Code 
source("scripts/descriptive_stats_22_lags_cluster_AWM_v7_4_1.R")
quarto::quarto_render("levy_paper_tables_v7_4.qmd")

Table 1. Distribution of annual changes in levies and EAV

Code 
include_table("v22_table_01_distribution.html")
Table 1: Table 1. Distribution of annual percentage changes in levies and EAV
agency variable N p5 p25 p50 p75 p95
Muni d_levy 1572 -1.3 0.9 2.6 4.6 11.8
Muni d_eav 1575 -15.5 -6.9 -1.5 4.2 24.1
Other d_levy 2145 -4.6 1.1 2.3 3.4 8.3
Other d_eav 2145 -16.3 -7.2 -1.6 4.1 24.1
School d_levy 1995 -2.7 1.4 2.3 3.5 8.6
School d_eav 1995 -14.8 -7.1 -1.6 4.1 23.3
Township d_levy 420 -2.0 1.2 2.7 4.1 6.8
Township d_eav 420 -14.1 -6.8 -1.4 3.9 21.6
All d_levy 6132 -2.9 1.1 2.4 3.8 9.2
All d_eav 6135 -15.4 -7.0 -1.6 4.1 23.8
2009-2023. Changes are 100 times annual log ratios. N counts finite agency-year observations separately for each variable. Percentiles use the untrimmed sample and R quantile type 7.

Figure 1. Observed levy changes and EAV changes

Each point represents an agency-year. Both growth rates are restricted to ±20 (100 times annual log ratios). Red lines fit OLS with an intercept to the displayed observations; brown lines show equal levy and EAV growth. Each panel’s N counts unique agencies (distinct agency_group values) with at least one observation remaining after trimming during 2009–2023. An agency is counted once within each panel regardless of how many years it contributes. Home-rule municipalities are also included in the municipal and all-agency panels, so counts across panels should not be added together.

Figure 1
Six panels plotting levy growth against EAV growth for all agencies, municipalities, home-rule municipalities, other districts, schools, and townships, with OLS and equal-growth reference lines.

Figure 1 — clean presentation

This version shows the same observations and fits, with equal axis scales and the explanatory text outside the image. Teal points represent agency-years, terracotta lines show OLS fits, and dashed gray diagonals mark equal levy and EAV growth. The years and sample restrictions are the same as above.

Figure 2
Clean six-panel comparison of levy and EAV growth, with teal observations, terracotta fitted lines, and dashed gray equal-growth diagonals.

Table 2. All-agency OLS models

Code 
include_table("v22_table_02_OLS_all_agencies.html")
Table 2
tinytable_r67wcz2x2v8v8ndnfwa2
Table 2: All agencies OLS Predict levy using d_eav
M1 M2 M3
* p < 0.1, ** p < 0.05, *** p < 0.01
Cook County, Illinois data from 2008 through 2023.
Columns 2 and 3 include year-by-home-rule fixed effects; column 3 includes unit fixed effects.
Standard errors are clustered by taxing agency. upper_bound is estimate + 1.645 * clustered SE.
Change in EAV 0.071*** 0.090*** 0.074***
(0.010) (0.025) (0.022)
N 6132 6132 6130
upper_bound 0.088 0.131 0.110

Table 2 upper-bound summary

Code 
include_table("v22_table_02_OLS_all_agencies_upper_bounds.html")
Table 3: Table 2: Upper bounds for all-agency OLS
model N estimate std_error p_value upper_bound
M1 6132 0.071*** 0.010 0.000 0.088
M2 6132 0.090*** 0.025 0.000 0.131
M3 6130 0.074*** 0.022 0.001 0.110
Where shown: * p < 0.10; ** p < 0.05; *** p < 0.01. Stars refer to the underlying estimate, not the upper bound. Upper-bound cells are 95% one-sided upper confidence bounds.

Table 3. All-agency IV models

Code 
include_table("v22_table_03_IV_all_agencies.html")
Table 4
tinytable_45ipqoa2f7wku7f3znb5
Table 3: All agencies IV Predict levy using d_eav
M1 M2 M3
* p < 0.1, ** p < 0.05, *** p < 0.01
Cook County, Illinois data from 2008 through 2023.
Columns 2 and 3 include year-by-home-rule fixed effects; column 3 includes unit fixed effects.
d_eav is treated as endogenous and instrumented by reassessment year.
Standard errors are clustered by taxing agency. p_exog is a cluster-robust control-function test of the null that d_eav is exogenous. regF is the first-stage F statistic.
Change in EAV −0.011 −0.091 −0.092
(0.031) (0.063) (0.063)
N 6132 6132 6130
upper_bound 0.040 0.012 0.011
p_exog 0.010 0.010 0.014
regF 3690.6 2381.2 2368.1

Table 3 upper-bound and diagnostic summary

The coefficient test asks whether the IV estimate differs from zero. p_exog asks whether OLS treatment of EAV growth as exogenous appears defensible. The first-stage statistic asks whether reassessment year predicts EAV growth strongly enough to identify the IV coefficient.

Failure to reject exogeneity also does not prove exogeneity, especially when the instrument is weak

Code 
include_table("v22_table_03_IV_all_agencies_upper_bounds.html")
Table 5: Table 3: Upper bounds for all-agency IV
model N estimate std_error p_value upper_bound first_stage_F p_exog
M1 6132 -0.011 0.031 0.724 0.040 3690.6 0.010
M2 6132 -0.091 0.063 0.146 0.012 2381.2 0.010
M3 6130 -0.092 0.063 0.144 0.011 2368.1 0.014

Table 4. OLS models by agency type

Comparing stars across columns does not show if agency types differ. It does not imply that the school and township coefficients differ significantly. A formal comparison requires a pooled interaction model or an appropriate cross-group Wald test. Text shouldn’t say significant for schools and not significant for townships, for example.

Wide upper-bound table

Code 
include_table("v22_table_04_OLS_by_type_upper_bounds.html")
Table 6: Table 4: By agency type OLS upper-bound estimates
row Home-rule municipalities Non-home-rule municipalities Other Townships Schools
No controls 0.058 0.075 0.123 0.054 0.126
Year FE 0.163 0.045 0.180 0.063 0.187
Year + unit FE 0.149 0.027 0.145 0.064 0.167
IG revenue + year + unit FE 0.424 0.353 NA NA 0.265
Same sample as IG row, no IG control 0.422 0.326 NA NA 0.264
Enrollment + year + unit FE NA NA NA NA 0.264

Long upper-bound table

Code 
include_table("v22_table_04_OLS_by_type_upper_bounds_long.html")
Table 7: Table 4: By agency type OLS upper-bound estimates
row agency N estimate std_error p_value upper_bound
No controls Home-rule municipalities 1001 0.015 0.026 0.578 0.058
No controls Non-home-rule municipalities 571 0.039* 0.022 0.087 0.075
No controls Other 2145 0.093*** 0.018 0.000 0.123
No controls Townships 420 0.009 0.027 0.737 0.054
No controls Schools 1995 0.096*** 0.018 0.000 0.126
Year FE Home-rule municipalities 1001 0.077 0.052 0.148 0.163
Year FE Non-home-rule municipalities 571 -0.034 0.048 0.484 0.045
Year FE Other 2145 0.109** 0.044 0.014 0.180
Year FE Townships 420 0.018 0.027 0.502 0.063
Year FE Schools 1995 0.114** 0.044 0.010 0.187
Year + unit FE Home-rule municipalities 1001 0.071 0.048 0.140 0.149
Year + unit FE Non-home-rule municipalities 571 -0.051 0.047 0.286 0.027
Year + unit FE Other 2143 0.085** 0.036 0.021 0.145
Year + unit FE Townships 420 0.018 0.028 0.509 0.064
Year + unit FE Schools 1995 0.102** 0.040 0.011 0.167
IG revenue + year + unit FE Home-rule municipalities 479 0.229* 0.119 0.059 0.424
IG revenue + year + unit FE Non-home-rule municipalities 98 0.127 0.137 0.365 0.353
IG revenue + year + unit FE Schools 1722 0.157** 0.066 0.019 0.265
Same sample as IG row, no IG control Home-rule municipalities 479 0.229* 0.118 0.057 0.422
Same sample as IG row, no IG control Non-home-rule municipalities 98 0.117 0.127 0.371 0.326
Same sample as IG row, no IG control Schools 1722 0.156** 0.066 0.019 0.264
Enrollment + year + unit FE Schools 1722 0.155** 0.066 0.021 0.264
Where shown: * p < 0.10; ** p < 0.05; *** p < 0.01. Stars refer to the underlying estimate, not the upper bound. Upper-bound cells are 95% one-sided upper confidence bounds.

Table 5. IV models by agency type

Shows whether the IV coefficient differs from zero separately by agency type.

The table should report, for every model:

  • a cluster-robust first-stage Wald/Kleibergen–Paap statistic;

  • first-stage partial R2, if practical;

  • a weak-IV-robust confidence interval or test when strength is questionable;

  • optionally the exogeneity-test p-value.

Wide upper-bound table

Code 
include_table("v22_table_05_IV_by_type_upper_bounds.html")
Table 8: Table 5: By agency type IV upper-bound estimates
row Home-rule municipalities Non-home-rule municipalities Other Townships Schools
No controls 0.027 0.061 0.080 0.043 0.108
Year FE 0.085 0.167 0.087 0.081 0.065
Year + unit FE 0.085 0.167 0.087 0.081 0.065
IG revenue + year + unit FE 0.136 0.240 NA NA 0.128
Same sample as IG row, no IG control 0.138 0.228 NA NA 0.130
Enrollment + year + unit FE NA NA NA NA 0.121

Long upper-bound table

Code 
include_table("v22_table_05_IV_by_type_upper_bounds_long.html")
Table 9: Table 5: By agency type IV upper-bound estimates
row agency N estimate std_error p_value upper_bound first_stage_F p_exog
No controls Home-rule municipalities 1001 -0.045 0.044 0.311 0.027 530.4 0.289
No controls Non-home-rule municipalities 571 -0.029 0.055 0.597 0.061 604.6 0.311
No controls Other 2145 -0.052 0.080 0.515 0.080 1018.2 0.068
No controls Townships 420 -0.058 0.061 0.350 0.043 687.3 0.239
No controls Schools 1995 0.065** 0.026 0.015 0.108 1307.5 0.330
Year FE Home-rule municipalities 1001 -0.030 0.070 0.668 0.085 366.2 0.283
Year FE Non-home-rule municipalities 571 -0.053 0.134 0.694 0.167 25.8 0.895
Year FE Other 2145 -0.208 0.179 0.248 0.087 698.3 0.088
Year FE Townships 420 -0.064 0.088 0.473 0.081 1202.9 0.277
Year FE Schools 1995 0.011 0.033 0.734 0.065 814.6 0.075
Year + unit FE Home-rule municipalities 1001 -0.031 0.070 0.664 0.085 362.2 0.289
Year + unit FE Non-home-rule municipalities 571 -0.061 0.139 0.661 0.167 23.9 0.947
Year + unit FE Other 2143 -0.208 0.180 0.248 0.087 698.9 0.103
Year + unit FE Townships 420 -0.064 0.088 0.473 0.081 1202.9 0.261
Year + unit FE Schools 1995 0.011 0.033 0.742 0.065 816.4 0.090
IG revenue + year + unit FE Home-rule municipalities 479 -0.514 0.395 0.199 0.136 7.3 0.060
IG revenue + year + unit FE Non-home-rule municipalities 98 -0.108 0.212 0.616 0.240 2.4 0.086
IG revenue + year + unit FE Schools 1722 -0.037 0.100 0.714 0.128 122.0 0.135
Same sample as IG row, no IG control Home-rule municipalities 479 -0.514 0.396 0.200 0.138 7.3 0.062
Same sample as IG row, no IG control Non-home-rule municipalities 98 -0.109 0.205 0.601 0.228 2.3 0.085
Same sample as IG row, no IG control Schools 1722 -0.035 0.101 0.728 0.130 122.4 0.141
Enrollment + year + unit FE Schools 1722 -0.045 0.101 0.659 0.121 123.1 0.122
Where shown: * p < 0.10; ** p < 0.05; *** p < 0.01. Stars refer to the underlying estimate, not the upper bound. Upper-bound cells are 95% one-sided upper confidence bounds.

Table 5C. IV diagnostics by agency type

Code 
include_table(
  "v22_table_05_C_IV_by_type_diagnostics.html"
)
Table 10: Table 5C: First-stage strength and endogeneity diagnostics by agency type
row agency N first_stage_F p_exog
No controls Home-rule municipalities 1001 530.4 0.289
No controls Non-home-rule municipalities 571 604.6 0.311
No controls Other 2145 1018.2 0.068
No controls Townships 420 687.3 0.239
No controls Schools 1995 1307.5 0.330
Year FE Home-rule municipalities 1001 366.2 0.283
Year FE Non-home-rule municipalities 571 25.8 0.895
Year FE Other 2145 698.3 0.088
Year FE Townships 420 1202.9 0.277
Year FE Schools 1995 814.6 0.075
Year + unit FE Home-rule municipalities 1001 362.2 0.289
Year + unit FE Non-home-rule municipalities 571 23.9 0.947
Year + unit FE Other 2143 698.9 0.103
Year + unit FE Townships 420 1202.9 0.261
Year + unit FE Schools 1995 816.4 0.090
IG revenue + year + unit FE Home-rule municipalities 479 7.3 0.060
IG revenue + year + unit FE Non-home-rule municipalities 98 2.4 0.086
IG revenue + year + unit FE Schools 1722 122.0 0.135
Same sample as IG row, no IG control Home-rule municipalities 479 7.3 0.062
Same sample as IG row, no IG control Non-home-rule municipalities 98 2.3 0.085
Same sample as IG row, no IG control Schools 1722 122.4 0.141
Enrollment + year + unit FE Schools 1722 123.1 0.122
first_stage_F is the cluster-robust first-stage Wald statistic. p_exog is the cluster-robust control-function test of the null that d_eav is exogenous.

Table 6. All-agency asymmetric OLS models

Code 
include_table("v22_table_06_OLS_all_agencies_asym.html")
Table 11
tinytable_e2pz2h1qsscltm0k3eqp
Table 6: All agencies OLS allowing for asymmetry
M1 M2 M3
* p < 0.1, ** p < 0.05, *** p < 0.01
Standard errors are clustered by taxing agency.
p_equal tests equality of the positive- and negative-EAV coefficients.
pos_upper_bound and neg_upper_bound equal the corresponding coefficient plus 1.645 times the clustered SE. neg_d_eav is coded as the actual negative value of d_eav when EAV declines.
Positive EAV change 0.068*** 0.092** 0.082**
(0.017) (0.036) (0.035)
Negative EAV change 0.078*** 0.081 0.044
(0.025) (0.086) (0.089)
N 6132 6132 6130
p_equal 0.777 0.920 0.744
pos_upper_bound 0.096 0.152 0.140
neg_upper_bound 0.119 0.223 0.191

Table 6 positive and negative upper-bound summary

Code 
include_table("v22_table_06_OLS_all_agencies_asym_upper_bounds.html")
Table 12: Table 6: Upper bounds for positive and negative EAV changes
model component N estimate std_error p_value upper_bound p_equal
M1 Positive EAV change 6132 0.068*** 0.017 0.000 0.096 0.777
M1 Negative EAV change 6132 0.078*** 0.025 0.002 0.119 0.777
M2 Positive EAV change 6132 0.092** 0.036 0.012 0.152 0.920
M2 Negative EAV change 6132 0.081 0.086 0.350 0.223 0.920
M3 Positive EAV change 6130 0.082** 0.035 0.020 0.140 0.744
M3 Negative EAV change 6130 0.044 0.089 0.618 0.191 0.744
Where shown: * p < 0.10; ** p < 0.05; *** p < 0.01. Stars refer to the underlying estimate, not the upper bound. Upper-bound cells are 95% one-sided upper confidence bounds.

Table 6B. OLS upper bounds for positive and negative EAV changes

Code 
include_table("v22_table_06_B_OLS_all_agencies_asym_pos_neg_upper_bounds.html")
Table 13: Table 6B: OLS upper bounds for positive and negative EAV changes
model component N estimate std_error p_value upper_bound p_equal
M1 Positive EAV change 6132 0.068*** 0.017 0.000 0.096 0.777
M1 Negative EAV change 6132 0.078*** 0.025 0.002 0.119 0.777
M2 Positive EAV change 6132 0.092** 0.036 0.012 0.152 0.920
M2 Negative EAV change 6132 0.081 0.086 0.350 0.223 0.920
M3 Positive EAV change 6130 0.082** 0.035 0.020 0.140 0.744
M3 Negative EAV change 6130 0.044 0.089 0.618 0.191 0.744
Where shown: * p < 0.10; ** p < 0.05; *** p < 0.01. Stars refer to the underlying estimate, not the upper bound. Upper-bound cells are 95% one-sided upper confidence bounds.

Table 7. All-agency asymmetric IV models

Code 
include_table("v22_table_07_IV_all_agencies_asym.html")
Table 14
tinytable_7j4nskha7eav8onrr7ip
Table 7: All agencies IV allowing for asymmetry
M1 M2 M3
* p < 0.1, ** p < 0.05, *** p < 0.01
Standard errors are clustered by taxing agency.
F_neg and F_pos are equation-specific, cluster-robust first-stage Wald statistics.
SW_F_neg and SW_F_pos are Sanderson-Windmeijer conditional F statistics for the two endogenous variables.
p_exog is a cluster-robust joint control-function test that the positive and negative EAV variables are exogenous.
p_equal tests equality of the positive- and negative-EAV coefficients.
pos_upper_bound and neg_upper_bound equal the corresponding coefficient plus 1.645 times the clustered SE. neg_d_eav is coded as the actual negative value of d_eav when EAV declines.
Positive EAV change 0.004 −0.051 −0.051
(0.027) (0.042) (0.042)
Negative EAV change 0.114*** 0.060 0.061
(0.034) (0.050) (0.050)
N 6132 6132 6130
F_neg 8756.8 1082.0 1068.0
F_pos 4244.6 2646.1 2628.1
SW_F_neg 3909.9 1959.8 1945.7
SW_F_pos 4633.3 5118.0 5099.5
p_exog 0.156 0.009 0.012
p_equal 0.014 0.107 0.104
pos_upper_bound 0.048 0.019 0.019
neg_upper_bound 0.170 0.142 0.144

Table 7 positive and negative upper-bound summary

Code 
include_table("v22_table_07_IV_all_agencies_asym_upper_bounds.html")
Table 15: Table 7: Upper bounds for positive and negative EAV changes, IV
model component N estimate std_error p_value upper_bound F_neg F_pos SW_F_neg SW_F_pos p_exog p_equal
M1 Positive EAV change 6132 0.004 0.027 0.883 0.048 8756.8 4244.6 3909.9 4633.3 0.156 0.014
M1 Negative EAV change 6132 0.114*** 0.034 0.001 0.170 8756.8 4244.6 3909.9 4633.3 0.156 0.014
M2 Positive EAV change 6132 -0.051 0.042 0.233 0.019 1082.0 2646.1 1959.8 5118.0 0.009 0.107
M2 Negative EAV change 6132 0.060 0.050 0.227 0.142 1082.0 2646.1 1959.8 5118.0 0.009 0.107
M3 Positive EAV change 6130 -0.051 0.042 0.234 0.019 1068.0 2628.1 1945.7 5099.5 0.012 0.104
M3 Negative EAV change 6130 0.061 0.050 0.220 0.144 1068.0 2628.1 1945.7 5099.5 0.012 0.104
Where shown: * p < 0.10; ** p < 0.05; *** p < 0.01. Stars refer to the underlying estimate, not the upper bound. Upper-bound cells are 95% one-sided upper confidence bounds.

Table 7B. IV upper bounds for positive and negative EAV changes

Code 
include_table("v22_table_07_B_IV_all_agencies_asym_pos_neg_upper_bounds.html")
Table 16: Table 7B: IV upper bounds for positive and negative EAV changes
model component N estimate std_error p_value upper_bound F_neg F_pos SW_F_neg SW_F_pos p_exog p_equal
M1 Positive EAV change 6132 0.004 0.027 0.883 0.048 8756.8 4244.6 3909.9 4633.3 0.156 0.014
M1 Negative EAV change 6132 0.114*** 0.034 0.001 0.170 8756.8 4244.6 3909.9 4633.3 0.156 0.014
M2 Positive EAV change 6132 -0.051 0.042 0.233 0.019 1082.0 2646.1 1959.8 5118.0 0.009 0.107
M2 Negative EAV change 6132 0.060 0.050 0.227 0.142 1082.0 2646.1 1959.8 5118.0 0.009 0.107
M3 Positive EAV change 6130 -0.051 0.042 0.234 0.019 1068.0 2628.1 1945.7 5099.5 0.012 0.104
M3 Negative EAV change 6130 0.061 0.050 0.220 0.144 1068.0 2628.1 1945.7 5099.5 0.012 0.104
Where shown: * p < 0.10; ** p < 0.05; *** p < 0.01. Stars refer to the underlying estimate, not the upper bound. Upper-bound cells are 95% one-sided upper confidence bounds.

Table 7C. First-stage diagnostics

Code 
include_table("v22_table_07_C_IV_all_agencies_asym_first_stage.html")
Table 17: Table 7C: First-stage and conditional weak-identification diagnostics for asymmetric IV models
model N F_neg F_pos SW_F_neg SW_F_pos p_exog
M1 6132 8756.8 4244.6 3909.9 4633.3 0.156
M2 6132 1082.0 2646.1 1959.8 5118.0 0.009
M3 6130 1068.0 2628.1 1945.7 5099.5 0.012

Table 7D. Alternative negative generated instrument

Code 
include_table("v22_table_07_D_IV_all_agencies_asym_alt_neg_instrument_pos_neg_upper_bounds.html")
Table 18: Table 7D: Alternative asymmetric IV upper bounds using neg_d_eav_hat
model component N estimate std_error p_value upper_bound F_neg F_pos SW_F_neg SW_F_pos p_exog p_equal
M1 Positive EAV change 6132 0.007 0.023 0.764 0.045 2640.9 2185.9 5915.4 5239.9 0.085 0.037
M1 Negative EAV change 6132 0.140*** 0.044 0.002 0.213 2640.9 2185.9 5915.4 5239.9 0.085 0.037
M2 Positive EAV change 6132 -0.051 0.042 0.233 0.019 1082.0 2646.1 1959.8 5118.0 0.009 0.107
M2 Negative EAV change 6132 0.060 0.050 0.227 0.142 1082.0 2646.1 1959.8 5118.0 0.009 0.107
M3 Positive EAV change 6130 -0.051 0.042 0.234 0.019 1068.0 2628.1 1945.7 5099.5 0.012 0.104
M3 Negative EAV change 6130 0.061 0.050 0.220 0.144 1068.0 2628.1 1945.7 5099.5 0.012 0.104
Where shown: * p < 0.10; ** p < 0.05; *** p < 0.01. Stars refer to the underlying estimate, not the upper bound. Upper-bound cells are 95% one-sided upper confidence bounds.

Table 8. Asymmetric OLS models by agency type

Wide upper-bound table

Code 
include_table("v22_table_08_OLS_by_type_asym_upper_bounds.html")
Table 19: Table 8: OLS upper-bound estimates allowing for asymmetry
row Home-rule municipalities Non-home-rule municipalities Other Townships Schools
No controls 0.080 0.030 0.135 0.053 0.165
Year FE 0.188 0.116 0.238 0.038 0.191
Year + unit FE 0.168 0.105 0.223 0.037 0.177
IG revenue + year + unit FE 0.576 0.510 NA NA 0.331
Same sample as IG row, no IG control 0.574 0.490 NA NA 0.330
Enrollment + year + unit FE NA NA NA NA 0.330

Long upper-bound table

Code 
include_table("v22_table_08_OLS_by_type_asym_upper_bounds_long.html")
Table 20: Table 8: OLS upper-bound estimates allowing for asymmetry
row agency N estimate std_error p_value upper_bound p_equal
No controls Home-rule municipalities 1001 0.004 0.046 0.932 0.080 0.649
No controls Non-home-rule municipalities 571 -0.007 0.022 0.763 0.030 0.201
No controls Other 2145 0.087*** 0.030 0.004 0.135 0.813
No controls Townships 420 0.014 0.024 0.573 0.053 0.815
No controls Schools 1995 0.111*** 0.033 0.001 0.165 0.392
Year FE Home-rule municipalities 1001 0.083 0.064 0.199 0.188 0.713
Year FE Non-home-rule municipalities 571 0.027 0.054 0.625 0.116 0.170
Year FE Other 2145 0.123* 0.070 0.083 0.238 0.803
Year FE Townships 420 -0.022 0.037 0.555 0.038 0.037
Year FE Schools 1995 0.102* 0.054 0.059 0.191 0.334
Year + unit FE Home-rule municipalities 1001 0.070 0.060 0.249 0.168 0.937
Year + unit FE Non-home-rule municipalities 571 0.016 0.054 0.765 0.105 0.132
Year + unit FE Other 2143 0.110 0.069 0.113 0.223 0.667
Year + unit FE Townships 420 -0.025 0.037 0.513 0.037 0.032
Year + unit FE Schools 1995 0.096* 0.049 0.053 0.177 0.632
IG revenue + year + unit FE Home-rule municipalities 479 0.293* 0.172 0.095 0.576 0.320
IG revenue + year + unit FE Non-home-rule municipalities 98 0.195 0.192 0.322 0.510 0.581
IG revenue + year + unit FE Schools 1722 0.175* 0.095 0.067 0.331 0.594
Same sample as IG row, no IG control Home-rule municipalities 479 0.293* 0.171 0.093 0.574 0.319
Same sample as IG row, no IG control Non-home-rule municipalities 98 0.182 0.188 0.345 0.490 0.587
Same sample as IG row, no IG control Schools 1722 0.175* 0.095 0.067 0.330 0.585
Enrollment + year + unit FE Schools 1722 0.174* 0.095 0.070 0.330 0.580
Where shown: * p < 0.10; ** p < 0.05; *** p < 0.01. Stars refer to the underlying estimate, not the upper bound. Upper-bound cells are 95% one-sided upper confidence bounds.

Table 8B wide positive and negative upper-bound table

Code 
include_table("v22_table_08_B_OLS_by_type_asym_pos_neg_upper_bounds.html")
Table 21: Table 8B: OLS upper bounds for positive and negative EAV changes by agency type
row Home-rule municipalities Non-home-rule municipalities Other Townships Schools
No controls - Positive EAV change 0.080 0.030 0.135 0.053 0.165
No controls - Negative EAV change 0.090 0.267 0.196 0.091 0.105
Year FE - Positive EAV change 0.188 0.116 0.238 0.038 0.191
Year FE - Negative EAV change 0.152 0.042 0.378 0.365 0.236
Year + unit FE - Positive EAV change 0.168 0.105 0.223 0.037 0.177
Year + unit FE - Negative EAV change 0.182 0.005 0.328 0.386 0.204
IG revenue + year + unit FE - Positive EAV change 0.576 0.510 NA NA 0.331
IG revenue + year + unit FE - Negative EAV change 0.198 0.342 NA NA 0.189
Same sample as IG row, no IG control - Positive EAV change 0.574 0.490 NA NA 0.330
Same sample as IG row, no IG control - Negative EAV change 0.198 0.312 NA NA 0.187
Enrollment + year + unit FE - Positive EAV change NA NA NA NA 0.330
Enrollment + year + unit FE - Negative EAV change NA NA NA NA 0.184

Table 8B long positive and negative upper-bound table

Code 
include_table("v22_table_08_B_OLS_by_type_asym_pos_neg_upper_bounds_long.html")
Table 22: Table 8B: OLS upper bounds for positive and negative EAV changes by agency type
row agency component N estimate std_error p_value upper_bound p_equal
No controls Home-rule municipalities Positive EAV change 1001 0.004 0.046 0.932 0.080 0.649
No controls Home-rule municipalities Negative EAV change 1001 0.036 0.033 0.289 0.090 0.649
No controls Non-home-rule municipalities Positive EAV change 571 -0.007 0.022 0.763 0.030 0.201
No controls Non-home-rule municipalities Negative EAV change 571 0.126 0.086 0.149 0.267 0.201
No controls Other Positive EAV change 2145 0.087*** 0.030 0.004 0.135 0.813
No controls Other Negative EAV change 2145 0.105* 0.055 0.061 0.196 0.813
No controls Townships Positive EAV change 420 0.014 0.024 0.573 0.053 0.815
No controls Townships Negative EAV change 420 0.001 0.055 0.990 0.091 0.815
No controls Schools Positive EAV change 1995 0.111*** 0.033 0.001 0.165 0.392
No controls Schools Negative EAV change 1995 0.067*** 0.023 0.004 0.105 0.392
Year FE Home-rule municipalities Positive EAV change 1001 0.083 0.064 0.199 0.188 0.713
Year FE Home-rule municipalities Negative EAV change 1001 0.049 0.063 0.438 0.152 0.713
Year FE Non-home-rule municipalities Positive EAV change 571 0.027 0.054 0.625 0.116 0.170
Year FE Non-home-rule municipalities Negative EAV change 571 -0.219 0.159 0.175 0.042 0.170
Year FE Other Positive EAV change 2145 0.123* 0.070 0.083 0.238 0.803
Year FE Other Negative EAV change 2145 0.063 0.192 0.745 0.378 0.803
Year FE Townships Positive EAV change 420 -0.022 0.037 0.555 0.038 0.037
Year FE Townships Negative EAV change 420 0.218** 0.089 0.021 0.365 0.037
Year FE Schools Positive EAV change 1995 0.102* 0.054 0.059 0.191 0.334
Year FE Schools Negative EAV change 1995 0.169*** 0.041 0.000 0.236 0.334
Year + unit FE Home-rule municipalities Positive EAV change 1001 0.070 0.060 0.249 0.168 0.937
Year + unit FE Home-rule municipalities Negative EAV change 1001 0.077 0.064 0.229 0.182 0.937
Year + unit FE Non-home-rule municipalities Positive EAV change 571 0.016 0.054 0.765 0.105 0.132
Year + unit FE Non-home-rule municipalities Negative EAV change 571 -0.264 0.164 0.115 0.005 0.132
Year + unit FE Other Positive EAV change 2143 0.110 0.069 0.113 0.223 0.667
Year + unit FE Other Negative EAV change 2143 0.001 0.199 0.998 0.328 0.667
Year + unit FE Townships Positive EAV change 420 -0.025 0.037 0.513 0.037 0.032
Year + unit FE Townships Negative EAV change 420 0.233** 0.093 0.018 0.386 0.032
Year + unit FE Schools Positive EAV change 1995 0.096* 0.049 0.053 0.177 0.632
Year + unit FE Schools Negative EAV change 1995 0.131*** 0.044 0.004 0.204 0.632
IG revenue + year + unit FE Home-rule municipalities Positive EAV change 479 0.293* 0.172 0.095 0.576 0.320
IG revenue + year + unit FE Home-rule municipalities Negative EAV change 479 0.094 0.063 0.143 0.198 0.320
IG revenue + year + unit FE Non-home-rule municipalities Positive EAV change 98 0.195 0.192 0.322 0.510 0.581
IG revenue + year + unit FE Non-home-rule municipalities Negative EAV change 98 0.031 0.189 0.871 0.342 0.581
IG revenue + year + unit FE Schools Positive EAV change 1722 0.175* 0.095 0.067 0.331 0.594
IG revenue + year + unit FE Schools Negative EAV change 1722 0.114** 0.046 0.014 0.189 0.594
Same sample as IG row, no IG control Home-rule municipalities Positive EAV change 479 0.293* 0.171 0.093 0.574 0.319
Same sample as IG row, no IG control Home-rule municipalities Negative EAV change 479 0.093 0.063 0.147 0.198 0.319
Same sample as IG row, no IG control Non-home-rule municipalities Positive EAV change 98 0.182 0.188 0.345 0.490 0.587
Same sample as IG row, no IG control Non-home-rule municipalities Negative EAV change 98 0.023 0.176 0.899 0.312 0.587
Same sample as IG row, no IG control Schools Positive EAV change 1722 0.175* 0.095 0.067 0.330 0.585
Same sample as IG row, no IG control Schools Negative EAV change 1722 0.112** 0.045 0.015 0.187 0.585
Enrollment + year + unit FE Schools Positive EAV change 1722 0.174* 0.095 0.070 0.330 0.580
Enrollment + year + unit FE Schools Negative EAV change 1722 0.111** 0.045 0.015 0.184 0.580
Where shown: * p < 0.10; ** p < 0.05; *** p < 0.01. Stars refer to the underlying estimate, not the upper bound. Upper-bound cells are 95% one-sided upper confidence bounds.

Table 9. Asymmetric IV models by agency type

Wide upper-bound table

Code 
include_table("v22_table_09_IV_by_type_asym_upper_bounds.html")
Table 23: Table 9: IV upper-bound estimates allowing for asymmetry
row Home-rule municipalities Non-home-rule municipalities Other Townships Schools
No controls 0.002 0.056 0.089 0.041 0.105
Year FE 0.040 -0.027 0.091 0.072 0.041
Year + unit FE 0.041 -0.028 0.092 0.071 0.040
IG revenue + year + unit FE 0.057 0.490 NA NA 0.043
Same sample as IG row, no IG control 0.058 0.442 NA NA 0.042
Enrollment + year + unit FE NA NA NA NA 0.037

Long upper-bound table

Code 
include_table("v22_table_09_IV_by_type_asym_upper_bounds_long.html")
Table 24: Table 9: IV upper-bound estimates allowing for asymmetry
row agency N estimate std_error p_value upper_bound p_equal
No controls Home-rule municipalities 1001 -0.047 0.030 0.123 0.002 0.904
No controls Non-home-rule municipalities 571 -0.003 0.036 0.944 0.056 0.236
No controls Other 2145 -0.029 0.072 0.687 0.089 0.050
No controls Townships 420 -0.031 0.044 0.484 0.041 0.160
No controls Schools 1995 0.072*** 0.020 0.000 0.105 0.316
Year FE Home-rule municipalities 1001 -0.013 0.032 0.688 0.040 0.707
Year FE Non-home-rule municipalities 571 -0.085** 0.035 0.021 -0.027 0.786
Year FE Other 2145 -0.113 0.124 0.364 0.091 0.141
Year FE Townships 420 -0.021 0.056 0.709 0.072 0.230
Year FE Schools 1995 0.010 0.019 0.615 0.041 0.936
Year + unit FE Home-rule municipalities 1001 -0.012 0.032 0.710 0.041 0.687
Year + unit FE Non-home-rule municipalities 571 -0.086** 0.036 0.020 -0.028 0.834
Year + unit FE Other 2143 -0.113 0.124 0.366 0.092 0.140
Year + unit FE Townships 420 -0.021 0.056 0.710 0.071 0.233
Year + unit FE Schools 1995 0.009 0.019 0.640 0.040 0.924
IG revenue + year + unit FE Home-rule municipalities 479 -0.064 0.074 0.387 0.057 0.096
IG revenue + year + unit FE Non-home-rule municipalities 98 0.299** 0.116 0.019 0.490 0.042
IG revenue + year + unit FE Schools 1722 -0.012 0.033 0.723 0.043 0.745
Same sample as IG row, no IG control Home-rule municipalities 479 -0.065 0.075 0.387 0.058 0.098
Same sample as IG row, no IG control Non-home-rule municipalities 98 0.270** 0.105 0.018 0.442 0.051
Same sample as IG row, no IG control Schools 1722 -0.012 0.033 0.713 0.042 0.769
Enrollment + year + unit FE Schools 1722 -0.016 0.032 0.614 0.037 0.717
Where shown: * p < 0.10; ** p < 0.05; *** p < 0.01. Stars refer to the underlying estimate, not the upper bound. Upper-bound cells are 95% one-sided upper confidence bounds.

Table 9B wide positive and negative upper-bound table

Code 
include_table("v22_table_09_B_IV_by_type_asym_pos_neg_upper_bounds.html")
Table 25: Table 9B: IV upper bounds for positive and negative EAV changes by agency type
row Home-rule municipalities Non-home-rule municipalities Other Townships Schools
No controls - Positive EAV change 0.002 0.056 0.089 0.041 0.105
No controls - Negative EAV change 0.079 0.325 0.276 0.344 0.194
Year FE - Positive EAV change 0.040 -0.027 0.091 0.072 0.041
Year FE - Negative EAV change 0.209 0.178 0.332 0.218 0.088
Year + unit FE - Positive EAV change 0.041 -0.028 0.092 0.071 0.040
Year + unit FE - Negative EAV change 0.215 0.197 0.334 0.220 0.086
IG revenue + year + unit FE - Positive EAV change 0.057 0.490 NA NA 0.043
IG revenue + year + unit FE - Negative EAV change 0.274 0.147 NA NA 0.092
Same sample as IG row, no IG control - Positive EAV change 0.058 0.442 NA NA 0.042
Same sample as IG row, no IG control - Negative EAV change 0.273 0.131 NA NA 0.090
Enrollment + year + unit FE - Positive EAV change NA NA NA NA 0.037
Enrollment + year + unit FE - Negative EAV change NA NA NA NA 0.090

Table 9B long positive and negative upper-bound table

Code 
include_table("v22_table_09_B_IV_by_type_asym_pos_neg_upper_bounds_long.html")
Table 26: Table 9B: IV upper bounds for positive and negative EAV changes by agency type
row agency component N estimate std_error p_value upper_bound F_neg F_pos SW_F_neg SW_F_pos p_exog p_equal
No controls Home-rule municipalities Positive EAV change 1001 -0.047 0.030 0.123 0.002 1581.1 558.4 566.5 667.8 0.010 0.904
No controls Home-rule municipalities Negative EAV change 1001 -0.060 0.084 0.480 0.079 1581.1 558.4 566.5 667.8 0.010 0.904
No controls Non-home-rule municipalities Positive EAV change 571 -0.003 0.036 0.944 0.056 823.7 1008.6 625.4 1150.3 0.760 0.236
No controls Non-home-rule municipalities Negative EAV change 571 0.156 0.103 0.137 0.325 823.7 1008.6 625.4 1150.3 0.760 0.236
No controls Other Positive EAV change 2145 -0.029 0.072 0.687 0.089 2998.9 1073.9 1130.5 1282.1 0.377 0.050
No controls Other Negative EAV change 2145 0.159** 0.071 0.027 0.276 2998.9 1073.9 1130.5 1282.1 0.377 0.050
No controls Townships Positive EAV change 420 -0.031 0.044 0.484 0.041 979.0 1099.4 582.6 794.5 0.172 0.160
No controls Townships Negative EAV change 420 0.168 0.107 0.126 0.344 979.0 1099.4 582.6 794.5 0.172 0.160
No controls Schools Positive EAV change 1995 0.072*** 0.020 0.000 0.105 2869.6 1631.7 1425.6 1536.0 0.339 0.316
No controls Schools Negative EAV change 1995 0.128*** 0.040 0.002 0.194 2869.6 1631.7 1425.6 1536.0 0.339 0.316
Year FE Home-rule municipalities Positive EAV change 1001 -0.013 0.032 0.688 0.040 202.1 404.2 325.9 809.3 0.398 0.707
Year FE Home-rule municipalities Negative EAV change 1001 0.034 0.107 0.754 0.209 202.1 404.2 325.9 809.3 0.398 0.707
Year FE Non-home-rule municipalities Positive EAV change 571 -0.085** 0.035 0.021 -0.027 125.6 1162.2 211.8 1768.2 0.154 0.786
Year FE Non-home-rule municipalities Negative EAV change 571 -0.143 0.195 0.470 0.178 125.6 1162.2 211.8 1768.2 0.154 0.786
Year FE Other Positive EAV change 2145 -0.113 0.124 0.364 0.091 373.0 918.5 657.2 1676.2 0.289 0.141
Year FE Other Negative EAV change 2145 0.148 0.112 0.189 0.332 373.0 918.5 657.2 1676.2 0.289 0.141
Year FE Townships Positive EAV change 420 -0.021 0.056 0.709 0.072 113.4 649.4 231.4 1301.0 0.237 0.230
Year FE Townships Negative EAV change 420 0.106 0.068 0.130 0.218 113.4 649.4 231.4 1301.0 0.237 0.230
Year FE Schools Positive EAV change 1995 0.010 0.019 0.615 0.041 346.1 725.5 599.2 1447.1 0.001 0.936
Year FE Schools Negative EAV change 1995 0.005 0.050 0.920 0.088 346.1 725.5 599.2 1447.1 0.001 0.936
Year + unit FE Home-rule municipalities Positive EAV change 1001 -0.012 0.032 0.710 0.041 199.9 400.6 322.9 806.3 0.410 0.687
Year + unit FE Home-rule municipalities Negative EAV change 1001 0.038 0.108 0.724 0.215 199.9 400.6 322.9 806.3 0.410 0.687
Year + unit FE Non-home-rule municipalities Positive EAV change 571 -0.086** 0.036 0.020 -0.028 119.3 1176.6 188.7 1765.3 0.238 0.834
Year + unit FE Non-home-rule municipalities Negative EAV change 571 -0.132 0.200 0.513 0.197 119.3 1176.6 188.7 1765.3 0.238 0.834
Year + unit FE Other Positive EAV change 2143 -0.113 0.124 0.366 0.092 370.6 910.9 657.5 1668.5 0.356 0.140
Year + unit FE Other Negative EAV change 2143 0.149 0.112 0.185 0.334 370.6 910.9 657.5 1668.5 0.356 0.140
Year + unit FE Townships Positive EAV change 420 -0.021 0.056 0.710 0.071 112.3 650.6 231.2 1311.5 0.179 0.233
Year + unit FE Townships Negative EAV change 420 0.107 0.069 0.134 0.220 112.3 650.6 231.2 1311.5 0.179 0.233
Year + unit FE Schools Positive EAV change 1995 0.009 0.019 0.640 0.040 342.7 722.3 593.0 1444.6 0.005 0.924
Year + unit FE Schools Negative EAV change 1995 0.004 0.050 0.942 0.086 342.7 722.3 593.0 1444.6 0.005 0.924
IG revenue + year + unit FE Home-rule municipalities Positive EAV change 479 -0.064 0.074 0.387 0.057 92.4 116.8 223.0 229.3 0.063 0.096
IG revenue + year + unit FE Home-rule municipalities Negative EAV change 479 0.144* 0.079 0.076 0.274 92.4 116.8 223.0 229.3 0.063 0.096
IG revenue + year + unit FE Non-home-rule municipalities Positive EAV change 98 0.299** 0.116 0.019 0.490 121.3 15.2 292.0 36.3 0.876 0.042
IG revenue + year + unit FE Non-home-rule municipalities Negative EAV change 98 -0.039 0.113 0.735 0.147 121.3 15.2 292.0 36.3 0.876 0.042
IG revenue + year + unit FE Schools Positive EAV change 1722 -0.012 0.033 0.723 0.043 287.7 367.8 604.1 769.4 0.004 0.745
IG revenue + year + unit FE Schools Negative EAV change 1722 0.009 0.050 0.866 0.092 287.7 367.8 604.1 769.4 0.004 0.745
Same sample as IG row, no IG control Home-rule municipalities Positive EAV change 479 -0.065 0.075 0.387 0.058 93.2 116.7 223.5 228.6 0.064 0.098
Same sample as IG row, no IG control Home-rule municipalities Negative EAV change 479 0.142* 0.080 0.080 0.273 93.2 116.7 223.5 228.6 0.064 0.098
Same sample as IG row, no IG control Non-home-rule municipalities Positive EAV change 98 0.270** 0.105 0.018 0.442 150.3 16.4 311.6 38.2 0.884 0.051
Same sample as IG row, no IG control Non-home-rule municipalities Negative EAV change 98 -0.047 0.108 0.668 0.131 150.3 16.4 311.6 38.2 0.884 0.051
Same sample as IG row, no IG control Schools Positive EAV change 1722 -0.012 0.033 0.713 0.042 288.8 366.9 606.0 772.5 0.004 0.769
Same sample as IG row, no IG control Schools Negative EAV change 1722 0.006 0.051 0.905 0.090 288.8 366.9 606.0 772.5 0.004 0.769
Enrollment + year + unit FE Schools Positive EAV change 1722 -0.016 0.032 0.614 0.037 287.5 368.6 609.4 774.4 0.003 0.717
Enrollment + year + unit FE Schools Negative EAV change 1722 0.006 0.051 0.903 0.090 287.5 368.6 609.4 774.4 0.003 0.717
Where shown: * p < 0.10; ** p < 0.05; *** p < 0.01. Stars refer to the underlying estimate, not the upper bound. Upper-bound cells are 95% one-sided upper confidence bounds.

Table 9C. First-stage diagnostics by agency type

Code 
include_table("v22_table_09_C_IV_by_type_asym_first_stage_long.html")
Table 27: Table 9C: First-stage and conditional weak-identification diagnostics by agency type
row agency N F_neg F_pos SW_F_neg SW_F_pos p_exog
No controls Home-rule municipalities 1001 1581.1 558.4 566.5 667.8 0.010
No controls Non-home-rule municipalities 571 823.7 1008.6 625.4 1150.3 0.760
No controls Other 2145 2998.9 1073.9 1130.5 1282.1 0.377
No controls Townships 420 979.0 1099.4 582.6 794.5 0.172
No controls Schools 1995 2869.6 1631.7 1425.6 1536.0 0.339
Year FE Home-rule municipalities 1001 202.1 404.2 325.9 809.3 0.398
Year FE Non-home-rule municipalities 571 125.6 1162.2 211.8 1768.2 0.154
Year FE Other 2145 373.0 918.5 657.2 1676.2 0.289
Year FE Townships 420 113.4 649.4 231.4 1301.0 0.237
Year FE Schools 1995 346.1 725.5 599.2 1447.1 0.001
Year + unit FE Home-rule municipalities 1001 199.9 400.6 322.9 806.3 0.410
Year + unit FE Non-home-rule municipalities 571 119.3 1176.6 188.7 1765.3 0.238
Year + unit FE Other 2143 370.6 910.9 657.5 1668.5 0.356
Year + unit FE Townships 420 112.3 650.6 231.2 1311.5 0.179
Year + unit FE Schools 1995 342.7 722.3 593.0 1444.6 0.005
IG revenue + year + unit FE Home-rule municipalities 479 92.4 116.8 223.0 229.3 0.063
IG revenue + year + unit FE Non-home-rule municipalities 98 121.3 15.2 292.0 36.3 0.876
IG revenue + year + unit FE Schools 1722 287.7 367.8 604.1 769.4 0.004
Same sample as IG row, no IG control Home-rule municipalities 479 93.2 116.7 223.5 228.6 0.064
Same sample as IG row, no IG control Non-home-rule municipalities 98 150.3 16.4 311.6 38.2 0.884
Same sample as IG row, no IG control Schools 1722 288.8 366.9 606.0 772.5 0.004
Enrollment + year + unit FE Schools 1722 287.5 368.6 609.4 774.4 0.003

Table 9D. Alternative negative generated instrument by agency type

Code 
include_table("v22_table_09_D_IV_by_type_asym_alt_neg_instrument_long.html")
Table 28: Table 9D: Alternative asymmetric IV by agency type using neg_d_eav_hat
row agency component N estimate std_error p_value upper_bound F_neg F_pos SW_F_neg SW_F_pos p_exog p_equal
No controls Home-rule municipalities Positive EAV change 1001 -0.030 0.038 0.426 0.032 420.9 298.4 899.3 680.1 0.643 0.078
No controls Home-rule municipalities Negative EAV change 1001 0.076* 0.040 0.063 0.141 420.9 298.4 899.3 680.1 0.643 0.078
No controls Non-home-rule municipalities Positive EAV change 571 0.011 0.041 0.795 0.078 394.5 466.3 998.3 1255.5 0.109 0.104
No controls Non-home-rule municipalities Negative EAV change 571 0.247* 0.128 0.060 0.457 394.5 466.3 998.3 1255.5 0.109 0.104
No controls Other Positive EAV change 2145 -0.021 0.059 0.723 0.076 719.6 655.6 1574.9 1555.6 0.131 0.150
No controls Other Negative EAV change 2145 0.230* 0.119 0.056 0.426 719.6 655.6 1574.9 1555.6 0.131 0.150
No controls Townships Positive EAV change 420 -0.053 0.054 0.330 0.035 548.1 523.5 1268.5 1375.3 0.119 0.686
No controls Townships Negative EAV change 420 -0.016 0.078 0.839 0.112 548.1 523.5 1268.5 1375.3 0.119 0.686
No controls Schools Positive EAV change 1995 0.066*** 0.022 0.003 0.102 989.0 729.9 2243.2 1714.7 0.268 0.818
No controls Schools Negative EAV change 1995 0.075*** 0.025 0.003 0.116 989.0 729.9 2243.2 1714.7 0.268 0.818
Year FE Home-rule municipalities Positive EAV change 1001 -0.013 0.032 0.688 0.040 202.1 404.2 325.9 809.3 0.398 0.707
Year FE Home-rule municipalities Negative EAV change 1001 0.034 0.107 0.754 0.209 202.1 404.2 325.9 809.3 0.398 0.707
Year FE Non-home-rule municipalities Positive EAV change 571 -0.085** 0.035 0.021 -0.027 125.6 1162.2 211.8 1768.2 0.154 0.786
Year FE Non-home-rule municipalities Negative EAV change 571 -0.143 0.195 0.470 0.178 125.6 1162.2 211.8 1768.2 0.154 0.786
Year FE Other Positive EAV change 2145 -0.113 0.124 0.364 0.091 373.0 918.5 657.2 1676.2 0.289 0.141
Year FE Other Negative EAV change 2145 0.148 0.112 0.189 0.332 373.0 918.5 657.2 1676.2 0.289 0.141
Year FE Townships Positive EAV change 420 -0.021 0.056 0.709 0.072 113.4 649.4 231.4 1301.0 0.237 0.230
Year FE Townships Negative EAV change 420 0.106 0.068 0.130 0.218 113.4 649.4 231.4 1301.0 0.237 0.230
Year FE Schools Positive EAV change 1995 0.010 0.019 0.615 0.041 346.1 725.5 599.2 1447.1 0.001 0.936
Year FE Schools Negative EAV change 1995 0.005 0.050 0.920 0.088 346.1 725.5 599.2 1447.1 0.001 0.936
Year + unit FE Home-rule municipalities Positive EAV change 1001 -0.012 0.032 0.710 0.041 199.9 400.6 322.9 806.3 0.410 0.687
Year + unit FE Home-rule municipalities Negative EAV change 1001 0.038 0.108 0.724 0.215 199.9 400.6 322.9 806.3 0.410 0.687
Year + unit FE Non-home-rule municipalities Positive EAV change 571 -0.086** 0.036 0.020 -0.028 119.3 1176.6 188.7 1765.3 0.238 0.834
Year + unit FE Non-home-rule municipalities Negative EAV change 571 -0.132 0.200 0.513 0.197 119.3 1176.6 188.7 1765.3 0.238 0.834
Year + unit FE Other Positive EAV change 2143 -0.113 0.124 0.366 0.092 370.6 910.9 657.5 1668.5 0.356 0.140
Year + unit FE Other Negative EAV change 2143 0.149 0.112 0.185 0.334 370.6 910.9 657.5 1668.5 0.356 0.140
Year + unit FE Townships Positive EAV change 420 -0.021 0.056 0.710 0.071 112.3 650.6 231.2 1311.5 0.179 0.233
Year + unit FE Townships Negative EAV change 420 0.107 0.069 0.134 0.220 112.3 650.6 231.2 1311.5 0.179 0.233
Year + unit FE Schools Positive EAV change 1995 0.009 0.019 0.640 0.040 342.7 722.3 593.0 1444.6 0.005 0.924
Year + unit FE Schools Negative EAV change 1995 0.004 0.050 0.942 0.086 342.7 722.3 593.0 1444.6 0.005 0.924
IG revenue + year + unit FE Home-rule municipalities Positive EAV change 479 -0.064 0.074 0.387 0.057 92.4 116.8 223.0 229.3 0.063 0.096
IG revenue + year + unit FE Home-rule municipalities Negative EAV change 479 0.144* 0.079 0.076 0.274 92.4 116.8 223.0 229.3 0.063 0.096
IG revenue + year + unit FE Non-home-rule municipalities Positive EAV change 98 0.299** 0.116 0.019 0.490 121.3 15.2 292.0 36.3 0.876 0.042
IG revenue + year + unit FE Non-home-rule municipalities Negative EAV change 98 -0.039 0.113 0.735 0.147 121.3 15.2 292.0 36.3 0.876 0.042
IG revenue + year + unit FE Schools Positive EAV change 1722 -0.012 0.033 0.723 0.043 287.7 367.8 604.1 769.4 0.004 0.745
IG revenue + year + unit FE Schools Negative EAV change 1722 0.009 0.050 0.866 0.092 287.7 367.8 604.1 769.4 0.004 0.745
Same sample as IG row, no IG control Home-rule municipalities Positive EAV change 479 -0.065 0.075 0.387 0.058 93.2 116.7 223.5 228.6 0.064 0.098
Same sample as IG row, no IG control Home-rule municipalities Negative EAV change 479 0.142* 0.080 0.080 0.273 93.2 116.7 223.5 228.6 0.064 0.098
Same sample as IG row, no IG control Non-home-rule municipalities Positive EAV change 98 0.270** 0.105 0.018 0.442 150.3 16.4 311.6 38.2 0.884 0.051
Same sample as IG row, no IG control Non-home-rule municipalities Negative EAV change 98 -0.047 0.108 0.668 0.131 150.3 16.4 311.6 38.2 0.884 0.051
Same sample as IG row, no IG control Schools Positive EAV change 1722 -0.012 0.033 0.713 0.042 288.8 366.9 606.0 772.5 0.004 0.769
Same sample as IG row, no IG control Schools Negative EAV change 1722 0.006 0.051 0.905 0.090 288.8 366.9 606.0 772.5 0.004 0.769
Enrollment + year + unit FE Schools Positive EAV change 1722 -0.016 0.032 0.614 0.037 287.5 368.6 609.4 774.4 0.003 0.717
Enrollment + year + unit FE Schools Negative EAV change 1722 0.006 0.051 0.903 0.090 287.5 368.6 609.4 774.4 0.003 0.717
Where shown: * p < 0.10; ** p < 0.05; *** p < 0.01. Stars refer to the underlying estimate, not the upper bound. Upper-bound cells are 95% one-sided upper confidence bounds.

Table 10. All-agency OLS models with lagged EAV effects

Code 
include_table("v22_table_10_OLS_all_agencies_lagged.html")
Table 29
tinytable_rv8ex5khrxmtsx9emx2s
Table 10: All agencies OLS allowing for two lags of d_eav
M1 M2 M3
* p < 0.1, ** p < 0.05, *** p < 0.01
Standard errors are clustered by taxing agency.
p_lags tests whether the sum of the two lag coefficients equals zero.
joint_lag_F and joint_lag_p test whether both lag coefficients jointly equal zero.
upper_bound is computed for the cumulative sum of current and lagged EAV coefficients.
Current EAV change 0.077*** 0.120*** 0.099***
(0.011) (0.034) (0.033)
Lag 1 EAV change −0.037 0.048 0.024
(0.031) (0.031) (0.034)
Lag 2 EAV change 0.080** 0.135** 0.115
(0.037) (0.067) (0.072)
N 6132 6132 6130
p_lags 0.002 0.031 0.141
joint_lag_F 4.9 2.3 1.3
joint_lag_p 0.007 0.097 0.271
upper_bound 0.152 0.489 0.437

Table 10 cumulative current and lagged effects

Code 
include_table("v22_table_10_OLS_all_agencies_lagged_sums.html")
Table 30: Table 10: Sum of current and lagged EAV effects
model N lagged_magnitude std_error lagged_p joint_lag_F joint_lag_p A_upper_bound
M1 6132 0.120*** 0.019 0.000 4.9 0.007 0.152
M2 6132 0.304*** 0.113 0.007 2.3 0.097 0.489
M3 6130 0.238** 0.121 0.049 1.3 0.271 0.437
Where shown: * p < 0.10; ** p < 0.05; *** p < 0.01. Stars refer to the underlying estimate, not the upper bound. Upper-bound cells are 95% one-sided upper confidence bounds.

Table 11. OLS models with lagged EAV effects by agency type

Wide upper-bound table

Code 
include_table("v22_table_11_OLS_by_type_lagged_upper_bounds.html")
Table 31: Table 11: OLS upper-bound estimates allowing for two lags of d_eav
row Home-rule municipalities Non-home-rule municipalities Other Townships Schools
No controls 0.138 0.136 0.241 0.095 0.149
Year FE 0.413 0.447 0.793 0.197 0.478
Year + unit FE 0.422 0.392 0.739 0.280 0.385
IG revenue + year + unit FE 0.833 0.441 NA NA 0.438
Same sample as IG row, no IG control 0.832 0.380 NA NA 0.436
Enrollment + year + unit FE NA NA NA NA 0.433

Long upper-bound table

Code 
include_table("v22_table_11_OLS_by_type_lagged_upper_bounds_long.html")
Table 32: Table 11: OLS upper-bound estimates allowing for two lags of d_eav
row agency N lagged_magnitude std_error lagged_p joint_lag_F joint_lag_p A_upper_bound
No controls Home-rule municipalities 1001 0.073* 0.039 0.067 3.1 0.044 0.138
No controls Non-home-rule municipalities 571 0.065 0.043 0.146 4.8 0.008 0.136
No controls Other 2145 0.171*** 0.043 0.000 2.1 0.126 0.241
No controls Townships 420 0.054** 0.025 0.040 5.4 0.005 0.095
No controls Schools 1995 0.114*** 0.021 0.000 4.9 0.008 0.149
Year FE Home-rule municipalities 1001 0.110 0.184 0.554 0.1 0.948 0.413
Year FE Non-home-rule municipalities 571 0.233* 0.130 0.080 3.7 0.026 0.447
Year FE Other 2145 0.391 0.244 0.111 0.9 0.408 0.793
Year FE Townships 420 0.036 0.098 0.712 0.7 0.518 0.197
Year FE Schools 1995 0.343*** 0.082 0.000 12.0 0.000 0.478
Year + unit FE Home-rule municipalities 1001 0.146 0.168 0.385 0.3 0.730 0.422
Year + unit FE Non-home-rule municipalities 571 0.161 0.140 0.259 2.8 0.061 0.392
Year + unit FE Other 2143 0.300 0.267 0.262 0.8 0.447 0.739
Year + unit FE Townships 420 0.070 0.127 0.585 0.7 0.489 0.280
Year + unit FE Schools 1995 0.257*** 0.078 0.001 5.4 0.005 0.385
IG revenue + year + unit FE Home-rule municipalities 479 0.361 0.287 0.214 0.3 0.765 0.833
IG revenue + year + unit FE Non-home-rule municipalities 98 0.098 0.208 0.643 0.0 0.962 0.441
IG revenue + year + unit FE Schools 1722 0.280*** 0.096 0.004 3.8 0.023 0.438
Same sample as IG row, no IG control Home-rule municipalities 479 0.360 0.287 0.214 0.3 0.759 0.832
Same sample as IG row, no IG control Non-home-rule municipalities 98 0.075 0.185 0.688 0.1 0.939 0.380
Same sample as IG row, no IG control Schools 1722 0.278*** 0.096 0.005 3.7 0.025 0.436
Enrollment + year + unit FE Schools 1722 0.278*** 0.095 0.004 4.0 0.019 0.433
Where shown: * p < 0.10; ** p < 0.05; *** p < 0.01. Stars refer to the underlying estimate, not the upper bound. Upper-bound cells are 95% one-sided upper confidence bounds.

Appendix and diagnostic tables

Table A1. Instrument relevance

The excluded variable F-test (often called the first-stage F-test) in an Instrumental Variable (IV) model is a diagnostic test used to check the strength of your instruments. It determines if your excluded instruments are sufficiently correlated with the endogenous explanatory variables to avoid weak instrument bias

Interpreting the Test:

The Null Hypothesis ((H_{0})): The excluded instruments are irrelevant (i.e., they do not jointly predict the endogenous variable in the first stage).

Rule of Thumb: A standard rule of thumb is that the F-statistic should be greater than 10. If the F-test yields a value below 10, it generally indicates weak instruments, meaning your IV estimates may be biased and standard errors may be unreliable.

Modern Standards: Modern econometric guidance (such as Lee et al., 2022) suggests targeting higher F-statistics (e.g., >104.7) to guarantee robustness at a 5% significance level.

Code 
include_table("instrument_validity_checks.html")
Table 33
tinytable_2wx7e32zk87jagpes5tx
Table A1: Instrument relevance: EAV changes predicted by reassessment year
All agencies All Munis HR Munis Non-HR Munis Other Schools Townships
* p < 0.1, ** p < 0.05, *** p < 0.01
Standard errors are clustered by taxing agency.
The excluded-instrument F statistic tests whether reassessment year predicts the change in EAV.
Reassessment year 6.719*** 6.587*** 6.646*** 5.833*** 6.726*** 6.824*** 6.900***
(0.138) (0.283) (0.349) (1.192) (0.254) (0.239) (0.199)
Num.Obs. 6133 1575 1002 573 2143 1995 420
R2 0.688 0.709 0.673 0.817 0.659 0.697 0.734
R2 Within 0.141 0.135 0.140 0.082 0.130 0.151 0.205
FE: n_uniqueid X X X X X X X
FE: year X X X X X X
FE: year^home_rule_ind X
Excluded-instrument F 2367.7 540.5 362.2 24.0 698.9 816.4 1202.9
Year-by-home-rule FE Yes No No No No No No
Year FE Yes Yes Yes Yes Yes Yes Yes
Agency FE Yes Yes Yes Yes Yes Yes Yes

Table A2. OLS point estimates by agency type

Wide point-estimate table

Code 
include_table("v22_table_A2_OLS_by_type_point_estimates_wide.html")
Table 34: Table A2: By agency type OLS point estimates
row Home-rule municipalities Non-home-rule municipalities Schools Townships Other
No controls 0.015 0.039* 0.096*** 0.009 0.093***
Year FE 0.077 -0.034 0.114** 0.018 0.109**
Year + unit FE 0.071 -0.051 0.102** 0.018 0.085**
IG revenue + year + unit FE 0.229* 0.127 0.157** NA NA
Same sample as IG row, no IG control 0.229* 0.117 0.156** NA NA
Enrollment + year + unit FE NA NA 0.155** NA NA
Where shown: * p < 0.10; ** p < 0.05; *** p < 0.01. Stars refer to the underlying estimate, not the upper bound. Upper-bound cells are 95% one-sided upper confidence bounds.

Long point-estimate table

Code 
include_table("v22_table_A2_OLS_by_type_point_estimates.html")
Table 35: Table A2: By agency type OLS point estimates
row agency N estimate std_error p_value
No controls Home-rule municipalities 1001 0.015 0.026 0.578
No controls Non-home-rule municipalities 571 0.039* 0.022 0.087
No controls Other 2145 0.093*** 0.018 0.000
No controls Townships 420 0.009 0.027 0.737
No controls Schools 1995 0.096*** 0.018 0.000
Year FE Home-rule municipalities 1001 0.077 0.052 0.148
Year FE Non-home-rule municipalities 571 -0.034 0.048 0.484
Year FE Other 2145 0.109** 0.044 0.014
Year FE Townships 420 0.018 0.027 0.502
Year FE Schools 1995 0.114** 0.044 0.010
Year + unit FE Home-rule municipalities 1001 0.071 0.048 0.140
Year + unit FE Non-home-rule municipalities 571 -0.051 0.047 0.286
Year + unit FE Other 2143 0.085** 0.036 0.021
Year + unit FE Townships 420 0.018 0.028 0.509
Year + unit FE Schools 1995 0.102** 0.040 0.011
IG revenue + year + unit FE Home-rule municipalities 479 0.229* 0.119 0.059
IG revenue + year + unit FE Non-home-rule municipalities 98 0.127 0.137 0.365
IG revenue + year + unit FE Schools 1722 0.157** 0.066 0.019
Same sample as IG row, no IG control Home-rule municipalities 479 0.229* 0.118 0.057
Same sample as IG row, no IG control Non-home-rule municipalities 98 0.117 0.127 0.371
Same sample as IG row, no IG control Schools 1722 0.156** 0.066 0.019
Enrollment + year + unit FE Schools 1722 0.155** 0.066 0.021
Where shown: * p < 0.10; ** p < 0.05; *** p < 0.01. Stars refer to the underlying estimate, not the upper bound. Upper-bound cells are 95% one-sided upper confidence bounds.

Table A3. IV point estimates by agency type

Wide point-estimate table

Code 
include_table("v22_table_A3_IV_by_type_point_estimates_wide.html")
Table 36: Table A3: By agency type IV point estimates
row Home-rule municipalities Non-home-rule municipalities Schools Townships Other
No controls -0.045 -0.029 0.065** -0.058 -0.052
Year FE -0.030 -0.053 0.011 -0.064 -0.208
Year + unit FE -0.031 -0.061 0.011 -0.064 -0.208
IG revenue + year + unit FE -0.514 -0.108 -0.037 NA NA
Same sample as IG row, no IG control -0.514 -0.109 -0.035 NA NA
Enrollment + year + unit FE NA NA -0.045 NA NA
Where shown: * p < 0.10; ** p < 0.05; *** p < 0.01. Stars refer to the underlying estimate, not the upper bound. Upper-bound cells are 95% one-sided upper confidence bounds.

Long point-estimate table

Code 
include_table("v22_table_A3_IV_by_type_point_estimates.html")
Table 37: Table A3: By agency type IV point estimates
row agency N estimate std_error p_value
No controls Home-rule municipalities 1001 -0.045 0.044 0.311
No controls Non-home-rule municipalities 571 -0.029 0.055 0.597
No controls Other 2145 -0.052 0.080 0.515
No controls Townships 420 -0.058 0.061 0.350
No controls Schools 1995 0.065** 0.026 0.015
Year FE Home-rule municipalities 1001 -0.030 0.070 0.668
Year FE Non-home-rule municipalities 571 -0.053 0.134 0.694
Year FE Other 2145 -0.208 0.179 0.248
Year FE Townships 420 -0.064 0.088 0.473
Year FE Schools 1995 0.011 0.033 0.734
Year + unit FE Home-rule municipalities 1001 -0.031 0.070 0.664
Year + unit FE Non-home-rule municipalities 571 -0.061 0.139 0.661
Year + unit FE Other 2143 -0.208 0.180 0.248
Year + unit FE Townships 420 -0.064 0.088 0.473
Year + unit FE Schools 1995 0.011 0.033 0.742
IG revenue + year + unit FE Home-rule municipalities 479 -0.514 0.395 0.199
IG revenue + year + unit FE Non-home-rule municipalities 98 -0.108 0.212 0.616
IG revenue + year + unit FE Schools 1722 -0.037 0.100 0.714
Same sample as IG row, no IG control Home-rule municipalities 479 -0.514 0.396 0.200
Same sample as IG row, no IG control Non-home-rule municipalities 98 -0.109 0.205 0.601
Same sample as IG row, no IG control Schools 1722 -0.035 0.101 0.728
Enrollment + year + unit FE Schools 1722 -0.045 0.101 0.659
Where shown: * p < 0.10; ** p < 0.05; *** p < 0.01. Stars refer to the underlying estimate, not the upper bound. Upper-bound cells are 95% one-sided upper confidence bounds.

Table A4. OLS symmetry tests

Code 
include_table("v22_table_A4_OLS_by_type_asym_p_equal.html")
Table 38: Table A4: P-values for OLS symmetry tests
row agency p_equal
No controls Home-rule municipalities 0.649
No controls Non-home-rule municipalities 0.201
No controls Other 0.813
No controls Townships 0.815
No controls Schools 0.392
Year FE Home-rule municipalities 0.713
Year FE Non-home-rule municipalities 0.170
Year FE Other 0.803
Year FE Townships 0.037
Year FE Schools 0.334
Year + unit FE Home-rule municipalities 0.937
Year + unit FE Non-home-rule municipalities 0.132
Year + unit FE Other 0.667
Year + unit FE Townships 0.032
Year + unit FE Schools 0.632
IG revenue + year + unit FE Home-rule municipalities 0.320
IG revenue + year + unit FE Non-home-rule municipalities 0.581
IG revenue + year + unit FE Schools 0.594
Same sample as IG row, no IG control Home-rule municipalities 0.319
Same sample as IG row, no IG control Non-home-rule municipalities 0.587
Same sample as IG row, no IG control Schools 0.585
Enrollment + year + unit FE Schools 0.580

Table A5. OLS asymmetric point estimates for positive EAV changes

Wide point-estimate table

Code 
include_table("v22_table_A5_OLS_by_type_asym_point_estimates_wide.html")
Table 39: Table A5: OLS point estimates for positive EAV changes
row Home-rule municipalities Non-home-rule municipalities Schools Townships Other
No controls 0.004 -0.007 0.111*** 0.014 0.087***
Year FE 0.083 0.027 0.102* -0.022 0.123*
Year + unit FE 0.070 0.016 0.096* -0.025 0.110
IG revenue + year + unit FE 0.293* 0.195 0.175* NA NA
Same sample as IG row, no IG control 0.293* 0.182 0.175* NA NA
Enrollment + year + unit FE NA NA 0.174* NA NA
Where shown: * p < 0.10; ** p < 0.05; *** p < 0.01. Stars refer to the underlying estimate, not the upper bound. Upper-bound cells are 95% one-sided upper confidence bounds.

Long point-estimate table

Code 
include_table("v22_table_A5_OLS_by_type_asym_point_estimates.html")
Table 40: Table A5: OLS point estimates for positive EAV changes
row agency N estimate std_error p_value
No controls Home-rule municipalities 1001 0.004 0.046 0.932
No controls Non-home-rule municipalities 571 -0.007 0.022 0.763
No controls Other 2145 0.087*** 0.030 0.004
No controls Townships 420 0.014 0.024 0.573
No controls Schools 1995 0.111*** 0.033 0.001
Year FE Home-rule municipalities 1001 0.083 0.064 0.199
Year FE Non-home-rule municipalities 571 0.027 0.054 0.625
Year FE Other 2145 0.123* 0.070 0.083
Year FE Townships 420 -0.022 0.037 0.555
Year FE Schools 1995 0.102* 0.054 0.059
Year + unit FE Home-rule municipalities 1001 0.070 0.060 0.249
Year + unit FE Non-home-rule municipalities 571 0.016 0.054 0.765
Year + unit FE Other 2143 0.110 0.069 0.113
Year + unit FE Townships 420 -0.025 0.037 0.513
Year + unit FE Schools 1995 0.096* 0.049 0.053
IG revenue + year + unit FE Home-rule municipalities 479 0.293* 0.172 0.095
IG revenue + year + unit FE Non-home-rule municipalities 98 0.195 0.192 0.322
IG revenue + year + unit FE Schools 1722 0.175* 0.095 0.067
Same sample as IG row, no IG control Home-rule municipalities 479 0.293* 0.171 0.093
Same sample as IG row, no IG control Non-home-rule municipalities 98 0.182 0.188 0.345
Same sample as IG row, no IG control Schools 1722 0.175* 0.095 0.067
Enrollment + year + unit FE Schools 1722 0.174* 0.095 0.070
Where shown: * p < 0.10; ** p < 0.05; *** p < 0.01. Stars refer to the underlying estimate, not the upper bound. Upper-bound cells are 95% one-sided upper confidence bounds.

Table A5B. OLS asymmetric point estimates for positive and negative EAV changes

Wide point-estimate table

Code 
include_table("v22_table_A5_B_OLS_by_type_asym_point_estimates_pos_neg_wide.html")
Table 41: Table A5B: OLS asymmetric point estimates for positive and negative EAV changes
row Home-rule municipalities Non-home-rule municipalities Schools Townships Other
No controls - Positive EAV change 0.004 -0.007 0.111*** 0.014 0.087***
No controls - Negative EAV change 0.036 0.126 0.067*** 0.001 0.105*
Year FE - Positive EAV change 0.083 0.027 0.102* -0.022 0.123*
Year FE - Negative EAV change 0.049 -0.219 0.169*** 0.218** 0.063
Year + unit FE - Positive EAV change 0.070 0.016 0.096* -0.025 0.110
Year + unit FE - Negative EAV change 0.077 -0.264 0.131*** 0.233** 0.001
IG revenue + year + unit FE - Positive EAV change 0.293* 0.195 0.175* NA NA
IG revenue + year + unit FE - Negative EAV change 0.094 0.031 0.114** NA NA
Same sample as IG row, no IG control - Positive EAV change 0.293* 0.182 0.175* NA NA
Same sample as IG row, no IG control - Negative EAV change 0.093 0.023 0.112** NA NA
Enrollment + year + unit FE - Positive EAV change NA NA 0.174* NA NA
Enrollment + year + unit FE - Negative EAV change NA NA 0.111** NA NA
Where shown: * p < 0.10; ** p < 0.05; *** p < 0.01. Stars refer to the underlying estimate, not the upper bound. Upper-bound cells are 95% one-sided upper confidence bounds.

Long point-estimate table

Code 
include_table("v22_table_A5_B_OLS_by_type_asym_point_estimates_pos_neg.html")
Table 42: Table A5B: OLS asymmetric point estimates for positive and negative EAV changes
row agency component N estimate std_error p_value upper_bound p_equal
No controls Home-rule municipalities Positive EAV change 1001 0.004 0.046 0.932 0.080 0.649
No controls Home-rule municipalities Negative EAV change 1001 0.036 0.033 0.289 0.090 0.649
No controls Non-home-rule municipalities Positive EAV change 571 -0.007 0.022 0.763 0.030 0.201
No controls Non-home-rule municipalities Negative EAV change 571 0.126 0.086 0.149 0.267 0.201
No controls Other Positive EAV change 2145 0.087*** 0.030 0.004 0.135 0.813
No controls Other Negative EAV change 2145 0.105* 0.055 0.061 0.196 0.813
No controls Townships Positive EAV change 420 0.014 0.024 0.573 0.053 0.815
No controls Townships Negative EAV change 420 0.001 0.055 0.990 0.091 0.815
No controls Schools Positive EAV change 1995 0.111*** 0.033 0.001 0.165 0.392
No controls Schools Negative EAV change 1995 0.067*** 0.023 0.004 0.105 0.392
Year FE Home-rule municipalities Positive EAV change 1001 0.083 0.064 0.199 0.188 0.713
Year FE Home-rule municipalities Negative EAV change 1001 0.049 0.063 0.438 0.152 0.713
Year FE Non-home-rule municipalities Positive EAV change 571 0.027 0.054 0.625 0.116 0.170
Year FE Non-home-rule municipalities Negative EAV change 571 -0.219 0.159 0.175 0.042 0.170
Year FE Other Positive EAV change 2145 0.123* 0.070 0.083 0.238 0.803
Year FE Other Negative EAV change 2145 0.063 0.192 0.745 0.378 0.803
Year FE Townships Positive EAV change 420 -0.022 0.037 0.555 0.038 0.037
Year FE Townships Negative EAV change 420 0.218** 0.089 0.021 0.365 0.037
Year FE Schools Positive EAV change 1995 0.102* 0.054 0.059 0.191 0.334
Year FE Schools Negative EAV change 1995 0.169*** 0.041 0.000 0.236 0.334
Year + unit FE Home-rule municipalities Positive EAV change 1001 0.070 0.060 0.249 0.168 0.937
Year + unit FE Home-rule municipalities Negative EAV change 1001 0.077 0.064 0.229 0.182 0.937
Year + unit FE Non-home-rule municipalities Positive EAV change 571 0.016 0.054 0.765 0.105 0.132
Year + unit FE Non-home-rule municipalities Negative EAV change 571 -0.264 0.164 0.115 0.005 0.132
Year + unit FE Other Positive EAV change 2143 0.110 0.069 0.113 0.223 0.667
Year + unit FE Other Negative EAV change 2143 0.001 0.199 0.998 0.328 0.667
Year + unit FE Townships Positive EAV change 420 -0.025 0.037 0.513 0.037 0.032
Year + unit FE Townships Negative EAV change 420 0.233** 0.093 0.018 0.386 0.032
Year + unit FE Schools Positive EAV change 1995 0.096* 0.049 0.053 0.177 0.632
Year + unit FE Schools Negative EAV change 1995 0.131*** 0.044 0.004 0.204 0.632
IG revenue + year + unit FE Home-rule municipalities Positive EAV change 479 0.293* 0.172 0.095 0.576 0.320
IG revenue + year + unit FE Home-rule municipalities Negative EAV change 479 0.094 0.063 0.143 0.198 0.320
IG revenue + year + unit FE Non-home-rule municipalities Positive EAV change 98 0.195 0.192 0.322 0.510 0.581
IG revenue + year + unit FE Non-home-rule municipalities Negative EAV change 98 0.031 0.189 0.871 0.342 0.581
IG revenue + year + unit FE Schools Positive EAV change 1722 0.175* 0.095 0.067 0.331 0.594
IG revenue + year + unit FE Schools Negative EAV change 1722 0.114** 0.046 0.014 0.189 0.594
Same sample as IG row, no IG control Home-rule municipalities Positive EAV change 479 0.293* 0.171 0.093 0.574 0.319
Same sample as IG row, no IG control Home-rule municipalities Negative EAV change 479 0.093 0.063 0.147 0.198 0.319
Same sample as IG row, no IG control Non-home-rule municipalities Positive EAV change 98 0.182 0.188 0.345 0.490 0.587
Same sample as IG row, no IG control Non-home-rule municipalities Negative EAV change 98 0.023 0.176 0.899 0.312 0.587
Same sample as IG row, no IG control Schools Positive EAV change 1722 0.175* 0.095 0.067 0.330 0.585
Same sample as IG row, no IG control Schools Negative EAV change 1722 0.112** 0.045 0.015 0.187 0.585
Enrollment + year + unit FE Schools Positive EAV change 1722 0.174* 0.095 0.070 0.330 0.580
Enrollment + year + unit FE Schools Negative EAV change 1722 0.111** 0.045 0.015 0.184 0.580
Where shown: * p < 0.10; ** p < 0.05; *** p < 0.01. Stars refer to the underlying estimate, not the upper bound. Upper-bound cells are 95% one-sided upper confidence bounds.

Table A6. IV symmetry tests

Code 
include_table("v22_table_A6_IV_by_type_asym_p_equal.html")
Table 43: Table A6: P-values for IV symmetry tests
row agency p_equal
No controls Home-rule municipalities 0.904
No controls Non-home-rule municipalities 0.236
No controls Other 0.050
No controls Townships 0.160
No controls Schools 0.316
Year FE Home-rule municipalities 0.707
Year FE Non-home-rule municipalities 0.786
Year FE Other 0.141
Year FE Townships 0.230
Year FE Schools 0.936
Year + unit FE Home-rule municipalities 0.687
Year + unit FE Non-home-rule municipalities 0.834
Year + unit FE Other 0.140
Year + unit FE Townships 0.233
Year + unit FE Schools 0.924
IG revenue + year + unit FE Home-rule municipalities 0.096
IG revenue + year + unit FE Non-home-rule municipalities 0.042
IG revenue + year + unit FE Schools 0.745
Same sample as IG row, no IG control Home-rule municipalities 0.098
Same sample as IG row, no IG control Non-home-rule municipalities 0.051
Same sample as IG row, no IG control Schools 0.769
Enrollment + year + unit FE Schools 0.717

Table A7. IV asymmetric point estimates for positive EAV changes

Wide point-estimate table

Code 
include_table("v22_table_A7_IV_by_type_asym_point_estimates_wide.html")
Table 44: Table A7: IV point estimates for positive EAV changes
row Home-rule municipalities Non-home-rule municipalities Schools Townships Other
No controls -0.047 -0.003 0.072*** -0.031 -0.029
Year FE -0.013 -0.085** 0.010 -0.021 -0.113
Year + unit FE -0.012 -0.086** 0.009 -0.021 -0.113
IG revenue + year + unit FE -0.064 0.299** -0.012 NA NA
Same sample as IG row, no IG control -0.065 0.270** -0.012 NA NA
Enrollment + year + unit FE NA NA -0.016 NA NA
Where shown: * p < 0.10; ** p < 0.05; *** p < 0.01. Stars refer to the underlying estimate, not the upper bound. Upper-bound cells are 95% one-sided upper confidence bounds.

Long point-estimate table

Code 
include_table("v22_table_A7_IV_by_type_asym_point_estimates.html")
Table 45: Table A7: IV point estimates for positive EAV changes
row agency N estimate std_error p_value
No controls Home-rule municipalities 1001 -0.047 0.030 0.123
No controls Non-home-rule municipalities 571 -0.003 0.036 0.944
No controls Other 2145 -0.029 0.072 0.687
No controls Townships 420 -0.031 0.044 0.484
No controls Schools 1995 0.072*** 0.020 0.000
Year FE Home-rule municipalities 1001 -0.013 0.032 0.688
Year FE Non-home-rule municipalities 571 -0.085** 0.035 0.021
Year FE Other 2145 -0.113 0.124 0.364
Year FE Townships 420 -0.021 0.056 0.709
Year FE Schools 1995 0.010 0.019 0.615
Year + unit FE Home-rule municipalities 1001 -0.012 0.032 0.710
Year + unit FE Non-home-rule municipalities 571 -0.086** 0.036 0.020
Year + unit FE Other 2143 -0.113 0.124 0.366
Year + unit FE Townships 420 -0.021 0.056 0.710
Year + unit FE Schools 1995 0.009 0.019 0.640
IG revenue + year + unit FE Home-rule municipalities 479 -0.064 0.074 0.387
IG revenue + year + unit FE Non-home-rule municipalities 98 0.299** 0.116 0.019
IG revenue + year + unit FE Schools 1722 -0.012 0.033 0.723
Same sample as IG row, no IG control Home-rule municipalities 479 -0.065 0.075 0.387
Same sample as IG row, no IG control Non-home-rule municipalities 98 0.270** 0.105 0.018
Same sample as IG row, no IG control Schools 1722 -0.012 0.033 0.713
Enrollment + year + unit FE Schools 1722 -0.016 0.032 0.614
Where shown: * p < 0.10; ** p < 0.05; *** p < 0.01. Stars refer to the underlying estimate, not the upper bound. Upper-bound cells are 95% one-sided upper confidence bounds.

Table A7B. IV asymmetric point estimates for positive and negative EAV changes

Wide point-estimate table

Code 
include_table("v22_table_A7_B_IV_by_type_asym_point_estimates_pos_neg_wide.html")
Table 46: Table A7B: IV asymmetric point estimates for positive and negative EAV changes
row Home-rule municipalities Non-home-rule municipalities Schools Townships Other
No controls - Positive EAV change -0.047 -0.003 0.072*** -0.031 -0.029
No controls - Negative EAV change -0.060 0.156 0.128*** 0.168 0.159**
Year FE - Positive EAV change -0.013 -0.085** 0.010 -0.021 -0.113
Year FE - Negative EAV change 0.034 -0.143 0.005 0.106 0.148
Year + unit FE - Positive EAV change -0.012 -0.086** 0.009 -0.021 -0.113
Year + unit FE - Negative EAV change 0.038 -0.132 0.004 0.107 0.149
IG revenue + year + unit FE - Positive EAV change -0.064 0.299** -0.012 NA NA
IG revenue + year + unit FE - Negative EAV change 0.144* -0.039 0.009 NA NA
Same sample as IG row, no IG control - Positive EAV change -0.065 0.270** -0.012 NA NA
Same sample as IG row, no IG control - Negative EAV change 0.142* -0.047 0.006 NA NA
Enrollment + year + unit FE - Positive EAV change NA NA -0.016 NA NA
Enrollment + year + unit FE - Negative EAV change NA NA 0.006 NA NA
Where shown: * p < 0.10; ** p < 0.05; *** p < 0.01. Stars refer to the underlying estimate, not the upper bound. Upper-bound cells are 95% one-sided upper confidence bounds.

Long point-estimate table

Code 
include_table("v22_table_A7_B_IV_by_type_asym_point_estimates_pos_neg.html")
Table 47: Table A7B: IV asymmetric point estimates for positive and negative EAV changes
row agency component N estimate std_error p_value upper_bound F_neg F_pos SW_F_neg SW_F_pos p_exog p_equal
No controls Home-rule municipalities Positive EAV change 1001 -0.047 0.030 0.123 0.002 1581.1 558.4 566.5 667.8 0.010 0.904
No controls Home-rule municipalities Negative EAV change 1001 -0.060 0.084 0.480 0.079 1581.1 558.4 566.5 667.8 0.010 0.904
No controls Non-home-rule municipalities Positive EAV change 571 -0.003 0.036 0.944 0.056 823.7 1008.6 625.4 1150.3 0.760 0.236
No controls Non-home-rule municipalities Negative EAV change 571 0.156 0.103 0.137 0.325 823.7 1008.6 625.4 1150.3 0.760 0.236
No controls Other Positive EAV change 2145 -0.029 0.072 0.687 0.089 2998.9 1073.9 1130.5 1282.1 0.377 0.050
No controls Other Negative EAV change 2145 0.159** 0.071 0.027 0.276 2998.9 1073.9 1130.5 1282.1 0.377 0.050
No controls Townships Positive EAV change 420 -0.031 0.044 0.484 0.041 979.0 1099.4 582.6 794.5 0.172 0.160
No controls Townships Negative EAV change 420 0.168 0.107 0.126 0.344 979.0 1099.4 582.6 794.5 0.172 0.160
No controls Schools Positive EAV change 1995 0.072*** 0.020 0.000 0.105 2869.6 1631.7 1425.6 1536.0 0.339 0.316
No controls Schools Negative EAV change 1995 0.128*** 0.040 0.002 0.194 2869.6 1631.7 1425.6 1536.0 0.339 0.316
Year FE Home-rule municipalities Positive EAV change 1001 -0.013 0.032 0.688 0.040 202.1 404.2 325.9 809.3 0.398 0.707
Year FE Home-rule municipalities Negative EAV change 1001 0.034 0.107 0.754 0.209 202.1 404.2 325.9 809.3 0.398 0.707
Year FE Non-home-rule municipalities Positive EAV change 571 -0.085** 0.035 0.021 -0.027 125.6 1162.2 211.8 1768.2 0.154 0.786
Year FE Non-home-rule municipalities Negative EAV change 571 -0.143 0.195 0.470 0.178 125.6 1162.2 211.8 1768.2 0.154 0.786
Year FE Other Positive EAV change 2145 -0.113 0.124 0.364 0.091 373.0 918.5 657.2 1676.2 0.289 0.141
Year FE Other Negative EAV change 2145 0.148 0.112 0.189 0.332 373.0 918.5 657.2 1676.2 0.289 0.141
Year FE Townships Positive EAV change 420 -0.021 0.056 0.709 0.072 113.4 649.4 231.4 1301.0 0.237 0.230
Year FE Townships Negative EAV change 420 0.106 0.068 0.130 0.218 113.4 649.4 231.4 1301.0 0.237 0.230
Year FE Schools Positive EAV change 1995 0.010 0.019 0.615 0.041 346.1 725.5 599.2 1447.1 0.001 0.936
Year FE Schools Negative EAV change 1995 0.005 0.050 0.920 0.088 346.1 725.5 599.2 1447.1 0.001 0.936
Year + unit FE Home-rule municipalities Positive EAV change 1001 -0.012 0.032 0.710 0.041 199.9 400.6 322.9 806.3 0.410 0.687
Year + unit FE Home-rule municipalities Negative EAV change 1001 0.038 0.108 0.724 0.215 199.9 400.6 322.9 806.3 0.410 0.687
Year + unit FE Non-home-rule municipalities Positive EAV change 571 -0.086** 0.036 0.020 -0.028 119.3 1176.6 188.7 1765.3 0.238 0.834
Year + unit FE Non-home-rule municipalities Negative EAV change 571 -0.132 0.200 0.513 0.197 119.3 1176.6 188.7 1765.3 0.238 0.834
Year + unit FE Other Positive EAV change 2143 -0.113 0.124 0.366 0.092 370.6 910.9 657.5 1668.5 0.356 0.140
Year + unit FE Other Negative EAV change 2143 0.149 0.112 0.185 0.334 370.6 910.9 657.5 1668.5 0.356 0.140
Year + unit FE Townships Positive EAV change 420 -0.021 0.056 0.710 0.071 112.3 650.6 231.2 1311.5 0.179 0.233
Year + unit FE Townships Negative EAV change 420 0.107 0.069 0.134 0.220 112.3 650.6 231.2 1311.5 0.179 0.233
Year + unit FE Schools Positive EAV change 1995 0.009 0.019 0.640 0.040 342.7 722.3 593.0 1444.6 0.005 0.924
Year + unit FE Schools Negative EAV change 1995 0.004 0.050 0.942 0.086 342.7 722.3 593.0 1444.6 0.005 0.924
IG revenue + year + unit FE Home-rule municipalities Positive EAV change 479 -0.064 0.074 0.387 0.057 92.4 116.8 223.0 229.3 0.063 0.096
IG revenue + year + unit FE Home-rule municipalities Negative EAV change 479 0.144* 0.079 0.076 0.274 92.4 116.8 223.0 229.3 0.063 0.096
IG revenue + year + unit FE Non-home-rule municipalities Positive EAV change 98 0.299** 0.116 0.019 0.490 121.3 15.2 292.0 36.3 0.876 0.042
IG revenue + year + unit FE Non-home-rule municipalities Negative EAV change 98 -0.039 0.113 0.735 0.147 121.3 15.2 292.0 36.3 0.876 0.042
IG revenue + year + unit FE Schools Positive EAV change 1722 -0.012 0.033 0.723 0.043 287.7 367.8 604.1 769.4 0.004 0.745
IG revenue + year + unit FE Schools Negative EAV change 1722 0.009 0.050 0.866 0.092 287.7 367.8 604.1 769.4 0.004 0.745
Same sample as IG row, no IG control Home-rule municipalities Positive EAV change 479 -0.065 0.075 0.387 0.058 93.2 116.7 223.5 228.6 0.064 0.098
Same sample as IG row, no IG control Home-rule municipalities Negative EAV change 479 0.142* 0.080 0.080 0.273 93.2 116.7 223.5 228.6 0.064 0.098
Same sample as IG row, no IG control Non-home-rule municipalities Positive EAV change 98 0.270** 0.105 0.018 0.442 150.3 16.4 311.6 38.2 0.884 0.051
Same sample as IG row, no IG control Non-home-rule municipalities Negative EAV change 98 -0.047 0.108 0.668 0.131 150.3 16.4 311.6 38.2 0.884 0.051
Same sample as IG row, no IG control Schools Positive EAV change 1722 -0.012 0.033 0.713 0.042 288.8 366.9 606.0 772.5 0.004 0.769
Same sample as IG row, no IG control Schools Negative EAV change 1722 0.006 0.051 0.905 0.090 288.8 366.9 606.0 772.5 0.004 0.769
Enrollment + year + unit FE Schools Positive EAV change 1722 -0.016 0.032 0.614 0.037 287.5 368.6 609.4 774.4 0.003 0.717
Enrollment + year + unit FE Schools Negative EAV change 1722 0.006 0.051 0.903 0.090 287.5 368.6 609.4 774.4 0.003 0.717
Where shown: * p < 0.10; ** p < 0.05; *** p < 0.01. Stars refer to the underlying estimate, not the upper bound. Upper-bound cells are 95% one-sided upper confidence bounds.

Table A8. P-values for cumulative current and lagged effects

Code 
include_table("v22_table_A8_OLS_by_type_lagged_p_values.html")
Table 48: Table A8: Cumulative-effect, lag-sum, and joint lag-exclusion tests
row agency cumulative_effect_p p_lags joint_lag_F joint_lag_p
No controls Home-rule municipalities 0.067 0.016 3.1 0.044
No controls Non-home-rule municipalities 0.146 0.679 4.8 0.008
No controls Other 0.000 0.046 2.1 0.126
No controls Townships 0.040 0.011 5.4 0.005
No controls Schools 0.000 0.196 4.9 0.008
Year FE Home-rule municipalities 0.554 0.809 0.1 0.948
Year FE Non-home-rule municipalities 0.080 0.026 3.7 0.026
Year FE Other 0.111 0.209 0.9 0.408
Year FE Townships 0.712 0.728 0.7 0.518
Year FE Schools 0.000 0.000 12.0 0.000
Year + unit FE Home-rule municipalities 0.385 0.552 0.3 0.730
Year + unit FE Non-home-rule municipalities 0.259 0.095 2.8 0.061
Year + unit FE Other 0.262 0.392 0.8 0.447
Year + unit FE Townships 0.585 0.574 0.7 0.489
Year + unit FE Schools 0.001 0.001 5.4 0.005
IG revenue + year + unit FE Home-rule municipalities 0.214 0.477 0.3 0.765
IG revenue + year + unit FE Non-home-rule municipalities 0.643 0.784 0.0 0.962
IG revenue + year + unit FE Schools 0.004 0.007 3.8 0.023
Same sample as IG row, no IG control Home-rule municipalities 0.214 0.478 0.3 0.759
Same sample as IG row, no IG control Non-home-rule municipalities 0.688 0.736 0.1 0.939
Same sample as IG row, no IG control Schools 0.005 0.008 3.7 0.025
Enrollment + year + unit FE Schools 0.004 0.006 4.0 0.019
cumulative_effect_p tests H0: beta_current + beta_lag1 + beta_lag2 = 0. p_lags tests H0: beta_lag1 + beta_lag2 = 0 (excludes the current coefficient). joint_lag_F and joint_lag_p test H0: beta_lag1 = beta_lag2 = 0. Tests use agency-clustered covariance estimates; the lag-sum test includes covariance between the two lag estimates.

Table A9. Cumulative current and lagged point estimates

Wide point-estimate table

Code 
include_table("v22_table_A9_OLS_by_type_lagged_point_estimates_wide.html")
Table 49: Table A9: Sum of current and lagged EAV coefficients
row Home-rule municipalities Non-home-rule municipalities Schools Townships Other
No controls 0.073* 0.065 0.114*** 0.054** 0.171***
Year FE 0.110 0.233* 0.343*** 0.036 0.391
Year + unit FE 0.146 0.161 0.257*** 0.070 0.300
IG revenue + year + unit FE 0.361 0.098 0.280*** NA NA
Same sample as IG row, no IG control 0.360 0.075 0.278*** NA NA
Enrollment + year + unit FE NA NA 0.278*** NA NA
Where shown: * p < 0.10; ** p < 0.05; *** p < 0.01. Stars refer to the underlying estimate, not the upper bound. Upper-bound cells are 95% one-sided upper confidence bounds.

Long point-estimate table

Code 
include_table("v22_table_A9_OLS_by_type_lagged_point_estimates.html")
Table 50: Table A9: Sum of current and lagged EAV coefficients
row agency lagged_magnitude std_error lagged_p
No controls Home-rule municipalities 0.073* 0.039 0.067
No controls Non-home-rule municipalities 0.065 0.043 0.146
No controls Other 0.171*** 0.043 0.000
No controls Townships 0.054** 0.025 0.040
No controls Schools 0.114*** 0.021 0.000
Year FE Home-rule municipalities 0.110 0.184 0.554
Year FE Non-home-rule municipalities 0.233* 0.130 0.080
Year FE Other 0.391 0.244 0.111
Year FE Townships 0.036 0.098 0.712
Year FE Schools 0.343*** 0.082 0.000
Year + unit FE Home-rule municipalities 0.146 0.168 0.385
Year + unit FE Non-home-rule municipalities 0.161 0.140 0.259
Year + unit FE Other 0.300 0.267 0.262
Year + unit FE Townships 0.070 0.127 0.585
Year + unit FE Schools 0.257*** 0.078 0.001
IG revenue + year + unit FE Home-rule municipalities 0.361 0.287 0.214
IG revenue + year + unit FE Non-home-rule municipalities 0.098 0.208 0.643
IG revenue + year + unit FE Schools 0.280*** 0.096 0.004
Same sample as IG row, no IG control Home-rule municipalities 0.360 0.287 0.214
Same sample as IG row, no IG control Non-home-rule municipalities 0.075 0.185 0.688
Same sample as IG row, no IG control Schools 0.278*** 0.096 0.005
Enrollment + year + unit FE Schools 0.278*** 0.095 0.004
Where shown: * p < 0.10; ** p < 0.05; *** p < 0.01. Stars refer to the underlying estimate, not the upper bound. Upper-bound cells are 95% one-sided upper confidence bounds.