Rdd with categorical outcome variable

[Dayla Spassio]

HI everyone,
do you think it is possible to use the rdrobust command to perform an RDD but with the categorical dependent variable ordered? (my outcome variable takes values 0, 1, 2, 3, 4)

Thanks

[Jared Greathouse]

Mechanically it's likely possible, but it isn't an approach I endorse, necessarily.

[Dayla Spassio]

ok, thanks

[Dayla Spassio]

and what method would you use?

[Jared Greathouse]

I usually work with continuous data. Presumably there are good causal methods for categorical data, but I don't know of them.

[Cat Santos]

I have a similar question. Is it possible to run an RDD with a categorical ordinal dependent variable? My variable represents democratic preferences (+100 highest; -100 lowest). I was thinking of transforming it into a binary variable equal to 1 if democracy2>=50 and equal to 0 if democracy2<=-50. That is because I believe the values around 0 are more similar between themselves than the values on the extremes. Would this make sense?

Or should I simply just use another variable that is binary in nature, so that the survey responses between 0 and 1 are substantially different? I fear that in surveys it's hard to accuratey position onseself in a scale from -100 to 100, versus giving a binary response, particularly regarding democratic preferences.

Thank you!

 tab democracy2

 democracy2 |      Freq.     Percent        Cum.
------------+-----------------------------------
       -100 |        397        3.87        3.87
        -90 |        149        1.45        5.32
        -80 |        222        2.16        7.49
        -70 |        223        2.17        9.66
        -60 |        209        2.04       11.70
        -50 |        716        6.98       18.68
        -40 |        338        3.30       21.98
        -30 |        427        4.16       26.14
        -20 |        505        4.92       31.07
        -10 |        425        4.14       35.21
          0 |      1,277       12.45       47.66
         10 |        733        7.15       54.81
         20 |        884        8.62       63.43
         30 |        936        9.13       72.56
         40 |        676        6.59       79.15
         50 |      1,115       10.87       90.02
         60 |        375        3.66       93.68
         70 |        315        3.07       96.75
         80 |        218        2.13       98.88
         90 |         46        0.45       99.33
        100 |         69        0.67      100.00
------------+-----------------------------------
      Total |     10,255      100.00

.