FIT:
R² = 0.08
Adj. R² = -0.31 

Standard errors: Robust
--------------------------------------------------
                       Est.   S.E.   t val.      p
------------------- ------- ------ -------- ------
(Intercept)           24.32   2.09    11.63   0.00
WARMUS                 0.61   0.26     2.35   0.02
PTYID_CV              -0.76   0.16    -4.63   0.00
LRSELF_CV             -0.37   0.12    -3.03   0.00
age_CV                 0.01   0.02     0.37   0.72
education_CV2         -2.31   1.57    -1.47   0.15
education_CV3         -1.10   1.52    -0.72   0.48
education_CV4          0.04   1.58     0.02   0.98
education_CV5          0.46   1.75     0.26   0.79
income_CV              0.26   0.05     4.76   0.00
race_CV2              -2.41   1.09    -2.21   0.03
race_CV3               3.70   1.05     3.53   0.00
race_CV4               7.63   1.27     6.03   0.00
race_CV5              -0.77   1.69    -0.46   0.65
race_CV6               1.73   1.62     1.07   0.29
sex_CV2               -0.65   0.50    -1.30   0.20
--------------------------------------------------

Estimated dispersion parameter = 231.94 
Stratified 1 - level Cluster Sampling design (with replacement)
With (101) clusters.
Called via srvyr
Sampling variables:
  - ids: wgts_psu 
  - strata: wgts_ST_norata 
  - weights: Weight 

Call:  svyglm(formula = policies ~ WARMIMP + PTYID_CV + LRSELF_CV + 
    age_CV + education_CV + income_CV + race_CV + sex_CV, design = .)

Coefficients:
  (Intercept)        WARMIMP       PTYID_CV      LRSELF_CV         age_CV  
     24.23301        0.64868       -0.75734       -0.36399        0.00679  
education_CV2  education_CV3  education_CV4  education_CV5      income_CV  
     -2.37056       -1.15054       -0.05372        0.39017        0.26144  
     race_CV2       race_CV3       race_CV4       race_CV5       race_CV6  
     -2.40448        3.64607        7.63852       -0.86297        1.71269  
      sex_CV2  
     -0.62259  

Degrees of Freedom: 6311 Total (i.e. Null);  36 Residual
  (6 observations deleted due to missingness)
Null Deviance:      1589000 
Residual Deviance: 1464000  AIC: 54800
MODEL INFO:
Observations: 6312
Dependent Variable: policies
Type: Survey-weighted linear regression 

MODEL FIT:
R² = 0.08
Adj. R² = -0.30 

Standard errors: Robust
--------------------------------------------------
                       Est.   S.E.   t val.      p
------------------- ------- ------ -------- ------
(Intercept)           24.23   2.06    11.78   0.00
WARMIMP                0.65   0.25     2.57   0.01
PTYID_CV              -0.76   0.16    -4.65   0.00
LRSELF_CV             -0.36   0.13    -2.90   0.01
age_CV                 0.01   0.02     0.43   0.67
education_CV2         -2.37   1.57    -1.51   0.14
education_CV3         -1.15   1.53    -0.75   0.46
education_CV4         -0.05   1.59    -0.03   0.97
education_CV5          0.39   1.76     0.22   0.83
income_CV              0.26   0.05     4.82   0.00
race_CV2              -2.40   1.09    -2.21   0.03
race_CV3               3.65   1.04     3.51   0.00
race_CV4               7.64   1.27     6.03   0.00
race_CV5              -0.86   1.68    -0.51   0.61
race_CV6               1.71   1.61     1.06   0.30
sex_CV2               -0.62   0.49    -1.26   0.22
--------------------------------------------------

Estimated dispersion parameter = 231.96 
Stratified 1 - level Cluster Sampling design (with replacement)
With (101) clusters.
Called via srvyr
Sampling variables:
  - ids: wgts_psu 
  - strata: wgts_ST_norata 
  - weights: Weight 

Call:  svyglm(formula = policies ~ WARMDO + PTYID_CV + LRSELF_CV + age_CV + 
    education_CV + income_CV + race_CV + sex_CV, design = .)

Coefficients:
  (Intercept)         WARMDO       PTYID_CV      LRSELF_CV         age_CV  
    24.387588       0.613332      -0.762034      -0.397736       0.005111  
education_CV2  education_CV3  education_CV4  education_CV5      income_CV  
    -2.373521      -1.157976      -0.030008       0.461218       0.252483  
     race_CV2       race_CV3       race_CV4       race_CV5       race_CV6  
    -2.509603       3.803914       7.730878      -0.801208       1.715920  
      sex_CV2  
    -0.597005  

Degrees of Freedom: 6271 Total (i.e. Null);  36 Residual
  (46 observations deleted due to missingness)
Null Deviance:      1580000 
Residual Deviance: 1456000  AIC: 54450
MODEL INFO:
Observations: 6272
Dependent Variable: policies
Type: Survey-weighted linear regression 

MODEL FIT:
R² = 0.08
Adj. R² = -0.31 

Standard errors: Robust
--------------------------------------------------
                       Est.   S.E.   t val.      p
------------------- ------- ------ -------- ------
(Intercept)           24.39   2.32    10.50   0.00
WARMDO                 0.61   0.34     1.79   0.08
PTYID_CV              -0.76   0.16    -4.62   0.00
LRSELF_CV             -0.40   0.13    -3.18   0.00
age_CV                 0.01   0.02     0.32   0.75
education_CV2         -2.37   1.58    -1.51   0.14
education_CV3         -1.16   1.52    -0.76   0.45
education_CV4         -0.03   1.59    -0.02   0.99
education_CV5          0.46   1.75     0.26   0.79
income_CV              0.25   0.05     4.64   0.00
race_CV2              -2.51   1.10    -2.28   0.03
race_CV3               3.80   1.03     3.71   0.00
race_CV4               7.73   1.24     6.24   0.00
race_CV5              -0.80   1.70    -0.47   0.64
race_CV6               1.72   1.63     1.05   0.30
sex_CV2               -0.60   0.50    -1.19   0.24
--------------------------------------------------

Estimated dispersion parameter = 232.21 
Stratified 1 - level Cluster Sampling design (with replacement)
With (101) clusters.
Called via srvyr
Sampling variables:
  - ids: wgts_psu 
  - strata: wgts_ST_norata 
  - weights: Weight 

Call:  svyglm(formula = policies ~ REGGREEN + PTYID_CV + LRSELF_CV + 
    age_CV + education_CV + income_CV + race_CV + sex_CV, design = .)

Coefficients:
  (Intercept)       REGGREEN       PTYID_CV      LRSELF_CV         age_CV  
    24.499913       0.599117      -0.805170      -0.374780       0.006537  
education_CV2  education_CV3  education_CV4  education_CV5      income_CV  
    -2.368465      -1.185032      -0.039608       0.396069       0.252474  
     race_CV2       race_CV3       race_CV4       race_CV5       race_CV6  
    -2.306654       3.829304       7.619329      -0.919903       1.677453  
      sex_CV2  
    -0.623599  

Degrees of Freedom: 6307 Total (i.e. Null);  36 Residual
  (10 observations deleted due to missingness)
Null Deviance:      1588000 
Residual Deviance: 1465000  AIC: 54770
MODEL INFO:
Observations: 6308
Dependent Variable: policies
Type: Survey-weighted linear regression 

MODEL FIT:
R² = 0.08
Adj. R² = -0.31 

Standard errors: Robust
--------------------------------------------------
                       Est.   S.E.   t val.      p
------------------- ------- ------ -------- ------
(Intercept)           24.50   1.99    12.34   0.00
REGGREEN               0.60   0.26     2.28   0.03
PTYID_CV              -0.81   0.16    -4.93   0.00
LRSELF_CV             -0.37   0.12    -3.12   0.00
age_CV                 0.01   0.02     0.41   0.68
education_CV2         -2.37   1.57    -1.51   0.14
education_CV3         -1.19   1.54    -0.77   0.45
education_CV4         -0.04   1.59    -0.02   0.98
education_CV5          0.40   1.76     0.22   0.82
income_CV              0.25   0.05     4.59   0.00
race_CV2              -2.31   1.10    -2.09   0.04
race_CV3               3.83   1.02     3.74   0.00
race_CV4               7.62   1.23     6.19   0.00
race_CV5              -0.92   1.70    -0.54   0.59
race_CV6               1.68   1.62     1.04   0.31
sex_CV2               -0.62   0.50    -1.25   0.22
--------------------------------------------------

Estimated dispersion parameter = 232.22 
Stratified 1 - level Cluster Sampling design (with replacement)
With (101) clusters.
Called via srvyr
Sampling variables:
  - ids: wgts_psu 
  - strata: wgts_ST_norata 
  - weights: Weight 

Call:  svyglm(formula = policies ~ FSENV + PTYID_CV + LRSELF_CV + age_CV + 
    education_CV + income_CV + race_CV + sex_CV, design = .)

Coefficients:
  (Intercept)          FSENV       PTYID_CV      LRSELF_CV         age_CV  
    24.510559       0.537903      -0.813615      -0.388721       0.008728  
education_CV2  education_CV3  education_CV4  education_CV5      income_CV  
    -2.395933      -1.141873      -0.016450       0.451946       0.262369  
     race_CV2       race_CV3       race_CV4       race_CV5       race_CV6  
    -2.524851       3.727995       7.680157      -0.908957       1.763738  
      sex_CV2  
    -0.601715  

Degrees of Freedom: 6309 Total (i.e. Null);  36 Residual
  (8 observations deleted due to missingness)
Null Deviance:      1590000 
Residual Deviance: 1466000  AIC: 54780
MODEL INFO:
Observations: 6310
Dependent Variable: policies
Type: Survey-weighted linear regression 

MODEL FIT:
R² = 0.08
Adj. R² = -0.31 

Standard errors: Robust
--------------------------------------------------
                       Est.   S.E.   t val.      p
------------------- ------- ------ -------- ------
(Intercept)           24.51   2.26    10.85   0.00
FSENV                  0.54   0.31     1.73   0.09
PTYID_CV              -0.81   0.16    -5.12   0.00
LRSELF_CV             -0.39   0.13    -3.03   0.00
age_CV                 0.01   0.02     0.56   0.58
education_CV2         -2.40   1.55    -1.55   0.13
education_CV3         -1.14   1.51    -0.75   0.46
education_CV4         -0.02   1.57    -0.01   0.99
education_CV5          0.45   1.74     0.26   0.80
income_CV              0.26   0.05     4.83   0.00
race_CV2              -2.52   1.08    -2.33   0.03
race_CV3               3.73   1.04     3.58   0.00
race_CV4               7.68   1.26     6.11   0.00
race_CV5              -0.91   1.70    -0.53   0.60
race_CV6               1.76   1.61     1.10   0.28
sex_CV2               -0.60   0.50    -1.21   0.23
--------------------------------------------------

Estimated dispersion parameter = 232.42 
Stratified 1 - level Cluster Sampling design (with replacement)
With (101) clusters.
Called via srvyr
Sampling variables:
  - ids: wgts_psu 
  - strata: wgts_ST_norata 
  - weights: Weight 

Call:  svyglm(formula = policies ~ WARMUS + WARMIMP + WARMDO + REGGREEN + 
    FSENV + PTYID_CV + LRSELF_CV + age_CV + education_CV + income_CV + 
    race_CV + sex_CV, design = .)

Coefficients:
  (Intercept)         WARMUS        WARMIMP         WARMDO       REGGREEN  
    22.275271       0.239654       0.300151       0.164910       0.248054  
        FSENV       PTYID_CV      LRSELF_CV         age_CV  education_CV2  
     0.103897      -0.659772      -0.343578       0.004912      -2.346703  
education_CV3  education_CV4  education_CV5      income_CV       race_CV2  
    -1.137960      -0.001628       0.441770       0.251386      -2.283109  
     race_CV3       race_CV4       race_CV5       race_CV6        sex_CV2  
     3.718413       7.596191      -0.791075       1.693250      -0.677234  

Degrees of Freedom: 6251 Total (i.e. Null);  32 Residual
  (66 observations deleted due to missingness)
Null Deviance:      1576000 
Residual Deviance: 1450000  AIC: 54270
MODEL INFO:
Observations: 6252
Dependent Variable: policies
Type: Survey-weighted linear regression 

MODEL FIT:
R² = 0.08
Adj. R² = -0.47 

Standard errors: Robust
--------------------------------------------------
                       Est.   S.E.   t val.      p
------------------- ------- ------ -------- ------
(Intercept)           22.28   2.50     8.91   0.00
WARMUS                 0.24   0.38     0.64   0.53
WARMIMP                0.30   0.34     0.89   0.38
WARMDO                 0.16   0.44     0.38   0.71
REGGREEN               0.25   0.39     0.63   0.53
FSENV                  0.10   0.37     0.28   0.78
PTYID_CV              -0.66   0.17    -3.85   0.00
LRSELF_CV             -0.34   0.12    -2.77   0.01
age_CV                 0.00   0.02     0.31   0.76
education_CV2         -2.35   1.60    -1.47   0.15
education_CV3         -1.14   1.54    -0.74   0.47
education_CV4         -0.00   1.60    -0.00   1.00
education_CV5          0.44   1.77     0.25   0.80
income_CV              0.25   0.05     4.60   0.00
race_CV2              -2.28   1.11    -2.06   0.05
race_CV3               3.72   1.04     3.59   0.00
race_CV4               7.60   1.24     6.12   0.00
race_CV5              -0.79   1.69    -0.47   0.64
race_CV6               1.69   1.64     1.03   0.31
sex_CV2               -0.68   0.50    -1.34   0.19
--------------------------------------------------

Estimated dispersion parameter = 232.01

5.2.2 State mean attitudes and policies


Call:
lm(formula = policies ~ WARMUS_mean, data = .)

Coefficients:
(Intercept)  WARMUS_mean  
     -50.02        20.41  
MODEL INFO:
Observations: 50
Dependent Variable: policies
Type: OLS linear regression 

MODEL FIT:
F(1,48) = 25.62, p = 0.00
R² = 0.35
Adj. R² = 0.33 

Standard errors:OLS
--------------------------------------------------
                      Est.    S.E.   t val.      p
----------------- -------- ------- -------- ------
(Intercept)         -50.02   13.95    -3.59   0.00
WARMUS_mean          20.41    4.03     5.06   0.00
--------------------------------------------------

Call:
lm(formula = policies ~ WARMIMP_mean, data = .)

Coefficients:
 (Intercept)  WARMIMP_mean  
      -54.67         23.09  
MODEL INFO:
Observations: 50
Dependent Variable: policies
Type: OLS linear regression 

MODEL FIT:
F(1,48) = 24.71, p = 0.00
R² = 0.34
Adj. R² = 0.33 

Standard errors:OLS
---------------------------------------------------
                       Est.    S.E.   t val.      p
------------------ -------- ------- -------- ------
(Intercept)          -54.67   15.13    -3.61   0.00
WARMIMP_mean          23.09    4.65     4.97   0.00
---------------------------------------------------

Call:
lm(formula = policies ~ WARMDO_mean, data = .)

Coefficients:
(Intercept)  WARMDO_mean  
     -80.37        26.61  
MODEL INFO:
Observations: 50
Dependent Variable: policies
Type: OLS linear regression 

MODEL FIT:
F(1,48) = 30.89, p = 0.00
R² = 0.39
Adj. R² = 0.38 

Standard errors:OLS
--------------------------------------------------
                      Est.    S.E.   t val.      p
----------------- -------- ------- -------- ------
(Intercept)         -80.37   18.14    -4.43   0.00
WARMDO_mean          26.61    4.79     5.56   0.00
--------------------------------------------------

Call:
lm(formula = policies ~ REGGREEN_mean, data = .)

Coefficients:
  (Intercept)  REGGREEN_mean  
       -66.79          23.46  
MODEL INFO:
Observations: 50
Dependent Variable: policies
Type: OLS linear regression 

MODEL FIT:
F(1,48) = 24.68, p = 0.00
R² = 0.34
Adj. R² = 0.33 

Standard errors:OLS
----------------------------------------------------
                        Est.    S.E.   t val.      p
------------------- -------- ------- -------- ------
(Intercept)           -66.79   17.57    -3.80   0.00
REGGREEN_mean          23.46    4.72     4.97   0.00
----------------------------------------------------

Call:
lm(formula = policies ~ FSENV_mean, data = .)

Coefficients:
(Intercept)   FSENV_mean  
     -78.53        25.72  
MODEL INFO:
Observations: 50
Dependent Variable: policies
Type: OLS linear regression 

MODEL FIT:
F(1,48) = 22.16, p = 0.00
R² = 0.32
Adj. R² = 0.30 

Standard errors:OLS
--------------------------------------------------
                      Est.    S.E.   t val.      p
----------------- -------- ------- -------- ------
(Intercept)         -78.53   21.02    -3.74   0.00
FSENV_mean           25.72    5.46     4.71   0.00
--------------------------------------------------

Call:
lm(formula = policies ~ WARMUS_mean + WARMIMP_mean + WARMDO_mean + 
    REGGREEN_mean + FSENV_mean, data = .)

Coefficients:
  (Intercept)    WARMUS_mean   WARMIMP_mean    WARMDO_mean  REGGREEN_mean  
      -74.004          4.333          4.873         21.373          6.348  
   FSENV_mean  
      -10.629  
MODEL INFO:
Observations: 50
Dependent Variable: policies
Type: OLS linear regression 

MODEL FIT:
F(5,44) = 6.41, p = 0.00
R² = 0.42
Adj. R² = 0.36 

Standard errors:OLS
----------------------------------------------------
                        Est.    S.E.   t val.      p
------------------- -------- ------- -------- ------
(Intercept)           -74.00   22.97    -3.22   0.00
WARMUS_mean             4.33    9.42     0.46   0.65
WARMIMP_mean            4.87   10.93     0.45   0.66
WARMDO_mean            21.37   12.73     1.68   0.10
REGGREEN_mean           6.35    9.32     0.68   0.50
FSENV_mean            -10.63   14.49    -0.73   0.47
----------------------------------------------------

ANOVA for state attitudes and policies

Figure 3
variable Df Sum Sq Mean Sq F value Pr(>F)
WARMUS_mean  1  3,364.379 3,364.379 26.463 5.980 × 10−6
WARMIMP_mean  1    209.942   209.942  1.651 0.206
WARMDO_mean  1    364.183   364.183  2.865 0.098
REGGREEN_mean  1     67.598    67.598  0.532 0.47
FSENV_mean  1     68.429    68.429  0.538 0.467
Residuals 44  5,593.969   127.136 NA NA

5.3 Regressions without Sociodemographic Controls

5.3.1 Individuals’ attitudes and policies

Stratified 1 - level Cluster Sampling design (with replacement)
With (101) clusters.
Called via srvyr
Sampling variables:
  - ids: wgts_psu 
  - strata: wgts_ST_norata 
  - weights: Weight 

Call:  svyglm(formula = policies ~ WARMUS + PTYID_CV + LRSELF_CV, design = .)

Coefficients:
(Intercept)       WARMUS     PTYID_CV    LRSELF_CV  
    25.7634       1.0414      -0.4575      -0.5466  

Degrees of Freedom: 6970 Total (i.e. Null);  48 Residual
  (10 observations deleted due to missingness)
Null Deviance:      1770000 
Residual Deviance: 1699000  AIC: 60850
MODEL INFO:
Observations: 6971
Dependent Variable: policies
Type: Survey-weighted linear regression 

MODEL FIT:
R² = 0.04
Adj. R² = -0.02 

Standard errors: Robust
------------------------------------------------
                     Est.   S.E.   t val.      p
----------------- ------- ------ -------- ------
(Intercept)         25.76   1.32    19.46   0.00
WARMUS               1.04   0.24     4.30   0.00
PTYID_CV            -0.46   0.17    -2.77   0.01
LRSELF_CV           -0.55   0.12    -4.58   0.00
------------------------------------------------

Estimated dispersion parameter = 243.8 
Stratified 1 - level Cluster Sampling design (with replacement)
With (101) clusters.
Called via srvyr
Sampling variables:
  - ids: wgts_psu 
  - strata: wgts_ST_norata 
  - weights: Weight 

Call:  svyglm(formula = policies ~ WARMIMP + PTYID_CV + LRSELF_CV, design = .)

Coefficients:
(Intercept)      WARMIMP     PTYID_CV    LRSELF_CV  
    25.8509       1.0655      -0.4586      -0.5359  

Degrees of Freedom: 6972 Total (i.e. Null);  48 Residual
  (8 observations deleted due to missingness)
Null Deviance:      1771000 
Residual Deviance: 1700000  AIC: 60870
MODEL INFO:
Observations: 6973
Dependent Variable: policies
Type: Survey-weighted linear regression 

MODEL FIT:
R² = 0.04
Adj. R² = -0.02 

Standard errors: Robust
------------------------------------------------
                     Est.   S.E.   t val.      p
----------------- ------- ------ -------- ------
(Intercept)         25.85   1.36    19.07   0.00
WARMIMP              1.07   0.24     4.45   0.00
PTYID_CV            -0.46   0.16    -2.85   0.01
LRSELF_CV           -0.54   0.13    -4.21   0.00
------------------------------------------------

Estimated dispersion parameter = 243.84 
Stratified 1 - level Cluster Sampling design (with replacement)
With (101) clusters.
Called via srvyr
Sampling variables:
  - ids: wgts_psu 
  - strata: wgts_ST_norata 
  - weights: Weight 

Call:  svyglm(formula = policies ~ WARMDO + PTYID_CV + LRSELF_CV, design = .)

Coefficients:
(Intercept)       WARMDO     PTYID_CV    LRSELF_CV  
    26.4689       0.6380      -0.4958      -0.5901  

Degrees of Freedom: 6914 Total (i.e. Null);  48 Residual
  (66 observations deleted due to missingness)
Null Deviance:      1758000 
Residual Deviance: 1691000  AIC: 60370
MODEL INFO:
Observations: 6915
Dependent Variable: policies
Type: Survey-weighted linear regression 

MODEL FIT:
R² = 0.04
Adj. R² = -0.02 

Standard errors: Robust
------------------------------------------------
                     Est.   S.E.   t val.      p
----------------- ------- ------ -------- ------
(Intercept)         26.47   1.75    15.10   0.00
WARMDO               0.64   0.22     2.87   0.01
PTYID_CV            -0.50   0.17    -2.90   0.01
LRSELF_CV           -0.59   0.12    -4.73   0.00
------------------------------------------------

Estimated dispersion parameter = 244.54 
Stratified 1 - level Cluster Sampling design (with replacement)
With (101) clusters.
Called via srvyr
Sampling variables:
  - ids: wgts_psu 
  - strata: wgts_ST_norata 
  - weights: Weight 

Call:  svyglm(formula = policies ~ REGGREEN + PTYID_CV + LRSELF_CV, 
    design = .)

Coefficients:
(Intercept)     REGGREEN     PTYID_CV    LRSELF_CV  
    26.2673       0.6827      -0.5578      -0.5329  

Degrees of Freedom: 6961 Total (i.e. Null);  48 Residual
  (19 observations deleted due to missingness)
Null Deviance:      1770000 
Residual Deviance: 1700000  AIC: 60780
MODEL INFO:
Observations: 6962
Dependent Variable: policies
Type: Survey-weighted linear regression 

MODEL FIT:
R² = 0.04
Adj. R² = -0.02 

Standard errors: Robust
------------------------------------------------
                     Est.   S.E.   t val.      p
----------------- ------- ------ -------- ------
(Intercept)         26.27   1.26    20.85   0.00
REGGREEN             0.68   0.16     4.17   0.00
PTYID_CV            -0.56   0.16    -3.57   0.00
LRSELF_CV           -0.53   0.12    -4.45   0.00
------------------------------------------------

Estimated dispersion parameter = 244.21 
Stratified 1 - level Cluster Sampling design (with replacement)
With (101) clusters.
Called via srvyr
Sampling variables:
  - ids: wgts_psu 
  - strata: wgts_ST_norata 
  - weights: Weight 

Call:  svyglm(formula = policies ~ FSENV + PTYID_CV + LRSELF_CV, design = .)

Coefficients:
(Intercept)        FSENV     PTYID_CV    LRSELF_CV  
    28.0023       0.6011      -0.5983      -0.6141  

Degrees of Freedom: 6966 Total (i.e. Null);  48 Residual
  (14 observations deleted due to missingness)
Null Deviance:      1770000 
Residual Deviance: 1707000  AIC: 60840
MODEL INFO:
Observations: 6967
Dependent Variable: policies
Type: Survey-weighted linear regression 

MODEL FIT:
R² = 0.04
Adj. R² = -0.03 

Standard errors: Robust
------------------------------------------------
                     Est.   S.E.   t val.      p
----------------- ------- ------ -------- ------
(Intercept)         28.00   1.78    15.71   0.00
FSENV                0.60   0.32     1.90   0.06
PTYID_CV            -0.60   0.16    -3.78   0.00
LRSELF_CV           -0.61   0.13    -4.72   0.00
------------------------------------------------

Estimated dispersion parameter = 245.03 
Stratified 1 - level Cluster Sampling design (with replacement)
With (101) clusters.
Called via srvyr
Sampling variables:
  - ids: wgts_psu 
  - strata: wgts_ST_norata 
  - weights: Weight 

Call:  svyglm(formula = policies ~ FSENV + WARMDO + WARMUS + WARMIMP + 
    REGGREEN + PTYID_CV + LRSELF_CV, design = .)

Coefficients:
(Intercept)        FSENV       WARMDO       WARMUS      WARMIMP     REGGREEN  
    23.5435      -0.1830       0.1822       0.4342       0.4942       0.3565  
   PTYID_CV    LRSELF_CV  
    -0.3667      -0.4911  

Degrees of Freedom: 6887 Total (i.e. Null);  44 Residual
  (93 observations deleted due to missingness)
Null Deviance:      1753000 
Residual Deviance: 1678000  AIC: 60110
MODEL INFO:
Observations: 6888
Dependent Variable: policies
Type: Survey-weighted linear regression 

MODEL FIT:
R² = 0.04
Adj. R² = -0.11 

Standard errors: Robust
------------------------------------------------
                     Est.   S.E.   t val.      p
----------------- ------- ------ -------- ------
(Intercept)         23.54   1.98    11.90   0.00
FSENV               -0.18   0.37    -0.49   0.63
WARMDO               0.18   0.28     0.64   0.53
WARMUS               0.43   0.37     1.18   0.24
WARMIMP              0.49   0.36     1.38   0.17
REGGREEN             0.36   0.24     1.50   0.14
PTYID_CV            -0.37   0.18    -2.08   0.04
LRSELF_CV           -0.49   0.12    -4.03   0.00
------------------------------------------------

Estimated dispersion parameter = 243.61

5.3.2 State mean attitudes and policies


Call:
lm(formula = policies ~ WARMUS_mean, data = .)

Coefficients:
(Intercept)  WARMUS_mean  
     -56.26        22.59  
MODEL INFO:
Observations: 50
Dependent Variable: policies
Type: OLS linear regression 

MODEL FIT:
F(1,48) = 33.57, p = 0.00
R² = 0.41
Adj. R² = 0.40 

Standard errors:OLS
--------------------------------------------------
                      Est.    S.E.   t val.      p
----------------- -------- ------- -------- ------
(Intercept)         -56.26   13.27    -4.24   0.00
WARMUS_mean          22.59    3.90     5.79   0.00
--------------------------------------------------

Call:
lm(formula = policies ~ WARMIMP_mean, data = .)

Coefficients:
 (Intercept)  WARMIMP_mean  
      -60.89         25.28  
MODEL INFO:
Observations: 50
Dependent Variable: policies
Type: OLS linear regression 

MODEL FIT:
F(1,48) = 30.77, p = 0.00
R² = 0.39
Adj. R² = 0.38 

Standard errors:OLS
---------------------------------------------------
                       Est.    S.E.   t val.      p
------------------ -------- ------- -------- ------
(Intercept)          -60.89   14.68    -4.15   0.00
WARMIMP_mean          25.28    4.56     5.55   0.00
---------------------------------------------------

Call:
lm(formula = policies ~ WARMDO_mean, data = .)

Coefficients:
(Intercept)  WARMDO_mean  
     -75.07        18.60  
MODEL INFO:
Observations: 50
Dependent Variable: policies
Type: OLS linear regression 

MODEL FIT:
F(1,48) = 35.91, p = 0.00
R² = 0.43
Adj. R² = 0.42 

Standard errors:OLS
--------------------------------------------------
                      Est.    S.E.   t val.      p
----------------- -------- ------- -------- ------
(Intercept)         -75.07   15.95    -4.71   0.00
WARMDO_mean          18.60    3.10     5.99   0.00
--------------------------------------------------

Call:
lm(formula = policies ~ REGGREEN_mean, data = .)

Coefficients:
  (Intercept)  REGGREEN_mean  
       -71.42          18.28  
MODEL INFO:
Observations: 50
Dependent Variable: policies
Type: OLS linear regression 

MODEL FIT:
F(1,48) = 33.27, p = 0.00
R² = 0.41
Adj. R² = 0.40 

Standard errors:OLS
----------------------------------------------------
                        Est.    S.E.   t val.      p
------------------- -------- ------- -------- ------
(Intercept)           -71.42   15.94    -4.48   0.00
REGGREEN_mean          18.28    3.17     5.77   0.00
----------------------------------------------------

Call:
lm(formula = policies ~ FSENV_mean, data = .)

Coefficients:
(Intercept)   FSENV_mean  
     -76.55        25.37  
MODEL INFO:
Observations: 50
Dependent Variable: policies
Type: OLS linear regression 

MODEL FIT:
F(1,48) = 21.82, p = 0.00
R² = 0.31
Adj. R² = 0.30 

Standard errors:OLS
--------------------------------------------------
                      Est.    S.E.   t val.      p
----------------- -------- ------- -------- ------
(Intercept)         -76.55   20.76    -3.69   0.00
FSENV_mean           25.37    5.43     4.67   0.00
--------------------------------------------------

Call:
lm(formula = policies ~ WARMUS_mean + WARMIMP_mean + WARMDO_mean + 
    REGGREEN_mean + FSENV_mean, data = .)

Coefficients:
  (Intercept)    WARMUS_mean   WARMIMP_mean    WARMDO_mean  REGGREEN_mean  
      -63.920          8.342          5.097         18.633          4.535  
   FSENV_mean  
      -20.624  
MODEL INFO:
Observations: 50
Dependent Variable: policies
Type: OLS linear regression 

MODEL FIT:
F(5,44) = 9.05, p = 0.00
R² = 0.51
Adj. R² = 0.45 

Standard errors:OLS
----------------------------------------------------
                        Est.    S.E.   t val.      p
------------------- -------- ------- -------- ------
(Intercept)           -63.92   20.38    -3.14   0.00
WARMUS_mean             8.34    8.83     0.94   0.35
WARMIMP_mean            5.10   10.59     0.48   0.63
WARMDO_mean            18.63    7.97     2.34   0.02
REGGREEN_mean           4.53    6.87     0.66   0.51
FSENV_mean            -20.62   13.00    -1.59   0.12
----------------------------------------------------

5.3.3 Model Comparison

Models with sociodemographic control variables vs. models without

model_combined_IN model_combined_IN_no model_WARMUS_IN model_WARMUS_IN_no model_WARMIMP_IN model_WARMIMP_IN_no model_WARMDO_IN model_WARMDO_IN_no model_REGGREEN_IN model_REGGREEN_IN_no model_FSENV_IN model_FSENV_IN_no
(Intercept) 22.275**** 23.543**** 24.320**** 25.763**** 24.233**** 25.851**** 24.388**** 26.469**** 24.500**** 26.267**** 24.511**** 28.002****
WARMUS 0.240 0.434 0.610** 1.041****
WARMIMP 0.300 0.494 0.649** 1.065****
WARMDO 0.165 0.182 0.613* 0.638***
REGGREEN 0.248 0.356 0.599** 0.683****
FSENV 0.104 -0.183 0.538* 0.601*
PTYID_CV -0.660**** -0.367** -0.762**** -0.458*** -0.757**** -0.459*** -0.762**** -0.496*** -0.805**** -0.558**** -0.814**** -0.598****
LRSELF_CV -0.344*** -0.491**** -0.373*** -0.547**** -0.364*** -0.536**** -0.398*** -0.590**** -0.375*** -0.533**** -0.389*** -0.614****
age_CV 0.005 0.006 0.007 0.005 0.007 0.009
education_CV2 -2.347 -2.309 -2.371 -2.374 -2.368 -2.396
education_CV3 -1.138 -1.098 -1.151 -1.158 -1.185 -1.142
education_CV4 -0.002 0.038 -0.054 -0.030 -0.040 -0.016
education_CV5 0.442 0.458 0.390 0.461 0.396 0.452
income_CV 0.251**** 0.259**** 0.261**** 0.252**** 0.252**** 0.262****
race_CV2 -2.283** -2.414** -2.404** -2.510** -2.307** -2.525**
race_CV3 3.718*** 3.699*** 3.646*** 3.804**** 3.829**** 3.728***
race_CV4 7.596**** 7.635**** 7.639**** 7.731**** 7.619**** 7.680****
race_CV5 -0.791 -0.768 -0.863 -0.801 -0.920 -0.909
race_CV6 1.693 1.727 1.713 1.716 1.677 1.764
sex_CV2 -0.677 -0.652 -0.623 -0.597 -0.624 -0.602
Num.Obs. 6252 6888 6311 6971 6312 6973 6272 6915 6308 6962 6310 6967
R2 0.080 0.043 0.079 0.040 0.079 0.040 0.078 0.038 0.078 0.039 0.078 0.036
R2 Adj. -178.788 -148.784 -160.491 -138.387 -160.484 -138.415 -159.534 -137.507 -160.561 -138.307 -160.656 -138.926
AIC 51916.0 57509.3 52345.2 58103.6 52353.0 58127.2 52059.1 57712.6 52318.2 58063.9 52389.5 58155.6
BIC 60166.5 65778.2 60250.0 66356.3 60277.7 66397.1 59833.5 65823.4 60219.0 66272.7 60215.8 66333.1
RMSE 15.13 15.37 15.13 15.38 15.12 15.36 15.14 15.38 15.13 15.38 15.13 15.38
* p < 0.1, ** p < 0.05, *** p < 0.01, **** p < 0.001
model_combined_ST model_combined_ST_no model_WARMUS_ST model_WARMUS_ST_no model_WARMIMP_ST model_WARMIMP_ST_no model_WARMDO_ST model_WARMDO_ST_no model_REGGREEN_ST model_REGGREEN_ST_no model_FSENV_ST model_FSENV_ST_no
(Intercept) -74.004*** -63.920*** -50.023**** -56.264**** -54.668**** -60.893**** -80.368**** -75.070**** -66.794**** -71.419**** -78.534**** -76.553****
WARMUS_mean 4.333 8.342 20.410**** 22.590****
WARMIMP_mean 4.873 5.097 23.089**** 25.278****
WARMDO_mean 21.373 18.633** 26.613**** 18.600****
REGGREEN_mean 6.348 4.535 23.461**** 18.278****
FSENV_mean -10.629 -20.624 25.717**** 25.375****
Num.Obs. 50 50 50 50 50 50 50 50 50 50 50 50
R2 0.421 0.507 0.348 0.412 0.340 0.391 0.392 0.428 0.340 0.409 0.316 0.313
R2 Adj. 0.356 0.451 0.334 0.399 0.326 0.378 0.379 0.416 0.326 0.397 0.302 0.298
AIC 391.8 383.8 389.7 384.6 390.4 386.4 386.3 383.2 390.4 384.8 392.1 392.4
BIC 405.1 397.1 395.5 390.3 396.1 392.1 392.0 388.9 396.1 390.5 397.9 398.1
RMSE 10.58 9.76 11.23 10.67 11.30 10.85 10.85 10.52 11.30 10.69 11.50 11.53
* p < 0.1, ** p < 0.05, *** p < 0.01, **** p < 0.001