I am running the below syntax

Code:
xtabond2 generalcrime l.generalcrime  proactivity feb mar apr may jun jul aug sep oct nov dec, gmm( proactivity generalcrime , lag (2 4)) iv( feb mar apr may jun jul aug sep oct nov dec) nolevel

Code:
Dynamic panel-data estimation, two-step difference GMM
------------------------------------------------------------------------------
Group variable: id                              Number of obs      =     26150
Time variable : week                            Number of groups   =       523
Number of instruments = 305                     Obs per group: min =        50
Wald chi2(13) =    127.65                                      avg =     50.00
Prob > chi2   =     0.000                                      max =        50
------------------------------------------------------------------------------
             |              Corrected
generalcrime |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
generalcrime |
         L1. |   .0084245    .007643     1.10   0.270    -.0065556    .0234045
             |
 proactivity |   .0918542   .0309783     2.97   0.003     .0311378    .1525705
         feb |  -.3357907   .2197143    -1.53   0.126    -.7664227    .0948414
         mar |  -.1039943   .2222107    -0.47   0.640    -.5395193    .3315306
         apr |  -.1241196   .2364084    -0.53   0.600    -.5874715    .3392323
         may |   .3990998   .2487457     1.60   0.109    -.0884328    .8866324
         jun |   .6014529   .2578449     2.33   0.020     .0960862     1.10682
         jul |   .6369464   .2641964     2.41   0.016     .1191309    1.154762
         aug |   .3615521   .2706158     1.34   0.182    -.1688452    .8919494
         sep |  -.0556289   .2523996    -0.22   0.826    -.5503231    .4390653
         oct |  -.2701021   .2545519    -1.06   0.289    -.7690146    .2288105
         nov |  -.7050279   .2654339    -2.66   0.008    -1.225269    -.184787
         dec |   -.536009   .2914149    -1.84   0.066    -1.107172    .0351537
------------------------------------------------------------------------------
Instruments for first differences equation
  Standard
    D.(feb mar apr may jun jul aug sep oct nov dec)
  GMM-type (missing=0, separate instruments for each period unless collapsed)
    L(2/4).(proactivity generalcrime)
------------------------------------------------------------------------------
Arellano-Bond test for AR(1) in first differences: z =  -2.96  Pr > z =  0.003
Arellano-Bond test for AR(2) in first differences: z =  -1.01  Pr > z =  0.312
------------------------------------------------------------------------------
Sargan test of overid. restrictions: chi2(292)  = 558.93  Prob > chi2 =  0.000
  (Not robust, but not weakened by many instruments.)
Hansen test of overid. restrictions: chi2(292)  = 289.84  Prob > chi2 =  0.525
  (Robust, but weakened by many instruments.)

Difference-in-Hansen tests of exogeneity of instrument subsets:
  iv(feb mar apr may jun jul aug sep oct nov dec)
    Hansen test excluding group:     chi2(281)  = 280.43  Prob > chi2 =  0.498
    Difference (null H = exogenous): chi2(11)   =   9.40  Prob > chi2 =  0.585
However, when I change the main IV proactivity to l.proactivity, the direction of the effect changed. What does this mean? Should I use proactivity in level if I'm interested in the total effect proactivity has on crime?

Code:
. xtabond2 generalcrime l.generalcrime  l.proactivity feb mar apr may jun jul aug sep oct nov dec, gmm( proactivity generalcrime , lag (2
>  4)) iv( feb mar apr may jun jul aug sep oct nov dec) nolevel twostep robust
Favoring space over speed. To switch, type or click on mata: mata set matafavor speed, perm.

Dynamic panel-data estimation, two-step difference GMM
------------------------------------------------------------------------------
Group variable: id                              Number of obs      =     26150
Time variable : week                            Number of groups   =       523
Number of instruments = 305                     Obs per group: min =        50
Wald chi2(13) =    121.25                                      avg =     50.00
Prob > chi2   =     0.000                                      max =        50
------------------------------------------------------------------------------
             |              Corrected
generalcrime |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
generalcrime |
         L1. |   .0035523   .0066324     0.54   0.592    -.0094469    .0165516
             |
 proactivity |
         L1. |  -.0302871   .0110876    -2.73   0.006    -.0520183   -.0085559
             |
         feb |  -.3070974   .2207157    -1.39   0.164    -.7396923    .1254975
         mar |  -.0884815   .2224854    -0.40   0.691    -.5245449    .3475818
         apr |  -.1001165   .2368536    -0.42   0.673     -.564341    .3641081
         may |   .3628104    .256478     1.41   0.157    -.1398773    .8654981
         jun |   .5504192   .2667749     2.06   0.039     .0275501    1.073288
         jul |   .5248709   .2752155     1.91   0.057    -.0145415    1.064283
         aug |   .2397443   .2874894     0.83   0.404    -.3237246    .8032131
         sep |  -.2070373   .2700674    -0.77   0.443    -.7363596     .322285
         oct |   -.446874   .2807123    -1.59   0.111    -.9970599     .103312
         nov |   -.907495   .2868705    -3.16   0.002    -1.469751   -.3452392
         dec |   -.714167   .3153677    -2.26   0.024    -1.332276   -.0960577
------------------------------------------------------------------------------
Instruments for first differences equation
  Standard
    D.(feb mar apr may jun jul aug sep oct nov dec)
  GMM-type (missing=0, separate instruments for each period unless collapsed)
    L(2/4).(proactivity generalcrime)
------------------------------------------------------------------------------
Arellano-Bond test for AR(1) in first differences: z =  -2.96  Pr > z =  0.003
Arellano-Bond test for AR(2) in first differences: z =  -1.34  Pr > z =  0.181
------------------------------------------------------------------------------
Sargan test of overid. restrictions: chi2(292)  = 562.59  Prob > chi2 =  0.000
  (Not robust, but not weakened by many instruments.)
Hansen test of overid. restrictions: chi2(292)  = 292.47  Prob > chi2 =  0.481
  (Robust, but weakened by many instruments.)

Difference-in-Hansen tests of exogeneity of instrument subsets:
  iv(feb mar apr may jun jul aug sep oct nov dec)
    Hansen test excluding group:     chi2(281)  = 282.66  Prob > chi2 =  0.461
    Difference (null H = exogenous): chi2(11)   =   9.81  Prob > chi2 =  0.548


.