Rugged Individualism and Collective (In)action Appendix

Appendix Figure A.1: Total Frontier Experience (1790 to 1890) Notes: This figure is reproduced from Bazzi et al. (forthcoming). It is based on county-level data from NHGIS (Manson et al., 2019). Total frontier experience is the total number of years the county was within 100 km of the frontier line and its population density was below 6 people per square mile, between 1790–1890. The white areas to the east of the 1790 main frontier line are counties for which we do not know frontier history given the lack of Population Census data before 1790. The white areas to the west are beyond the 1890 frontier line. Figure A.2: TFE and Social Distancing controlling for Lagged Per Capita Cases and Deaths Specification: County and Week FE (a) Non-essential Visits (b) Time in Workplaces (Unacast) (Google Community Mobility) Specification: County and State-by-Week FE (c) Non-essential Visits (d) Time in Workplaces (Unacast) (Google Community Mobility) Notes: This figure displays estimation results for our event-study specification in equation 1 for two different outcomes, with and without state-by-week fixed effects. All panels display point estimates by week and 95% confidence intervals corresponding to standard errors clustered by arbitrary 60-square-mile grid-cells. Figure A.3: Robustness Check: TFE and Social Distancing Specification: County and Week FE (a) Non-essential Visits (b) Time in Workplaces (Unacast) (Google Community Mobility) Specification: County and State-by-Week FE (a) Non-essential Visits (b) Time in Workplaces (Unacast) (Google Community Mobility) Notes: This figure displays estimation results for our event-study specification in equation 1 for two different outcomes and controlling for, separately, county level population density, race (share of white), mean temperature, median income or education (share with post-secondary educatiob) interacted with time-to-event dummies. The first panel icnludes county and week fixed effects while the second panel uses county and state-by-week fixed effects. All panels display point estimates by week and 95% confidence intervals corresponding to standard errors clustered by arbitrary 60-square-mile grid-cells. Table A.1: Robustness Checks: TFE and Social Distancing (1) (2) (3) (4) (5) (6) (7) 1.003* (0.531) -1.027*** (0.061) 1.731*** (0.544) -0.007*** (0.001) 0.282*** (0.058) -0.469** (0.237) 0.000*** (0.000) -1.097*** (0.106) 35,828 -10.8 35,750 -10.8 (a) Non-essential Visits (Unacast) TFE × post-March-13 2.645*** (0.619) pop. density × post-March-13 1.746*** (0.529) -0.018*** (0.002) share white × post-March-13 3.238*** (0.579) 3.255*** (0.645) 2.057*** (0.593) 0.483*** (0.054) temperature × post-March-13 -0.685*** (0.218) median income × post-March-13 -0.000*** (0.000) share post-sec. educ × post-March-13 Number of County–Weeks Dep. Var. Mean 35,828 -10.8 35,828 -10.8 35,828 -10.8 35,750 -10.8 35,828 -10.8 (b) Time at Workplaces (Google Community Mobility) TFE × post-March-13 1.075*** (0.192) pop. density × post-March-13 0.787*** (0.151) -0.007*** (0.001) share white × post-March-13 1.119*** (0.191) 0.996*** (0.189) 0.635*** (0.131) -0.420*** (0.019) 0.390*** (0.121) -0.003*** (0.001) 0.007 (0.012) -0.045 (0.041) -0.000*** (0.000) -0.306*** (0.026) 49,938 -22.4 X X 49,857 -22.4 X X 0.028** (0.011) temperature × post-March-13 0.070 (0.053) median income × post-March-13 -0.000*** (0.000) share post-sec. educ × post-March-13 Number of County–Weeks Dep. Var. Mean County Fixed Effects Time Fixed Effects 0.413*** (0.129) 49,938 -22.4 X X 49,938 -22.4 X X 49,938 -22.4 X X 49,857 -22.4 X X 49,938 -22.4 X X Notes: This table reports estimates of the simpler difference-in-difference analogue of specification in equation (1) for two different outcomes and accounting for potential confounders interacted with an indicator for post-March-13. Standard errors are clustered by arbitrary 60-square-mile grid cells. Significance levels: ∗ : 10% ∗∗ : 5% ∗ ∗ ∗ : 1%. Table A.2: Robustness Checks: TFE and Mask Use Share who always wear a mask in public when they expect to be within six feet of another person total frontier experience (1) (2) (3) (4) (5) (6) (7) -0.019*** (0.003) -0.014*** (0.003) 0.000*** (0.000) -0.016*** (0.003) -0.019*** (0.003) -0.015*** (0.003) -0.012*** (0.003) 0.005*** (0.000) -0.008*** (0.003) 0.000*** (0.000) -0.002*** (0.000) -0.008 (0.007) 0.000*** (0.000) 0.002*** (0.000) 2,036 0.482 0.60 X X 2,036 0.482 0.63 X X pop. density share white -0.002*** (0.000) temperature -0.001 (0.009) median income 0.000*** (0.000) share post-sec. educ Number of Counties Dep. Var. Mean R2 State Fixed Effects Geographic/Agroclimatic Controls 2,036 0.482 0.54 X X 2,036 0.482 0.58 X X 2,036 0.482 0.56 X X 2,036 0.482 0.54 X X 2,036 0.482 0.58 X X Notes: This table reports estimates of equation (2) for the main mask use outcome while accounting for potential confounders at the county level. The regressions control for state fixed effects and the following additional predetermined controls: county area; county centroid latitude and longitude; distance to oceans,lakes and rivers from county centroid; mean county rainfall; elevation; and average potential agricultural yield. Standard errors are clustered by arbitrary 60square-mile grid cells. Standard errors are clustered by arbitrary 60-square-mile grid cells. Significance levels: ∗ : 10% ∗∗ : 5% ∗ ∗ ∗ : 1%. Table A.3: Robustness Checks: TFE and Local Policies (I) (1) (2) (3) (4) (5) (6) (7) (a) County Policies: Emergency Declarations total frontier experience -0.042*** (0.011) pop. density -0.025** (0.011) 0.000*** (0.000) share white -0.035*** (0.011) -0.042*** (0.011) -0.030*** (0.010) 0.012*** (0.001) -0.013 (0.010) 0.000*** (0.000) -0.002*** (0.001) -0.032* (0.017) 0.000*** (0.000) 0.005*** (0.002) 2,036 0.236 0.24 X X 2,036 0.236 0.27 X X -0.003*** (0.001) temperature -0.013 (0.021) median income 0.000*** (0.000) share post-sec. educ Number of Counties Dep. Var. Mean R2 State Fixed Effects Geographic/Agroclimatic Controls -0.023** (0.011) 2,036 0.236 0.20 X X 2,036 0.236 0.24 X X 2,036 0.236 0.20 X X 2,036 0.236 0.20 X X 2,036 0.236 0.23 X X (b) County Policies: Stay at Home Policies total frontier experience -0.025*** (0.007) pop. density -0.014** (0.006) 0.000*** (0.000) share white -0.021*** (0.007) -0.025*** (0.007) -0.019*** (0.006) 0.007*** (0.001) -0.009 (0.006) 0.000*** (0.000) -0.001* (0.000) 0.012 (0.014) 0.000** (0.000) 0.003** (0.001) 2,036 0.048 0.21 X X 2,036 0.048 0.25 X X -0.002*** (0.001) temperature 0.024* (0.014) median income 0.000*** (0.000) share post-sec. educ Number of Counties Dep. Var. Mean R2 State Fixed Effects Geographic/Agroclimatic Controls -0.014** (0.006) 2,036 0.048 0.16 X X 2,036 0.048 0.23 X X 2,036 0.048 0.16 X X 2,036 0.048 0.16 X X 2,036 0.048 0.19 X X (c) County Policies: Business Closure Policies total frontier experience -0.003 (0.003) pop. density -0.001 (0.003) 0.000 (0.000) share white -0.002 (0.003) -0.003 (0.003) -0.002 (0.003) 0.001* (0.000) -0.001 (0.003) 0.000 (0.000) -0.000 (0.000) -0.001 (0.003) 0.000 (0.000) 0.000 (0.000) 2,036 0.006 0.06 X X 2,036 0.006 0.07 X X -0.000 (0.000) temperature 0.001 (0.003) median income 0.000 (0.000) share post-sec. educ Number of Counties Dep. Var. Mean R2 State Fixed Effects Geographic/Agroclimatic Controls -0.002 (0.003) 2,036 0.006 0.06 X X 2,036 0.006 0.07 X X 2,036 0.006 0.06 X X 2,036 0.006 0.06 X X 2,036 0.006 0.06 X X Notes: This table reports estimates of equation (2) for local policy outcomes while accounting for potential confounders at the county level. The regressions control for state fixed effects and the following additional predetermined controls: county area; county centroid latitude and longitude; distance to oceans,lakes and rivers from county centroid; mean county rainfall; elevation; and average potential agricultural yield. Standard errors are clustered by arbitrary 60-squaremile grid cells. Standard errors are clustered by arbitrary 60-square-mile grid cells. Significance levels: ∗ : 10% ∗∗ : 5% ∗ ∗ ∗ : 1%. Table A.4: Robustness Checks: TFE and Local Policies (II) (1) (2) (3) (4) (5) (6) (7) -0.019** (0.008) 0.012*** (0.001) -0.012 (0.008) 0.000 (0.000) -0.003*** (0.001) -0.043** (0.017) -0.000 (0.000) 0.011*** (0.002) 2,035 0.405 0.64 X X 2,035 0.405 0.65 X X (d) County Policies: Mask Mandates total frontier experience -0.036*** (0.008) pop. density -0.025*** (0.008) 0.000*** (0.000) share white -0.028*** (0.008) -0.037*** (0.008) -0.030*** (0.008) -0.004*** (0.001) temperature -0.031 (0.020) median income 0.000*** (0.000) share post-sec. educ Number of Counties Dep. Var. Mean R2 State Fixed Effects Geographic/Agroclimatic Controls 2,035 0.405 0.61 X X 2,035 0.405 0.63 X X 2,035 0.405 0.62 X X 2,035 0.405 0.61 X X 2,035 0.405 0.62 X X Notes: This table reports estimates of equation (2) for local policy outcomes while accounting for potential confounders at the county level. The regressions control for state fixed effects and the following additional predetermined controls: county area; county centroid latitude and longitude; distance to oceans,lakes and rivers from county centroid; mean county rainfall; elevation; and average potential agricultural yield. Standard errors are clustered by arbitrary 60-squaremile grid cells. Standard errors are clustered by arbitrary 60-square-mile grid cells. Significance levels: ∗ : 10% ∗∗ : 5% ∗ ∗ ∗ : 1%. Table A.5: Disentangling Population Density: TFE, Social Distancing, and Mask Use (1) Pop Dens x post-March-13 Pop Dens Decile Within-State x post-March-13 FE Pop Dens Neighbor Matching Within-State x post-March-13 FE Sample Restriction (2) X (3) (4) (5) (6) X None None None X None > 90th ≤ 90th percentile urban pop. share, 2010 Panel (a): Non-Essential Visits (Unacast) TFE × post-March-13 Number of County–Weeks R2 2.645*** (0.619) 1.746*** (0.529) 0.748 (0.560) 0.845* (0.499) 1.191 (1.138) 1.681*** (0.591) 35,828 0.70 35,828 0.70 35,542 0.70 35,750 0.73 3,588 0.69 32,240 0.69 Panel (b): Time in Workplaces (Google Community Mobility) TFE × post-March-13 Number of County–Weeks R2 1.075*** (0.192) 0.787*** (0.151) 0.641*** (0.147) 0.558*** (0.111) 0.336 (0.440) 0.796*** (0.153) 49,938 0.91 49,938 0.91 49,537 0.91 49,852 0.92 4,995 0.96 44,943 0.90 Panel (c): Share who always wear a mask in public when they expect to be within six feet of another person total frontier experience Number of Counties R2 County Fixed Effects Week Fixed Effects -0.019*** (0.003) -0.014*** (0.003) -0.012*** (0.003) -0.010*** (0.003) -0.012 (0.007) -0.013*** (0.003) 2,036 0.54 X X 2,036 0.58 X X 2,021 0.59 X X 2,036 0.82 X X 201 0.59 X X 1,832 0.55 X X Notes: This table disentangles the effects of TFE on social distancing outcomes from the effects of contemporary population density by controlling for the differential effects of population density in several ways. Column 1 reports the baseline estimates with county and week fixed effects. Column 2 adds an interaction of the 2010 population density with postMarch 13 indicator. Column 3 includes fixed effect for the decile of within-state population density interacted with week indicators. Column 4 includes fixed effects indicators within-state pairs of counties that have the most similar population density in 2010 interacted with week indicators. Columns 5 and 6 split the sample into counties above and below the 90th percentile of contemporaneous urban population shares. Standard errors are clustered by arbitrary 60-square-mile grid cells. Column 4 and 5 additionally cluster (two-way) on the population density deciles and within-state county-pairs, respectively. Significance levels: ∗ : 10% ∗∗ : 5% ∗ ∗ ∗ : 1%. Table A.6: Disentangling Population Density: TFE and Local Policies (1) Pop Dens x post-March-13 Pop Dens Decile Within-State x post-March-13 FE Pop Dens Neighbor Matching Within-State x post-March-13 FE Sample Restriction (2) X (3) (4) (5) (6) X None None None X None > 90th ≤ 90th percentile urban pop. share, 2010 (a): Emergency Declaration total frontier experience Number of Counties R2 -0.042*** (0.011) -0.025** (0.011) -0.017 (0.011) -0.022 (0.014) 0.025 (0.035) -0.021* (0.012) 2,036 0.20 2,036 0.24 2,021 0.25 2,036 0.63 201 0.27 1,832 0.20 (b): Stay at Home Policies total frontier experience Number of Counties R2 -0.025*** (0.007) -0.014** (0.006) -0.012* (0.006) -0.012* (0.007) -0.051* (0.029) -0.009 (0.006) 2,036 0.16 2,036 0.23 2,021 0.25 2,036 0.69 201 0.48 1,832 0.13 (c): Business Closure Policies total frontier experience Number of Counties R2 -0.003 (0.003) -0.001 (0.003) -0.001 (0.003) 0.000 (0.004) 0.040 (0.029) -0.003* (0.002) 2,036 0.06 2,036 0.07 2,021 0.06 2,036 0.52 201 0.23 1,832 0.04 (d): Mask Mandates total frontier experience Number of Counties R2 State Fixed Effects Geographic/Agroclimatic Controls -0.036*** (0.008) -0.025*** (0.008) -0.019** (0.008) -0.022** (0.010) -0.037 (0.036) -0.018** (0.008) 2,035 0.61 X X 2,035 0.63 X X 2,020 0.65 X X 2,035 0.83 X X 201 0.44 X X 1,831 0.67 X X Notes: This table disentangles the effects of TFE on local policies from the effects of contemporary population density in several ways. Column 1 reports the baseline with state fixed effects and geographic and agroclimatic controls. Column 2 controls for contemporaneous population density. Column 3 includes fixed effects for the decile of within-state population density. Column 4 included fixed effects for the nearest-neighbor matching based on 2010 population density. Columns 5 and 6 split the sample into counties above and below the 90th percentile of contemporaneous urban population shares. The regressions control for state fixed effects and the following additional predetermined controls: county area; county centroid latitude and longitude; distance to oceans,lakes and rivers from county centroid; mean county rainfall; elevation; and average potential agricultural yield. Standard errors are clustered by arbitrary 60-square-mile grid cells. Standard errors are clustered by arbitrary 60-square-mile grid cells. Column 4 and 5 additionally cluster (two-way) on the population density deciles and within-state county-pairs, respectively. Significance levels: ∗ : 10% ∗∗ : 5% ∗ ∗ ∗ : 1%. Table A.7: Adding West Coast, Extended Time Frame, Regional Heterogeneity TFE, Social Distancing, and Mask Use Frontier Time Frame: Regional Sample Restriction: Baseline (1790–1890) Baseline + West Coast (1) Only Midwest (2) Only South (3) Extended (1790–1950) Only West (4) Extended Sample (5) Only Midwest (6) Only South (7) Only West (8) (a): Non-Essential Visits (Unacast) TFE × post-March-13 1.036** (0.466) 3.338** (1.408) 0.519 (0.409) -1.959* (1.010) 1.149*** (0.413) 4.581*** (1.552) 0.364 (0.380) 0.718 (0.593) Number of County–Weeks R2 38,168 0.04 15,106 0.01 18,460 0.03 3,120 0.07 42,926 0.03 15,340 0.02 20,410 0.03 5,694 0.08 (b): Time in Workplaces (Google Community Mobility) TFE × post-March-13 Number of County–Weeks R2 -4.021*** (0.239) -5.909*** (0.339) -3.056*** (0.265) -3.603*** (0.665) -2.778*** (0.212) -5.212*** (0.423) -2.794*** (0.208) -1.508*** (0.218) 52,615 0.15 23,275 0.19 23,884 0.12 3,757 0.09 59,965 0.12 23,681 0.17 26,696 0.12 7,889 0.10 (c): Share who always wear a mask in public when they expect to be within six feet of another person total frontier experience Number of Counties R2 County or State Fixed Effects Week FE Geographic/Agroclimatic Controls -0.019*** (0.003) -0.031*** (0.006) -0.018*** (0.004) -0.002 (0.009) -0.012*** (0.002) -0.016*** (0.006) -0.015*** (0.003) -0.005* (0.003) 2,141 0.58 X X X 987 0.40 X X X 936 0.44 X X X 152 0.55 X X X 2,499 0.59 X X X 1,037 0.42 X X X 1,074 0.46 X X X 322 0.72 X X X Notes: Focusing on the key social distancing outcomes, this table extends our baseline sample of counties and examines region-by-region sample splits. Column 1 adds counties along the secondary West Coast frontier. Column 2 restricts to counties in the Midwest Census region, column 3 restricts to the South region, and column 4 restricts to the West, which includes the counties added in column 1 plus others in states in the West region but falling inside the 1890 main east-towest frontier line. Column 5 expands the column 1 sample to include counties beyond the (main and secondary) 1890 frontier lines but inside the eventual frontier line realized by 1950. Columns 6–8 then proceed with the same region-byregion sample splits. Standard errors are clustered by arbitrary 60-square-mile grid cells. Significance levels: ∗ : 10% ∗∗ : 5% ∗ ∗ ∗ : 1%. Table A.8: Adding West Coast, Extended Time Frame, Regional Heterogeneity TFE and Local Policies Frontier Time Frame: Regional Sample Restriction: Baseline (1790–1890) Baseline + West Coast (1) Only Midwest (2) Extended (1790–1950) Only South (3) Only West (4) Extended Sample (5) Only Midwest (6) Only South (7) Only West (8) (a): Emergency Declaration total frontier experience Number of Counties R2 -0.039*** (0.011) -0.060*** (0.018) -0.039*** (0.015) 0.023 (0.026) -0.020*** (0.007) -0.041*** (0.014) -0.027*** (0.010) -0.004 (0.010) 2,141 0.26 987 0.10 936 0.19 152 0.45 2,500 0.28 1,038 0.09 1,074 0.19 322 0.42 (b): Stay at Home Policies total frontier experience Number of Counties R2 -0.027*** (0.007) -0.013 (0.010) -0.027*** (0.008) -0.034 (0.023) -0.021*** (0.004) -0.005 (0.009) -0.029*** (0.006) -0.018*** (0.005) 2,141 0.24 987 0.12 936 0.16 152 0.50 2,500 0.22 1,038 0.12 1,074 0.16 322 0.40 (c): Business Closure Policies total frontier experience Number of Counties R2 -0.002 (0.003) -0.002 (0.002) -0.002 (0.004) 0.012 (0.011) -0.004** (0.002) -0.001 (0.002) -0.003 (0.003) -0.004 (0.003) 2,141 0.09 987 0.02 936 0.08 152 0.35 2,500 0.11 1,038 0.02 1,074 0.07 322 0.22 (d): Mask Mandates total frontier experience Number of Counties R2 State Fixed Effects Geographic/Agroclimatic Controls -0.004*** (0.001) -0.000 (0.004) -0.002** (0.001) -0.008 (0.011) -0.003 (0.002) -0.000 (0.004) -0.003** (0.001) 0.000 (0.003) 1,307 0.10 X X 349 0.03 X X 771 0.05 X X 136 0.44 X X 1,494 0.09 X X 351 0.03 X X 827 0.05 X X 265 0.25 X X Notes: Focusing on the local policy outcomes, this table extends our baseline sample of counties and examines region-byregion sample splits. Column 1 adds counties along the secondary West Coast frontier. Column 2 restricts to counties in the Midwest Census region, column 3 restricts to the South region, and column 4 restricts to the West, which includes the counties added in column 1 plus others in states in the West region but falling inside the 1890 main east-to-west frontier line. Column 5 expands the column 1 sample to include counties beyond the (main and secondary) 1890 frontier lines but inside the eventual frontier line realized by 1950. Columns 6–8 then proceed with the same region-by-region sample splits. The regressions control for state fixed effects and the following additional predetermined controls: county area; county centroid latitude and longitude; distance to oceans,lakes and rivers from county centroid; mean county rainfall; elevation; and average potential agricultural yield. Standard errors are clustered by arbitrary 60-square-mile grid cells. Standard errors are clustered by arbitrary 60-square-mile grid cells. Significance levels: ∗ : 10% ∗∗ : 5% ∗ ∗ ∗ : 1%.