Why is it when I enter the following command:

"forvalues i = 1/5 {
regress edge percentFRL ipr_se_variation popdensity if schoolsize== `i'
eststo schoolsize`i'
}

esttab _all using schoolsize.csv"

that I am given a spreadsheet that includes a column (the first column) for the regression using all the observations instead of just divided up by schoolsize? Is there a way to prevent that? It's not a big deal as I can just delete the column, but I don't remember it doing this in the past.
(1) (2) (3) (4) (5) (6)
edge edge edge edge edge edge
percentFRL 0.618*** 0.385*** 0.591*** 0.636*** 0.675*** 0.578***
(277.02) (48.98) (90.98) (162.07) (183.46) (84.89)
ipr_se_variation 0.449*** 0.663*** 0.581*** 0.444*** 0.304*** 0.463***
(81.62) (35.99) (39.44) (46.75) (31.67) (28.72)
popdensity 0.000000196 0.000000485 0.00000102*** 0.00000141*** -0.00000122***
(0.24) (1.82) (5.90) (6.14) (-3.56)
_cons -0.184*** -0.108*** -0.184*** -0.192*** -0.188*** -0.181***
(-109.15) (-17.07) (-39.19) (-65.53) (-66.27) (-38.12)
N 173507 14666 24155 60282 58238 16139
t statistics in parentheses
0 ** p<0.01 *** p<0.001"