Do redistributive state taxes reduce inequality?

tax rates? No No One-Tailed F-Test: HO Is That Tax Redistribution Coefficients [greater than or equal to] 1 Can We Reject the Null That Pre-Tax Inequality Fully Adjusts in Response to ... Current taxes? Yes Yes

[P = 0.03] [P = 0.04] The sum of lagged taxes? Yes Yes

[P = 0.06] [P = 0.02] The sum of current and lagged

tax rates? Yes Yes

[P < 0.01] [P < 0.01]

[3] [4]

[delta] = 2.5 [delta] = 3.5

More Weight on Bottom of

Distribution Than Gini Tax (redistribution.sub.t] -0.394 -0.57

[0.820] [0.816] Tax (redistribution.sub.t-1] 0.94 0.885

[0.897] [1.017] Tax (redistribution.sub.t-2] -0.223 -0.176

[0.947] [1.026] Tax (redistribution.sub.t-3] -0.306 -0.41

[1.025] [1.215] Tax (redistribution.sub.t-4] -1.843 ** -2.160 ***

[0.738] [0.771] Tax (redistribution.sub.t-5] 0.281 0.345

[0.685] [0.757] Tax (redistribution.sub.t-6] -1.013 -0.883

[0.867] [0.953] Time-varying state characteristics? Yes Yes State and year fixed effects? Yes Yes Region-specific time trend? Yes Yes R-squared 0.61 0.61 Sum of lagged redistribution

coefficients -2.165 -2.399

[1.106] [1.131] Sum of all redistribution

coefficients -2.559 -2.969

[0.901] [0.906] One-Tailed F-test: HO Is That Tax Redistribution Coefficients [less than or equal to] 0 Can We Reject the Null That Pre-Tax Inequality Is Unaffected by ... Current taxes? No No The sum of lagged taxes? No No The sum of current and lagged

tax rates? No No One-Tailed F-Test: HO Is That Tax Redistribution Coefficients [greater than or equal to] 1 Can We Reject the Null That Pre-Tax Inequality Fully Adjusts in Response to ... Current taxes? Yes Yes

[P = 0.05] [P = 0.03] The sum of lagged taxes? Yes Yes

[P < 0.01] [P < 0.01] The sum of current and lagged

tax rates? Yes Yes

[P < 0.01] [P < 0.01] Note: Robust standard errors, clustered at the state level, in brackets. *, **, and *** denote statistical significance at the 10 %, 5%, and l% levels, respectively. "Tax Redistribution" is the negative of the Reynolds-Smolensky index, calculated as [[bar.SGB].sub.[delta]] - [[bar.SGA].sub.[delta]] (see text for details). Time-varying state characteristics are log real personal income per capita, the unemployment rate, the unionization rate, the sales tax rate, the maximum estate/inheritance tax rate, and an indicator for whether the state has an estate tax. Sample is restricted to inequality observations from 1983-2002. Sample size for all specifications is 1,020. TABLE 4 DOES TAX REDISTRIBUTION DRIVE INTERSTATE MIGRATION?

[1] [2] [3]

Incoming Outgoing Wage Ratio:

Migration Migration Incoming/ Dependent Variable Rate Rate Nonmovers [TR.sub.t] -0.670 -0.369 17.251

[0.650] [0.864] [11.760] [TR.sub.t-1] 1.063 -1.258 -29.954 *

[1.301] [0.884] [15.078] [TR.sub.t-2] -0.91 -0.603 16.575

[1.267] [0.821] [19.557] [TR.sub.t-3] -0.309 0.45 -8.149

[1.128] [0.973] [19.264] [TR.sub.t-4] -0.884 0.055 6.558

[1.008] [1.088] [15.297] [TR.sub.t-5] 1.965 * 0.351 -12.26

[1.101] [0.883] [12.736] [TR..sub.t-6] -1.269 0.487 4.095

[0.851] [0.757] [11.551] Time-varying state Yes Yes Yes

characteristics? State and year fixed Yes Yes Yes

effects (FE)? Region-specific time Yes Yes Yes

trend? Observations 918 918 918 R-squared 0.73 0.73 0.15 Sum of lagged TR -0.344 -0.518 3.135

coefficients [0.910] [1.278] [11.457] Sum of all TR -1.014 -0.887 -5.883

coefficients [0.806] [1.157] [11.283] Two-Tailed F-Test: HO Is That Tax Redistribution Coefficients = 0 Can We Reject the Null That the Dependent Variable Is Unaffected by ...

Current taxes? No No No

The sum of lagged No No Yes

taxes? [P = 0.05]

The sum of current and

lagged tax rates? No No No

[4] [5] [6]

Wage Ratio: Population

Outgoing/ Growth Log Dependent Variable Nonmovers Rate Population [TR.sub.t] -18.84 -0.704 -6.479 **

[24.131] [0.526] [3.078] [TR.sub.t-1] 63.088 ** 0.597 1.375

[25.599] [0.501] [2.723] [TR.sub.t-2] -21.978 -1.202 -1.345

[20.474] [0.760] [1.505] [TR.sub.t-3] 11.423 0.717 -1.824

[21.365] [0.793] [1.506] [TR.sub.t-4] -9.134 -0.012 1.862

[22.947] [1.002] [1.774] [TR.sub.t-5] 10.183 0.487 -2.683

[25.651] [0.894] (1.625] [TR..sub.t-6] -40.293* 0.107 -0.983

[23.171] [0.583] [3.506] Time-varying state Yes Yes Yes

characteristics? State and year fixed Yes Yes Yes

effects (FE)? Region-specific time Yes Yes Yes

trend? Observations 917 1,020 1,020 R-squared 0.13 0.64 0.99 Sum of lagged TR 13.288 0.694 -9.350

coefficients [23.724] [0.766] [6.480] Sum of all TR -5.552 -0.010 -15.829

coefficients [17.104] [0.590] [6.434] Two-Tailed F-Test: HO Is That Tax Redistribution Coefficients = 0 Can We Reject the Null That the Dependent Variable Is Unaffected by ...

Current taxes? No No Yes

[P = 0.04]

The sum of lagged No No No

taxes?

The sum of current and Yes

lagged tax rates? No No [P = 0.02] Note: Robust standard errors, clustered at the state level, in brackets. *, **, and *** denote statistical significance at the 10%, 5%, and 1%, levels, respectively. "TR" is the negative of the Reynolds-Smolensky index, calculated as [bar.GB] - [bar.GA] (see text for details). The population growth rate (column 5) is the change in log population from year t - 1 to year t. Time-varying state characteristics are log real personal income per capita, the unemployment rate, the unionization rate, the sales tax rate, the maximum estate/inheritance tax rate, and an indicator for whether the state has an estate tax. Dependent variables in columns I to 4 are measured from [March.sub.t] to [March.sub.t+1]. TABLE 5 HOW DO REDISTRIBUTIVE TAXES AFFECT POST-TAX INEQUALITY AND AVERAGE INCOME?

[1] [2]

Post-Tax Log Real