search

alarm-redist / fifty-states

Redistricting analysis for all 50 U.S. states
https://alarm-redist.github.io/fifty-states/
Other
9 stars 7 forks source link

2010 New Jersey Congressional Districts #160

Closed Jfer09 closed 1 year ago

Jfer09 commented 1 year ago

Redistricting requirements

In New Jersey, districts must:

  1. be contiguous
  2. have equal populations

Interpretation of requirements

We enforce a maximum population deviation of 0.5%. We use a pseudo-county constraint described below which attempts to mimic the norms in New Jersey of generally preserving county and municipal boundaries.

Data Sources

Data for New Jersey comes from the ALARM Project's 2010 Redistricting Data Files.

Pre-processing Notes

No manual pre-processing decisions were necessary.

Simulation Notes

We sample 10,000 districting plans for New Jersey across two independent runs of the SMC algorithm, and then thin the sample to down to 5,000 plans. To balance county and municipality splits, we create pseudocounties for use in the county constraint. These are counties, outside of Bergen, Burlington, Camden, Essex, Hudson, Mercer, Middlesex, Monmouth, Morris, Ocean, Passaic, and Union County, which are the counties larger than 50% of the target district population. WithinBergen, Burlington, Camden, Essex, Hudson, Mercer, Middlesex, Monmouth, Morris, Ocean, Passaic, and Union County, which are the counties larger than 50% of the target district population, each municipality is its own pseudocounty as well.

Validation

validation_20230120_2125

SMC: 10,000 sampled plans of 12 districts on 6,344 units
`adapt_k_thresh`=0.985 • `seq_alpha`=0.5
`est_label_mult`=1 • `pop_temper`=0

Plan diversity 80% range: 0.69 to 0.89

R-hat values for summary statistics:
   pop_overlap      total_vap       plan_dev      comp_edge    comp_polsby      pop_white 
      1.028225       1.008622       1.018606       1.001469       1.045092       1.039168 
     pop_black       pop_hisp       pop_aian      pop_asian       pop_nhpi      pop_other 
      1.040636       1.001220       1.044586       1.019275       1.016296       1.008444 
       pop_two      vap_white      vap_black       vap_hisp       vap_aian      vap_asian 
      1.030445       1.022881       1.031330       1.001028       1.035543       1.016141 
      vap_nhpi      vap_other        vap_two pre_16_dem_cli pre_16_rep_tru pre_20_dem_bid 
      1.003926       1.005651       1.025073       1.020200       1.035463       1.023895 
pre_20_rep_tru uss_18_dem_men uss_18_rep_hug uss_20_dem_boo uss_20_rep_meh         adv_16 
      1.040345       1.022220       1.012714       1.022041       1.031613       1.020200 
        adv_18         adv_20         arv_16         arv_18         arv_20  county_splits 
      1.022220       1.023329       1.035463       1.012714       1.044714       1.010899 
   muni_splits            ndv            nrv        ndshare          e_dvs         pr_dem 
      1.001378       1.022002       1.031914       1.026862       1.027341       1.007064 
         e_dem          pbias           egap 
      1.019315       1.019660       1.022577 

Sampling diagnostics for SMC run 1 of 2 (5,000 samples)
         Eff. samples (%) Acc. rate Log wgt. sd  Max. unique Est. k 
Split 1     4,907 (98.1%)     11.7%        0.28 3,251 (103%)     15 
Split 2     4,824 (96.5%)     17.6%        0.40 3,103 ( 98%)      9 
Split 3     4,737 (94.7%)     27.3%        0.49 3,092 ( 98%)      5 
Split 4     4,701 (94.0%)     35.1%        0.51 3,088 ( 98%)      3 
Split 5     4,650 (93.0%)     38.4%        0.54 3,088 ( 98%)      2 
Split 6     4,613 (92.3%)     34.8%        0.56 3,058 ( 97%)      2 
Split 7     4,519 (90.4%)     31.9%        0.59 3,055 ( 97%)      2 
Split 8     4,466 (89.3%)     22.4%        0.60 2,985 ( 94%)      3 
Split 9     4,503 (90.1%)     20.1%        0.59 2,900 ( 92%)      3 
Split 10    4,511 (90.2%)     16.7%        0.59 2,779 ( 88%)      3 
Split 11    4,510 (90.2%)      8.2%        0.58 2,444 ( 77%)      2 
Resample    2,916 (58.3%)       NA%        0.59 3,771 (119%)     NA 

Sampling diagnostics for SMC run 2 of 2 (5,000 samples)
         Eff. samples (%) Acc. rate Log wgt. sd  Max. unique Est. k 
Split 1     4,903 (98.1%)     14.1%        0.29 3,166 (100%)     12 
Split 2     4,817 (96.3%)     19.5%        0.41 3,140 ( 99%)      8 
Split 3     4,743 (94.9%)     27.4%        0.48 3,096 ( 98%)      5 
Split 4     4,652 (93.0%)     35.4%        0.53 3,099 ( 98%)      3 
Split 5     4,621 (92.4%)     31.5%        0.55 3,091 ( 98%)      3 
Split 6     4,589 (91.8%)     23.0%        0.57 3,057 ( 97%)      4 
Split 7     4,541 (90.8%)     25.7%        0.57 3,016 ( 95%)      3 
Split 8     4,546 (90.9%)     17.7%        0.58 2,978 ( 94%)      4 
Split 9     4,523 (90.5%)     19.5%        0.58 2,923 ( 92%)      3 
Split 10    4,593 (91.9%)     18.8%        0.56 2,851 ( 90%)      2 
Split 11    4,568 (91.4%)      6.9%        0.55 2,455 ( 78%)      2 
Resample    3,399 (68.0%)       NA%        0.57 3,788 (120%)     NA

Checklist

delete this line and all the tags except the reviewers you need @CoryMcCartan @christopherkenny

CoryMcCartan commented 1 year ago

Sims look good! Can you edit the top of the documentation to read:

## Redistricting requirements
In New Jersey, districts must:

1. be contiguous
1. have equal populations

### Interpretation of requirements
We enforce a maximum population deviation of 0.5%.
We use a pseudo-county constraint described below which attempts to mimic the norms in New Jersey of generally preserving county and municipal boundaries.

We have not been mentioning the VRA as a redistricting constraint, for whatever reason (maybe because it's federal?). And we do like to mention the pseudo-county constraint in both 'interpretation' and 'Simulation Notes'

Jfer09 commented 1 year ago

done! (I did change the second 1. to a 2. which I assumed was a typo haha)

CoryMcCartan commented 1 year ago

@christopherkenny good by me! feel free to merge when you're ready