Announcement

Collapse
No announcement yet.
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • IV regression issues

    I have the following school level data with MSA indicators metaread, and other controls. Here is a snippet:
    Code:
    * dataex year metaread schoolid indep ins ln_fteteachers ln_enroll_total ln_enroll_white ln_cty_pop ln_density e_schoolcommunitytype e_schoolsreligorientation schoollevelcode_e
    * Example generated by -dataex-. To install: ssc install dataex
    clear
    input double year long metaread double schoolid float(indep ins ln_fteteachers ln_enroll_total ln_enroll_white ln_cty_pop ln_density) int(e_schoolcommunitytype e_schoolsreligorientation) long schoollevelcode_e
    2000 8600  2   .6792825 17.453243         .  6.526495  6.198479 12.012943  4.826227   5   2 219
    2010 8600  2   .9151677 17.749481  3.708682  6.434546  6.095825  12.17899  4.992274   3   2 219
    2000 1000  5   .8440945 18.955835  3.135494  6.257668  6.107023 13.403091  6.389827   5   2   4
    2010 1000  5  1.1763699 19.150806   3.08191  6.070738  5.726848 13.397668  6.384404 118   2   4
    2000 2650  6 -.05712979 17.949232  2.646175  5.411646  5.278115 11.384706  4.880871   5   2   4
    2010 2650  6    .433094 18.177612 2.9496884  5.308268  4.962845  11.43722  4.933386   3   2   4
    2000 1000  7   .8440945 18.955835  2.459589  5.283204         . 13.403091  6.389827 119   2   4
    2010 1000  7  1.1763699 19.150806 1.4816046  4.158883         . 13.397668  6.384404 118   2   4
    2000 2030  8   .4822042 16.856903  2.451005  5.327876  5.123964 11.617862  5.255977   5   2   4
    2010 2030  8  1.1685236 17.023989  2.631889  5.030438  4.564348 11.690988  5.329103 277   2   4
    2000 1000 10   .8440945 18.955835  2.587764  5.455321  5.105946 13.403091  6.389827   5   2   4
    2010 1000 10  1.1763699 19.150806 2.5572274   5.31812  5.030438 13.397668  6.384404   3   2   4
    2000 1000 11   .8440945 18.955835  2.424803  5.416101 1.3862944 13.403091  6.389827   5   2   1
    2010 1000 11  1.1763699 19.150806         .         .         . 13.397668  6.384404   1   1   1
    2000 1000 12   .8440945 18.955835 2.6741486  5.389072  5.298317 13.403091  6.389827   5   2   4
    2010 1000 12  1.1763699 19.150806  2.797281  5.247024  5.062595 13.397668  6.384404   3   2   4
    2000 1000 13   .8440945 18.955835  2.424803  5.370638  5.010635 13.403091  6.389827   5   2   4
    2010 1000 13  1.1763699 19.150806 2.2407095 4.6151204  3.135494 13.397668  6.384404   3   2   4
    2000 1000 14   .8440945 18.955835  2.406945  5.347107         . 13.403091  6.389827   5   2   4
    2010 1000 14  1.1763699 19.150806 2.1972246  5.147494         0 13.397668  6.384404   3   2   4
    2000 1000 15   .8440945 18.955835  4.237001  6.844815         . 13.403091  6.389827   5   2 117
    2010 1000 15  1.1763699 19.150806   3.71113   6.39693  6.133398 13.397668  6.384404 118   2 117
    2000 1000 16   .8440945 18.955835 2.6246686  5.129899         . 13.403091  6.389827   5   2 117
    2010 1000 16  1.1763699 19.150806 2.2512918  5.075174         . 13.397668  6.384404   3   2 117
    2000 1000 17   .8440945 18.955835  2.815409  5.676754         . 13.403091  6.389827   5   2   4
    2010 1000 17  1.1763699 19.150806 2.1972246  4.804021         0 13.397668  6.384404   3   2   4
    2000 1000 18   .8440945 18.955835 1.3862944 4.1108737         0 13.403091  6.389827 119   2 219
    2010 1000 18  1.1763699 19.150806         .         .         . 13.397668  6.384404   1   1   1
    2000 1000 19   .8440945 18.955835 2.3025851  5.327876         . 13.403091  6.389827 119   2   4
    2010 1000 19  1.1763699 19.150806  2.424803  4.804021 4.3438053 13.397668  6.384404 277   2   4
    2000  450 20   .4542661 17.462225  2.397895  5.105946  4.859812 11.628475  5.221813   5   2   4
    2010  450 20   .7778646 17.600746 3.2108436   5.42495  5.068904 11.683275  5.276614 277   2 219
    2000 5160 22    .850684 19.189604  2.653242  5.062595         . 12.898828  5.784517 119   2   4
    2010 5160 22   1.045776 19.373346         .         .         . 12.931184  5.816873   1   1   1
    2000 5240 23  .52330875 18.145777 2.8449094  5.181784         .  12.31721  5.652487   5   2 117
    2010 5240 23  1.0506424 18.327234   2.70805  5.117994  .6931472 12.343061  5.678337   3   2 117
    2000 5240 24  .52330875 18.145777  2.980619  5.411646         .  12.31721  5.652487   5   2   4
    2010 5240 24  1.0506424 18.327234         .         .         . 12.343061  5.678337   1   1   1
    2000 5240 25  .52330875 18.145777  3.161247  6.052089  5.966147  12.31721  5.652487   5   2   4
    2010 5240 25  1.0506424 18.327234         .         .         . 12.343061  5.678337   1   1   1
    2000 5240 26  .52330875 18.145777  2.501436  5.081404  4.804021  12.31721  5.652487   5   2   4
    2010 5240 26  1.0506424 18.327234         .         .         . 12.343061  5.678337   1   1   1
    2000 5160 27    .850684 19.189604  3.148453  6.173786  6.025866 12.898828  5.784517   5   2   4
    2010 5160 27   1.045776 19.373346  3.034953  5.929589  5.686975 12.931184  5.816873   3   2   4
    2000 5160 28    .850684 19.189604 2.6246686   5.32301  4.882802 12.898828  5.784517   5   2   4
    2010 5160 28   1.045776 19.373346  3.505557  6.133398  5.645447 12.931184  5.816873   3   2   4
    2000 5160 29    .850684 19.189604  3.499533   6.55108  6.289716 12.898828  5.784517   5   2   4
    2010 5160 29   1.045776 19.373346  3.475067  6.224558  6.001415 12.931184  5.816873   3   2   4
    2000 5160 30    .850684 19.189604  3.054001  6.345636  6.253829 12.898828  5.784517   5   2   4
    2010 5160 30   1.045776 19.373346  3.280911  6.261492  6.146329 12.931184  5.816873   3   2   4
    2000 5160 31    .850684 19.189604 2.3125355  5.159055  5.099866 12.898828  5.784517   5   2   4
    2010 5160 31   1.045776 19.373346         .         .         . 12.931184  5.816873   1   1   1
    2000 5160 32    .850684 19.189604 2.4680996  5.429346         . 12.898828  5.784517   5   2   4
    2010 5160 32   1.045776 19.373346 2.3025851  5.283204 1.0986123 12.931184  5.816873   3   2   4
    2000 5160 33    .850684 19.189604  2.533697  5.463832 4.7004805 12.898828  5.784517   5   2   4
    2010 5160 33   1.045776 19.373346 2.3025851  4.969813 4.0073333 12.931184  5.816873   3   2   4
    2000 5160 34    .850684 19.189604 2.3125355  5.411646  5.236442 12.898828  5.784517   5   2   4
    2010 5160 34   1.045776 19.373346         .         .         . 12.931184  5.816873   1   1   1
    2000 5160 35    .850684 19.189604 2.3608541  5.451038  5.313206 11.852358  4.481004 221   2   4
    2010 5160 35   1.045776 19.373346  2.484907  5.241747  5.003946 12.113217 4.7418633 277   2   4
    2000 5160 36    .850684 19.189604 3.2027464  6.214608  6.186209 11.852358  4.481004 119   2   4
    2010 5160 36   1.045776 19.373346  3.182212   6.33328  6.073044 12.113217 4.7418633 223   2   4
    2000 5160 38    .850684 19.189604  2.525729  5.192957  5.075174 12.898828  5.784517   5   2   4
    2010 5160 38   1.045776 19.373346 2.3025851  5.198497  4.983607 12.931184  5.816873 118   2   4
    2000 5240 39  .52330875 18.145777  2.862201  5.480639  5.370638  12.31721  5.652487   5   2 117
    2010 5240 39  1.0506424 18.327234 4.0621657  6.682108  6.423247 12.343061  5.678337   3   2 219
    2000 5160 40    .850684 19.189604 4.7664385  7.360104   7.26333 12.898828  5.784517   5 116 219
    2010 5160 40   1.045776 19.373346  5.009968  7.247793  7.151485 12.931184  5.816873   3 116 219
    2000 1000 41   .8440945 18.955835 3.1904764     5.826         . 13.403091  6.389827   5 116   4
    2010 1000 41  1.1763699 19.150806  3.273364  5.774551  5.517453 13.397668  6.384404   3 116   4
    2000 1000 44   .8440945 18.955835  1.410987  4.204693  3.218876 13.403091  6.389827   5 116   4
    2010 1000 44  1.1763699 19.150806         .         .         . 13.397668  6.384404   1   1   1
    2000 1000 45   .8440945 18.955835 3.0910425  5.991465  5.799093 13.403091  6.389827 119 116 219
    2010 1000 45  1.1763699 19.150806 2.9704144  5.846439         . 13.397668  6.384404 118 116 219
    2000 1000 47   .8440945 18.955835  2.753661     5.826  5.192957 13.403091  6.389827   5 116 219
    2010 1000 47  1.1763699 19.150806   2.76001     5.826  5.117994 13.397668  6.384404 277 116 219
    2000 5240 48  .52330875 18.145777 2.1972246   5.26269 4.5325994  12.31721  5.652487   5 220   4
    2010 5240 48  1.0506424 18.327234  1.568616  4.912655  3.218876 12.343061  5.678337   3 220   4
    2000 5160 49    .850684 19.189604   1.94591   4.89784   3.89182 12.898828  5.784517 119 116   4
    2010 5160 49   1.045776 19.373346 1.9878744  5.010635  3.871201 12.931184  5.816873 118 116   4
    2000 2650 51 -.05712979 17.949232 4.0707345  6.739336  6.393591 11.384706  4.880871   5 116 219
    2010 2650 51    .433094 18.177612  3.624341  6.419995  6.335054  11.43722  4.933386   3 116 219
    2000 5160 57    .850684 19.189604         .  3.871201         . 12.898828  5.784517   5 116   4
    2010 5160 57   1.045776 19.373346         .         .         . 12.931184  5.816873   1   1   1
    2000 5160 58    .850684 19.189604 1.1631508  4.158883         . 12.898828  5.784517   5 116   4
    2010 5160 58   1.045776 19.373346         .         .         . 12.931184  5.816873   1   1   1
    2000 1000 60   .8440945 18.955835  1.609438  4.356709         . 13.403091  6.389827   5 116   4
    2010 1000 60  1.1763699 19.150806         0  .6931472         . 13.397668  6.384404   3 116   4
    2000 1000 62   .8440945 18.955835  .6931472  3.465736         . 13.403091  6.389827 119 116   4
    2010 1000 62  1.1763699 19.150806         .         .         . 13.397668  6.384404   1   1   1
    2000 1000 63   .8440945 18.955835  1.280934 4.3820267  2.772589 13.403091  6.389827   5 116   4
    2010 1000 63  1.1763699 19.150806         .         .         . 13.397668  6.384404   1   1   1
    2000  450 65   .4542661 17.462225         .         .         . 11.628475  5.221813   1   1   1
    2010  450 65   .7778646 17.600746  3.529297  5.986452  5.780744 11.683275  5.276614 277 220 219
    2000 5160 66    .850684 19.189604   3.73767  6.668228  6.566672 12.898828  5.784517   5 116 219
    2010 5160 66   1.045776 19.373346 3.6712246  6.306275  6.115892 12.931184  5.816873   3 116 219
    2000 1000 67   .8440945 18.955835  2.406945  5.459586   5.26269 13.403091  6.389827   5 116   4
    2010 1000 67  1.1763699 19.150806 2.1972246  4.997212  4.804021 13.397668  6.384404 118 116   4
    2000 1000 68   .8440945 18.955835 2.9231615  4.762174  4.744932 13.403091  6.389827   5 116   4
    2010 1000 68  1.1763699 19.150806  2.631889  4.477337 4.4308167 13.397668  6.384404   3 116   4
    end
    label values e_schoolsreligorientation libmediacentre
    label values e_schoolcommunitytype libmediacentre
    label def libmediacentre 1 ".", modify
    label def libmediacentre 2 "1-Catholic", modify
    label def libmediacentre 116 "2-Other religious", modify
    label def libmediacentre 220 "3-Nonsectarian", modify
    label def libmediacentre 3 "1-City (ulocale = 11 or 12 or 13)", modify
    label def libmediacentre 5 "1-Principle city (locale = 1 or 2)", modify
    label def libmediacentre 118 "2-Suburb (ulocale = 21 or 22 or 23)", modify
    label def libmediacentre 119 "2-Urban fringe large town (locale = 3 or 4 or 5)", modify
    label def libmediacentre 221 "3-Rural or small town (locale = 6 or 7 or 8)", modify
    label def libmediacentre 223 "3-Town (ulocale = 31 or 32 or 33)", modify
    label def libmediacentre 277 "4-Rural (ulocale = 41 or 42 or 43)", modify
    label values schoollevelcode_e schoollevelcode_e
    label def schoollevelcode_e 1 ".", modify
    label def schoollevelcode_e 4 "3-High", modify

    Now, I am running the following code:
    Code:
    egen panelid=group( metaread schoolid)
    xtset panelid year
    xi: ivreg2 ln_enroll_white i.metaread i.year i.metaread*i.year ln_fteteachers ln_cty_pop  i.e_schoolcommunitytype (indep=ins)  if privateindicator==1,first cluster(metaready1) partial( i.metaread i.year i.metaread*i.year ln_fteteachers ln_cty_pop  i.e_schoolcommunitytype)

    What this does is, it drops tons of dummies created due to fixed effects - i.metaread i.year i.metaread*i.year (but not all of any one specific type of fixed effect dummies).

    Then gives me this error:
    Code:
    Warning - endogenous variable(s) collinear with instruments
    Vars now exogenous: indep
    Warning - collinearities detected
    Vars dropped:       _ImetXyea_5000_2010 _ImetXyea_9360_2010 _Ie_schoolc_221
                        _Ie_schoolc_277 ins

    and then finally this:

    Code:
    Error: _ImetXyeaa80_2010 _ImetXyeaa160_2010 _ImetXyeaa200_2010 _ImetXyeaa220_2010 _ImetXyeaa240_2010 _ImetXyeaa280_2010 _ImetXyeaa320_2010 _ImetXyeaa380_20
    > 10 _ImetXyeaa440_2010 _ImetXyeaa450_2010 _ImetXyeaa460_2010 _ImetXyeaa480_2010 _ImetXyeaa520_2010 _ImetXyeaa560_2010 _ImetXyeaa600_2010 _ImetXyeaa640_201
    > 0 _ImetXyeaa680_2010 _ImetXyeaa720_2010 _ImetXyeaa760_2010 _ImetXyeaa840_2010 _ImetXyeaa860_2010 _ImetXyeaa870_2010 _ImetXyeaa880_2010 _ImetXyeaa920_2010
    >  _ImetXyeaa960_2010 _ImetXyeaa1000_2010 _ImetXyeaa1020_2010 _ImetXyeaa1040_2010 _ImetXyeaa1080_2010 _ImetXyeaa1150_2010 _ImetXyeaa1240_2010 _ImetXyeaa126
    > 0_2010 _ImetXyeaa1280_2010 _ImetXyeaa1320_2010 _ImetXyeaa1360_2010 _ImetXyeaa1400_2010 _ImetXyeaa1440_2010 _ImetXyeaa1520_2010 _ImetXyeaa1560_2010 _ImetX
    > yeaa1600_2010 _ImetXyeaa1620_2010 _ImetXyeaa1640_2010 _ImetXyeaa1660_2010 _ImetXyeaa1680_2010 _ImetXyeaa1720_2010 _ImetXyeaa1740_2010 _ImetXyeaa1760_2010
    >  _ImetXyeaa1840_2010 _ImetXyeaa1880_2010 _ImetXyeaa1920_2010 _ImetXyeaa1950_2010 _ImetXyeaa1960_2010 _ImetXyeaa2000_2010 _ImetXyeaa2020_2010 _ImetXyeaa20
    > 30_2010 _ImetXyeaa2040_2010 _ImetXyeaa2080_2010 _ImetXyeaa2120_2010 _ImetXyeaa2160_2010 _ImetXyeaa2240_2010 _ImetXyeaa2290_2010 _ImetXyeaa2320_2010 _Imet
    > Xyeaa2360_2010 _ImetXyeaa2400_2010 _ImetXyeaa2560_2010 _ImetXyeaa2580_2010 _ImetXyeaa2640_2010 _ImetXyeaa2650_2010 _ImetXyeaa2670_2010 _ImetXyeaa2680_201
    > 0 _ImetXyeaa2700_2010 _ImetXyeaa2710_2010 _ImetXyeaa2760_2010 _ImetXyeaa2840_2010 _ImetXyeaa2900_2010 _ImetXyeaa2920_2010 _ImetXyeaa3000_2010 _ImetXyeaa3
    > 060_2010 _ImetXyeaa3080_2010 _ImetXyeaa3120_2010 _ImetXyeaa3160_2010 _ImetXyeaa3180_2010 _ImetXyeaa3200_2010 _ImetXyeaa3240_2010 _ImetXyeaa3290_2010 _Ime
    > tXyeaa3320_2010 _ImetXyeaa3350_2010 _ImetXyeaa3360_2010 _ImetXyeaa3480_2010 _ImetXyeaa3520_2010 _ImetXyeaa3560_2010 _ImetXyeaa3600_2010 _ImetXyeaa3610_20
    > 10 _ImetXyeaa3620_2010 _ImetXyeaa3660_2010 _ImetXyeaa3680_2010 _ImetXyeaa3710_2010 _ImetXyeaa3720_2010 _ImetXyeaa3760_2010 _ImetXyeaa3800_2010 _ImetXyeaa
    > 3810_2010 _ImetXyeaa3840_2010 _ImetXyeaa3880_2010 _ImetXyeaa3920_2010 _ImetXyeaa3980_2010 _ImetXyeaa4000_2010 _ImetXyeaa4040_2010 _ImetXyeaa4100_2010 _Im
    > etXyeaa4120_2010 _ImetXyeaa4280_2010 _ImetXyeaa4320_2010 _ImetXyeaa4360_2010 _ImetXyeaa4400_2010 _ImetXyeaa4420_2010 _ImetXyeaa4480_2010 _ImetXyeaa4520_2
    > 010 _ImetXyeaa4600_2010 _ImetXyeaa4680_2010 _ImetXyeaa4720_2010 _ImetXyeaa4800_2010 _ImetXyeaa4880_2010 _ImetXyeaa4890_2010 _ImetXyeaa4900_2010 _ImetXyea
    > a4920_2010 _ImetXyeaa4940_2010 _ImetXyeaa5000_2010 _ImetXyeaa5080_2010 _ImetXyeaa5120_2010 _ImetXyeaa5160_2010 _ImetXyeaa5170_2010 _ImetXyeaa5190_2010 _I
    > metXyeaa5200_2010 _ImetXyeaa5240_2010 _ImetXyeaa5280_2010 _ImetXyeaa5360_2010 _ImetXyeaa5560_2010 _ImetXyeaa5600_2010 _ImetXyeaa5720_2010 _ImetXyeaa5790_
    > 2010 _ImetXyeaa5800_2010 _ImetXyeaa5880_2010 _ImetXyeaa5910_2010 _ImetXyeaa5920_2010 _ImetXyeaa5960_2010 _ImetXyeaa6080_2010 _ImetXyeaa6120_2010 _ImetXye
    > aa6160_2010 _ImetXyeaa6200_2010 _ImetXyeaa6280_2010 _ImetXyeaa6440_2010 _ImetXyeaa6520_2010 _ImetXyeaa6560_2010 _ImetXyeaa6600_2010 _ImetXyeaa6640_2010 _
    > ImetXyeaa6680_2010 _ImetXyeaa6690_2010 _ImetXyeaa6720_2010 _ImetXyeaa6740_2010 _ImetXyeaa6760_2010 _ImetXyeaa6780_2010 _ImetXyeaa6800_2010 _ImetXyeaa6820
    > _2010 _ImetXyeaa6840_2010 _ImetXyeaa6880_2010 _ImetXyeaa6920_2010 _ImetXyeaa6960_2010 _ImetXyeaa6980_2010 _ImetXyeaa7040_2010 _ImetXyeaa7080_2010 _ImetXy
    > eaa7120_2010 _ImetXyeaa7160_2010 _ImetXyeaa7240_2010 _ImetXyeaa7320_2010 _ImetXyeaa7360_2010 _ImetXyeaa7400_2010 _ImetXyeaa7480_2010 _ImetXyeaa7490_2010 
    > _ImetXyeaa7500_2010 _ImetXyeaa7510_2010 _ImetXyeaa7520_2010 _ImetXyeaa7560_2010 _ImetXyeaa7600_2010 _ImetXyeaa7610_2010 _ImetXyeaa7620_2010 _ImetXyeaa768
    > 0_2010 _ImetXyeaa7800_2010 _ImetXyeaa7840_2010 _ImetXyeaa7880_2010 _ImetXyeaa7920_2010 _ImetXyeaa8050_2010 _ImetXyeaa8120_2010 _ImetXyeaa8160_2010 _ImetX
    > yeaa8200_2010 _ImetXyeaa8280_2010 _ImetXyeaa8320_2010 _ImetXyeaa8400_2010 _ImetXyeaa8480_2010 _ImetXyeaa8520_2010 _ImetXyeaa8560_2010 _ImetXyeaa8600_2010
    >  _ImetXyeaa8640_2010 _ImetXyeaa8680_2010 _ImetXyeaa8760_2010 _ImetXyeaa8780_2010 _ImetXyeaa8800_2010 _ImetXyeaa8840_2010 _ImetXyeaa8920_2010 _ImetXyeaa89
    > 40_2010 _ImetXyeaa8960_2010 _ImetXyeaa9040_2010 _ImetXyeaa9080_2010 _ImetXyeaa9140_2010 _ImetXyeaa9160_2010 _ImetXyeaa9200_2010 _ImetXyeaa9260_2010 _Imet
    > Xyeaa9280_2010 _ImetXyeaa9320_2010 _ImetXyeaa9340_2010 _ImetXyeaa9360_2010 listed in partial() but not in list of regressors.
    invalid syntax
    r(198);

    Eariler than trying ivreg2 I was running the following code:
    Code:
    xi: ivreg ln_enroll_white (indep=ins ) i.metaread i.year i.metaread*i.year ln_density ln_fteteachers ln_cty_pop mfteteacherspubs e_religiousorientation e_schooltype e_schoolcommunitytype if privateindicator==1, first cluster(metaready1)
    It gives me appropriate results, but on the first stage all SE and p-values are returned .

    I really need interaction fixed effects, because the sign on the instrument coefficient turn negative if I take off interaction effects from the code that I have shown in the last.

  • #2
    I'm not sure why you use partial in ivreg2. I suspect something you have in the ivreg2 is resulting in a different sample than in ivreg. Your x's in the two estimates are different. The could cause problems by reducing the sample size or by simply including variables that are colinear. What happens if you run ivreg2 without partial - the ivreg2 documentation suggests "partial option is most useful when using cluster..." and with precisely the same x's as in the ivreg? I wonder about the colinearity of indep and ins in ivreg2 but not apparently in ivreg. If you run ivreg and than regress indep ins if e(sample), what is the r-square like? Are the two really colinear?


    If you get usable results with the ivreg formulation, I'm not sure I would worry about diagnosing what went wrong in ivreg2. By the way, it is easier to work with if you use line continuations /// instead of extremely long lines. Also, in recent Stata versions, you don't need xi to use i.x in a model.

    Comment


    • #3
      Thanks for your reply Phil. I was partial because some fixed effect dummies would drop due to multicollinearity and then first stage rank conditions wont be met. Any ways, that's beside the issue. I have another more basic issues which is driving me crazy.

      Data snippet:
      Code:
      * Example generated by -dataex-. To install: ssc install dataex
      clear
      input float(ln_enroll_total indep ins) double year long metaread float metaready1
       5.872118 -3.891767 17.58969 2000 40 1
       2.484907 -3.891767 17.58969 2000 40 1
       6.393591 -3.891767 17.58969 2000 40 1
       6.326149 -3.891767 17.58969 2000 40 1
       5.786897 -3.891767 17.58969 2000 40 1
       6.683361 -3.891767 17.58969 2000 40 1
       5.537334 -3.891767 17.58969 2000 40 1
       6.514713 -3.891767 17.58969 2000 40 1
       5.993961 -3.891767 17.58969 2000 40 1
       6.481577 -3.891767 17.58969 2000 40 1
       6.033086 -3.891767 17.58969 2000 40 1
       6.042633 -3.891767 17.58969 2000 40 1
       6.599871 -3.891767 17.58969 2000 40 1
       6.120297 -3.891767 17.58969 2000 40 1
        2.70805 -3.891767 17.58969 2000 40 1
       6.045005 -3.891767 17.58969 2000 40 1
        2.70805 -3.891767 17.58969 2000 40 1
       5.924256 -3.891767 17.58969 2000 40 1
       5.517453 -3.891767 17.58969 2000 40 1
       5.811141 -3.891767 17.58969 2000 40 1
       6.413459 -3.891767 17.58969 2000 40 1
       6.135565 -3.891767 17.58969 2000 40 1
         5.9428 -3.891767 17.58969 2000 40 1
        5.47227 -3.891767 17.58969 2000 40 1
       6.196444 -3.891767 17.58969 2000 40 1
       5.247024 -3.891767 17.58969 2000 40 1
       3.295837 -3.891767 17.58969 2000 40 1
       5.697093 -3.891767 17.58969 2000 40 1
       7.876638 -3.891767 17.58969 2000 40 1
       5.379897 -3.891767 17.58969 2000 40 1
       6.011267 -3.891767 17.58969 2000 40 1
       6.269096 -3.891767 17.58969 2000 40 1
         5.9428 -3.891767 17.58969 2000 40 1
       6.142037 -3.891767 17.58969 2000 40 1
       2.833213 -3.891767 17.58969 2000 40 1
       7.778212 -3.891767 17.58969 2000 40 1
       4.859812 -3.891767 17.58969 2000 40 1
       6.028278 -3.891767 17.58969 2000 40 1
       6.001415 -3.891767 17.58969 2000 40 1
      3.8066626 -3.891767 17.58969 2000 40 1
       5.758902 -3.891767 17.58969 2000 40 1
       5.783825 -3.891767 17.58969 2000 40 1
       5.356586 -3.891767 17.58969 2000 40 1
       6.124683 -3.891767 17.58969 2000 40 1
       6.679599 -3.891767 17.58969 2000 40 1
       6.533789 -3.891767 17.58969 2000 40 1
       6.107023 -4.302434 14.78755 2000 80 3
       6.732211 -4.302434 14.78755 2000 80 3
       6.804615 -4.302434 14.78755 2000 80 3
       5.986452 -4.302434 14.78755 2000 80 3
       6.180017 -4.302434 14.78755 2000 80 3
       6.580639 -4.302434 14.78755 2000 80 3
       6.133398 -4.302434 14.78755 2000 80 3
       6.111467 -4.302434 14.78755 2000 80 3
       5.831882 -4.302434 14.78755 2000 80 3
       6.042633 -4.302434 14.78755 2000 80 3
       6.854354 -4.302434 14.78755 2000 80 3
       5.863631 -4.302434 14.78755 2000 80 3
       6.300786 -4.302434 14.78755 2000 80 3
       6.066108 -4.302434 14.78755 2000 80 3
       6.403574 -4.302434 14.78755 2000 80 3
       5.476463 -4.302434 14.78755 2000 80 3
       6.045005 -4.302434 14.78755 2000 80 3
       5.743003 -4.302434 14.78755 2000 80 3
       6.854354 -4.302434 14.78755 2000 80 3
       5.771441 -4.302434 14.78755 2000 80 3
        6.49224 -4.302434 14.78755 2000 80 3
       6.030685 -4.302434 14.78755 2000 80 3
       6.876265 -4.302434 14.78755 2000 80 3
       6.186209 -4.302434 14.78755 2000 80 3
       7.072422 -4.302434 14.78755 2000 80 3
       6.873164 -4.302434 14.78755 2000 80 3
       5.659482 -4.302434 14.78755 2000 80 3
       6.654152 -4.302434 14.78755 2000 80 3
       5.556828 -4.302434 14.78755 2000 80 3
       6.876265 -4.302434 14.78755 2000 80 3
       6.257668 -4.302434 14.78755 2000 80 3
       5.961005 -4.302434 14.78755 2000 80 3
       7.554335 -4.302434 14.78755 2000 80 3
       5.634789 -4.302434 14.78755 2000 80 3
       6.028278 -4.302434 14.78755 2000 80 3
       6.230482 -4.302434 14.78755 2000 80 3
       5.877736 -4.302434 14.78755 2000 80 3
        5.83773 -4.302434 14.78755 2000 80 3
       6.698268 -4.302434 14.78755 2000 80 3
       6.428105 -4.302434 14.78755 2000 80 3
       6.079933 -4.302434 14.78755 2000 80 3
       6.424869 -4.302434 14.78755 2000 80 3
       6.459905 -4.302434 14.78755 2000 80 3
       6.356108 -4.302434 14.78755 2000 80 3
       6.095825 -4.302434 14.78755 2000 80 3
       6.559615 -4.302434 14.78755 2000 80 3
       5.831882 -4.302434 14.78755 2000 80 3
       5.749393 -4.302434 14.78755 2000 80 3
        5.47227 -4.302434 14.78755 2000 80 3
       5.170484 -4.302434 14.78755 2000 80 3
       6.095825 -4.302434 14.78755 2000 80 3
       5.476463 -4.302434 14.78755 2000 80 3
       6.306275 -4.302434 14.78755 2000 80 3
       6.659294 -4.302434 14.78755 2000 80 3
      end


      Instrument comes up strong, positive and significant with nice F-statistic,when I run any of the following:


      Code:
      xi: reg indep ins, cluster(metaready1)
      Code:
      xi: reg indep ins i.metaread, cluster(metaready1)
      Code:
      xi: reg indep ins i.year, cluster(metaready1)
      Code:
      xi: reg indep ins i.metaread i.year, cluster(metaready1)
      Code:
      xi: reg indep ins i.metaread*i.year, cluster(metaready1)

      But the moment I run this with actual dependent variable in there, the standard errors and p-values on everything are "." on the first stage.

      Code:
      xi: ivreg ln_enroll_total (indep=ins) i.metaread*i.year, first cluster(metaready1)


      Why would this happen? Why does this happen generally in STATA?



      Comment

      Working...
      X