{smcl}
{com}{sf}{ul off}{txt}{.-}
       log:  {res}E:\intdemdata\JOP Aug 01 replication.smcl
  {txt}log type:  {res}smcl
 {txt}opened on:  {res}21 Mar 2003, 15:28:34
{txt}
{com}. *This log file replicates the results reported in Dan Reiter, "Does Peace Nurture Democracy?"
. *Journal of Politics 63 (August) 2001, 935-948.  Analysis conducted using STATA 7.0.
. *Note that the results produced in Table 1 from p.943 are produced with a data set
. *different from the data set used to produce the results in Table 2 on p.945.
. set mem 70m
{txt}(71680k)

{com}. *First, I reproduce the Table 1 results, which are the factors which hasten the transition of
. *democracies.
. use "E:\intdemdata\dtnew02.dta", clear
{txt}
{com}. stweib midtot loswar dregfrc6 lnrgdp growth islam if year>1959 & year<1993, nohr robust cluster(panid)

         {txt}failure _d:  {res}dtrans
   {txt}analysis time _t:  {res}dtctr
                 {txt}id:  {res}panid


{txt}Fitting constant-only model:

Iteration 0:   log likelihood = {res}-143.78171
{txt}Iteration 1:   log likelihood = {res}-131.50681
{txt}Iteration 2:   log likelihood = {res}-131.00303
{txt}Iteration 3:   log likelihood = {res}-131.00139
{txt}Iteration 4:   log likelihood = {res}-131.00139

{txt}Fitting full model:

Iteration 0:   log likelihood = {res}-131.00139{txt}  
Iteration 1:   log likelihood = {res}-113.23526{txt}  
Iteration 2:   log likelihood = {res} -109.3401{txt}  
Iteration 3:   log likelihood = {res}-109.29224{txt}  
Iteration 4:   log likelihood = {res}-109.29214{txt}  
Iteration 5:   log likelihood = {res}-109.29214{txt}  

Weibull regression -- log relative-hazard form

No. of subjects = {res}         167                     {txt}Number of obs   ={res}      2564
{txt}No. of failures = {res}          39
{txt}Time at risk    = {res}        2564
                                                   {txt}Wald chi2({res}6{txt})    ={res}     48.52
{txt}Log likelihood  =   {res}-109.29214                     {txt}Prob > chi2     ={res}    0.0000

                            {txt}(standard errors adjusted for clustering on panid)
{hline 13}{c TT}{hline 64}
             {c |}               Robust
          _t {c |}      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
{hline 13}{c +}{hline 64}
      midtot {c |}  {res}-.2690823   .2049005    -1.31   0.189    -.6706799    .1325153
      {txt}loswar {c |}  {res} .3817912   1.548391     0.25   0.805    -2.652999    3.416581
    {txt}dregfrc6 {c |}  {res} 4.812115   1.117968     4.30   0.000     2.620937    7.003293
      {txt}lnrgdp {c |}  {res} .4684231   .2147294     2.18   0.029     .0475612    .8892849
      {txt}growth {c |}  {res}-.0254716   .2172975    -0.12   0.907     -.451367    .4004237
       {txt}islam {c |}  {res}-.0004386   .0071002    -0.06   0.951    -.0143547    .0134775
       {txt}_cons {c |}  {res}-7.134767   1.542086    -4.63   0.000     -10.1572   -4.112334
{txt}{hline 13}{c +}{hline 64}
       /ln_p {c |}  {res} -.557496   .1049813    -5.31   0.000{col 58}-.7632556   -.3517363
{txt}{hline 13}{c +}{hline 64}
           p {c |}  {res} .5726412   .0601166{col 58} .4661464    .7034656
         {txt}1/p {c |}  {res} 1.746294   .1833283{col 58} 1.421534    2.145249
{txt}{hline 13}{c BT}{hline 64}

{com}. *That replicates Model 1 of Table 1.
. stweib pbit loswar dregfrc6 lnrgdp growth islam if year>1959 & year<1989, nohr robust cluster(ccode)

         {txt}failure _d:  {res}dtrans
   {txt}analysis time _t:  {res}dtctr
                 {txt}id:  {res}panid


{txt}Fitting constant-only model:

Iteration 0:   log likelihood = {res}-97.899623
{txt}Iteration 1:   log likelihood = {res}-88.703125
{txt}Iteration 2:   log likelihood = {res}-88.317002
{txt}Iteration 3:   log likelihood = {res}-88.314911
{txt}Iteration 4:   log likelihood = {res}-88.314911

{txt}Fitting full model:

Iteration 0:   log likelihood = {res}-88.314911{txt}  
Iteration 1:   log likelihood = {res} -78.96665{txt}  
Iteration 2:   log likelihood = {res}-76.335099{txt}  
Iteration 3:   log likelihood = {res}-76.262878{txt}  
Iteration 4:   log likelihood = {res}-76.262731{txt}  
Iteration 5:   log likelihood = {res}-76.262731{txt}  

Weibull regression -- log relative-hazard form

No. of subjects = {res}         141                     {txt}Number of obs   ={res}      2274
{txt}No. of failures = {res}          23
{txt}Time at risk    = {res}        2274
                                                   {txt}Wald chi2({res}6{txt})    ={res}     43.60
{txt}Log likelihood  =   {res}-76.262731                     {txt}Prob > chi2     ={res}    0.0000

                            {txt}(standard errors adjusted for clustering on ccode)
{hline 13}{c TT}{hline 64}
             {c |}               Robust
          _t {c |}      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
{hline 13}{c +}{hline 64}
        pbit {c |}  {res}-.0630291   .2075476    -0.30   0.761    -.4698149    .3437566
      {txt}loswar {c |}  {res} 1.945241    1.18155     1.65   0.100    -.3705542    4.261037
    {txt}dregfrc6 {c |}  {res} 5.744824   2.322617     2.47   0.013     1.192578    10.29707
      {txt}lnrgdp {c |}  {res} .6558053   .2402742     2.73   0.006     .1848765    1.126734
      {txt}growth {c |}  {res}-.0331497   .4990038    -0.07   0.947    -1.011179    .9448797
       {txt}islam {c |}  {res} .0056607   .0096413     0.59   0.557    -.0132359    .0245573
       {txt}_cons {c |}  {res}-8.612328   1.536652    -5.60   0.000    -11.62411   -5.600546
{txt}{hline 13}{c +}{hline 64}
       /ln_p {c |}  {res}-.9418386   .0963077    -9.78   0.000{col 58}-1.130598    -.753079
{txt}{hline 13}{c +}{hline 64}
           p {c |}  {res} .3899103   .0375514{col 58} .3228401    .4709144
         {txt}1/p {c |}  {res} 2.564693   .2469996{col 58} 2.123528    3.097509
{txt}{hline 13}{c BT}{hline 64}

{com}. *That replicates Model 2 of Table 2.
. stweib pbitot loswar dregfrc6 lnrgdp growth islam if year>1959 & year<1989, nohr robust cluster(ccode)

         {txt}failure _d:  {res}dtrans
   {txt}analysis time _t:  {res}dtctr
                 {txt}id:  {res}panid


{txt}Fitting constant-only model:

Iteration 0:   log likelihood = {res}-97.899623
{txt}Iteration 1:   log likelihood = {res}-88.703125
{txt}Iteration 2:   log likelihood = {res}-88.317002
{txt}Iteration 3:   log likelihood = {res}-88.314911
{txt}Iteration 4:   log likelihood = {res}-88.314911

{txt}Fitting full model:

Iteration 0:   log likelihood = {res}-88.314911{txt}  
Iteration 1:   log likelihood = {res}-78.972983{txt}  
Iteration 2:   log likelihood = {res}-76.359932{txt}  
Iteration 3:   log likelihood = {res}-76.288325{txt}  
Iteration 4:   log likelihood = {res}-76.288178{txt}  
Iteration 5:   log likelihood = {res}-76.288178{txt}  

Weibull regression -- log relative-hazard form

No. of subjects = {res}         141                     {txt}Number of obs   ={res}      2274
{txt}No. of failures = {res}          23
{txt}Time at risk    = {res}        2274
                                                   {txt}Wald chi2({res}6{txt})    ={res}     42.31
{txt}Log likelihood  =   {res}-76.288178                     {txt}Prob > chi2     ={res}    0.0000

                            {txt}(standard errors adjusted for clustering on ccode)
{hline 13}{c TT}{hline 64}
             {c |}               Robust
          _t {c |}      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
{hline 13}{c +}{hline 64}
      pbitot {c |}  {res}-.0284425   .1483874    -0.19   0.848    -.3192766    .2623915
      {txt}loswar {c |}  {res} 1.916897   1.309268     1.46   0.143    -.6492209    4.483015
    {txt}dregfrc6 {c |}  {res} 5.733437   2.319435     2.47   0.013     1.187427    10.27945
      {txt}lnrgdp {c |}  {res} .6634493   .2457706     2.70   0.007     .1817477    1.145151
      {txt}growth {c |}  {res}-.0337533   .4957467    -0.07   0.946    -1.005399    .9378925
       {txt}islam {c |}  {res} .0055676   .0093721     0.59   0.552    -.0128014    .0239366
       {txt}_cons {c |}  {res}-8.679739   1.549911    -5.60   0.000    -11.71751   -5.641968
{txt}{hline 13}{c +}{hline 64}
       /ln_p {c |}  {res}-.9358022   .0946136    -9.89   0.000{col 58}-1.121241    -.750363
{txt}{hline 13}{c +}{hline 64}
           p {c |}  {res}  .392271   .0371142{col 58}  .325875    .4721951
         {txt}1/p {c |}  {res} 2.549258   .2411945{col 58} 2.117769    3.068662
{txt}{hline 13}{c BT}{hline 64}

{com}. *That replicates Model 3 of Table 1.
. stweib iw dregfrc6 lnrgdp growth islam if year>1959 & year<1993, nohr robust cluster(ccode)

         {txt}failure _d:  {res}dtrans
   {txt}analysis time _t:  {res}dtctr
                 {txt}id:  {res}panid


{txt}Fitting constant-only model:

Iteration 0:   log likelihood = {res}-143.78171
{txt}Iteration 1:   log likelihood = {res}-131.50681
{txt}Iteration 2:   log likelihood = {res}-131.00303
{txt}Iteration 3:   log likelihood = {res}-131.00139
{txt}Iteration 4:   log likelihood = {res}-131.00139

{txt}Fitting full model:

Iteration 0:   log likelihood = {res}-131.00139{txt}  
Iteration 1:   log likelihood = {res}-112.65672{txt}  
Iteration 2:   log likelihood = {res}-108.27519{txt}  
Iteration 3:   log likelihood = {res}-107.87288{txt}  
Iteration 4:   log likelihood = {res}-107.80652{txt}  
Iteration 5:   log likelihood = {res}-107.79056{txt}  
Iteration 6:   log likelihood = {res}-107.78696{txt}  
Iteration 7:   log likelihood = {res}-107.78618{txt}  
Iteration 8:   log likelihood = {res}-107.78604{txt}  
Iteration 9:   log likelihood = {res}-107.78603{txt}  
Iteration 10:  log likelihood = {res}-107.78603{txt}  

Weibull regression -- log relative-hazard form

No. of subjects = {res}         167                     {txt}Number of obs   ={res}      2564
{txt}No. of failures = {res}          39
{txt}Time at risk    = {res}        2564
                                                   {txt}Wald chi2({res}5{txt})    ={res}   1299.80
{txt}Log likelihood  =   {res}-107.78603                     {txt}Prob > chi2     ={res}    0.0000

                            {txt}(standard errors adjusted for clustering on ccode)
{hline 13}{c TT}{hline 64}
             {c |}               Robust
          _t {c |}      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
{hline 13}{c +}{hline 64}
          iw {c |}  {res} -15.2468    .534836   -28.51   0.000    -16.29506   -14.19854
    {txt}dregfrc6 {c |}  {res} 5.263675   1.084303     4.85   0.000      3.13848     7.38887
      {txt}lnrgdp {c |}  {res} .4528143   .1952429     2.32   0.020     .0701453    .8354833
      {txt}growth {c |}  {res}-.0111657   .2181786    -0.05   0.959    -.4387878    .4164565
       {txt}islam {c |}  {res} .0020841   .0073036     0.29   0.775    -.0122307     .016399
       {txt}_cons {c |}  {res}-7.205606   1.379117    -5.22   0.000    -9.908626   -4.502586
{txt}{hline 13}{c +}{hline 64}
       /ln_p {c |}  {res}-.5761715   .1022734    -5.63   0.000{col 58}-.7766237   -.3757193
{txt}{hline 13}{c +}{hline 64}
           p {c |}  {res} .5620461   .0574824{col 58} .4599563    .6867951
         {txt}1/p {c |}  {res} 1.779214   .1819663{col 58} 1.456038    2.174119
{txt}{hline 13}{c BT}{hline 64}

{com}. *That replicates Model 4 of Table 1.
. *Now, on to Table 2, which uses a different data set to analyze the factors which
. *hasten democratic breakdown.
. clear
{txt}
{com}. use "E:\intdemdata\dcnew01.dta", clear
{txt}
{com}. stweib  midtot  dregfrc6 lnrgdp growth islam if year>1959 & year<1993, robust cluster(ccode) nohr

         {txt}failure _d:  {res}autrev
   {txt}analysis time _t:  {res}dsctr
                 {txt}id:  {res}panelid


{txt}Fitting constant-only model:

Iteration 0:   log likelihood = {res}-54.304906
{txt}Iteration 1:   log likelihood = {res}-51.384946
{txt}Iteration 2:   log likelihood = {res}-51.339033
{txt}Iteration 3:   log likelihood = {res}-51.339021
{txt}Iteration 4:   log likelihood = {res}-51.339021

{txt}Fitting full model:

Iteration 0:   log likelihood = {res}-51.339021{txt}  
Iteration 1:   log likelihood = {res}-36.201247{txt}  
Iteration 2:   log likelihood = {res}-31.677484{txt}  
Iteration 3:   log likelihood = {res}-31.332918{txt}  
Iteration 4:   log likelihood = {res}-31.332092{txt}  
Iteration 5:   log likelihood = {res}-31.332092{txt}  

Weibull regression -- log relative-hazard form

No. of subjects = {res}          82                     {txt}Number of obs   ={res}      1277
{txt}No. of failures = {res}          16
{txt}Time at risk    = {res}        1277
                                                   {txt}Wald chi2({res}5{txt})    ={res}     48.38
{txt}Log likelihood  =   {res}-31.332092                     {txt}Prob > chi2     ={res}    0.0000

                            {txt}(standard errors adjusted for clustering on ccode)
{hline 13}{c TT}{hline 64}
             {c |}               Robust
          _t {c |}      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
{hline 13}{c +}{hline 64}
      midtot {c |}  {res}-.2490467   .4368666    -0.57   0.569     -1.10529    .6071961
    {txt}dregfrc6 {c |}  {res} 4.073772   3.488994     1.17   0.243    -2.764529    10.91207
      {txt}lnrgdp {c |}  {res}-1.860991   .3713526    -5.01   0.000    -2.588828   -1.133153
      {txt}growth {c |}  {res}-.3851388   1.277004    -0.30   0.763    -2.888021    2.117743
       {txt}islam {c |}  {res} .0184209   .0072628     2.54   0.011      .004186    .0326557
       {txt}_cons {c |}  {res}  8.09606   1.956796     4.14   0.000      4.26081    11.93131
{txt}{hline 13}{c +}{hline 64}
       /ln_p {c |}  {res} .1903058   .1291164     1.47   0.141{col 58}-.0627577    .4433692
{txt}{hline 13}{c +}{hline 64}
           p {c |}  {res} 1.209619   .1561817{col 58}  .939171    1.557947
         {txt}1/p {c |}  {res} .8267063   .1067413{col 58} .6418702    1.064769
{txt}{hline 13}{c BT}{hline 64}

{com}. *That replicates Model 5 of Table 2.
. stweib pbit dregfrc6 lnrgdp growth islam if year>1959 & year<1989, robust cluster(ccode) nohr

         {txt}failure _d:  {res}autrev
   {txt}analysis time _t:  {res}dsctr
                 {txt}id:  {res}panelid


{txt}Fitting constant-only model:

Iteration 0:   log likelihood = {res}-43.063836
{txt}Iteration 1:   log likelihood = {res}-40.678014
{txt}Iteration 2:   log likelihood = {res}-40.643988
{txt}Iteration 3:   log likelihood = {res}-40.643976
{txt}Iteration 4:   log likelihood = {res}-40.643976

{txt}Fitting full model:

Iteration 0:   log likelihood = {res}-40.643976{txt}  
Iteration 1:   log likelihood = {res}-27.741178{txt}  
Iteration 2:   log likelihood = {res}-23.919568{txt}  
Iteration 3:   log likelihood = {res}-23.586742{txt}  
Iteration 4:   log likelihood = {res}-23.585435{txt}  
Iteration 5:   log likelihood = {res}-23.585435{txt}  

Weibull regression -- log relative-hazard form

No. of subjects = {res}          63                     {txt}Number of obs   ={res}      1058
{txt}No. of failures = {res}          13
{txt}Time at risk    = {res}        1058
                                                   {txt}Wald chi2({res}5{txt})    ={res}     60.25
{txt}Log likelihood  =   {res}-23.585435                     {txt}Prob > chi2     ={res}    0.0000

                            {txt}(standard errors adjusted for clustering on ccode)
{hline 13}{c TT}{hline 64}
             {c |}               Robust
          _t {c |}      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
{hline 13}{c +}{hline 64}
        pbit {c |}  {res}-.7582944   .6604546    -1.15   0.251    -2.052762    .5361728
    {txt}dregfrc6 {c |}  {res} 1.731073   3.523074     0.49   0.623    -5.174026    8.636172
      {txt}lnrgdp {c |}  {res}-1.808032   .2632866    -6.87   0.000    -2.324064      -1.292
      {txt}growth {c |}  {res} .0235162   1.073567     0.02   0.983    -2.080636    2.127668
       {txt}islam {c |}  {res} .0191514   .0084254     2.27   0.023     .0026379    .0356649
       {txt}_cons {c |}  {res} 7.760785   1.458071     5.32   0.000     4.903018    10.61855
{txt}{hline 13}{c +}{hline 64}
       /ln_p {c |}  {res}  .306264   .1604111     1.91   0.056{col 58} -.008136    .6206641
{txt}{hline 13}{c +}{hline 64}
           p {c |}  {res} 1.358341    .217893{col 58}  .991897    1.860163
         {txt}1/p {c |}  {res} .7361922   .1180934{col 58} .5375873    1.008169
{txt}{hline 13}{c BT}{hline 64}

{com}. *That replicates Model 6 of Table 2.
. stweib pbitot dregfrc6 lnrgdp growth islam if year>1959 & year<1989, robust cluster(ccode) nohr

         {txt}failure _d:  {res}autrev
   {txt}analysis time _t:  {res}dsctr
                 {txt}id:  {res}panelid


{txt}Fitting constant-only model:

Iteration 0:   log likelihood = {res}-43.063836
{txt}Iteration 1:   log likelihood = {res}-40.678014
{txt}Iteration 2:   log likelihood = {res}-40.643988
{txt}Iteration 3:   log likelihood = {res}-40.643976
{txt}Iteration 4:   log likelihood = {res}-40.643976

{txt}Fitting full model:

Iteration 0:   log likelihood = {res}-40.643976{txt}  
Iteration 1:   log likelihood = {res}-28.423601{txt}  
Iteration 2:   log likelihood = {res}-25.012677{txt}  
Iteration 3:   log likelihood = {res}-24.746118{txt}  
Iteration 4:   log likelihood = {res}-24.745211{txt}  
Iteration 5:   log likelihood = {res}-24.745211{txt}  

Weibull regression -- log relative-hazard form

No. of subjects = {res}          63                     {txt}Number of obs   ={res}      1058
{txt}No. of failures = {res}          13
{txt}Time at risk    = {res}        1058
                                                   {txt}Wald chi2({res}5{txt})    ={res}     37.22
{txt}Log likelihood  =   {res}-24.745211                     {txt}Prob > chi2     ={res}    0.0000

                            {txt}(standard errors adjusted for clustering on ccode)
{hline 13}{c TT}{hline 64}
             {c |}               Robust
          _t {c |}      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
{hline 13}{c +}{hline 64}
      pbitot {c |}  {res}-.0836085   .2557773    -0.33   0.744    -.5849227    .4177057
    {txt}dregfrc6 {c |}  {res}  1.67088   3.344377     0.50   0.617    -4.883978    8.225738
      {txt}lnrgdp {c |}  {res}-1.692386   .3346012    -5.06   0.000    -2.348193    -1.03658
      {txt}growth {c |}  {res}-.0746846   1.151656    -0.06   0.948    -2.331888    2.182519
       {txt}islam {c |}  {res} .0185068   .0087361     2.12   0.034     .0013843    .0356292
       {txt}_cons {c |}  {res} 7.060131   2.035354     3.47   0.001     3.070911    11.04935
{txt}{hline 13}{c +}{hline 64}
       /ln_p {c |}  {res} .2302917   .1641619     1.40   0.161{col 58}-.0914596    .5520431
{txt}{hline 13}{c +}{hline 64}
           p {c |}  {res} 1.258967   .2066744{col 58} .9125982    1.736798
         {txt}1/p {c |}  {res} .7943018   .1303941{col 58} .5757723    1.095773
{txt}{hline 13}{c BT}{hline 64}

{com}. *That replicates Model 7 of Table 2.
. stweib iw dregfrc6 lnrgdp growth islam if year>1959 & year<1993, robust cluster(ccode) nohr

         {txt}failure _d:  {res}autrev
   {txt}analysis time _t:  {res}dsctr
                 {txt}id:  {res}panelid


{txt}Fitting constant-only model:

Iteration 0:   log likelihood = {res}-54.304906
{txt}Iteration 1:   log likelihood = {res}-51.384946
{txt}Iteration 2:   log likelihood = {res}-51.339033
{txt}Iteration 3:   log likelihood = {res}-51.339021
{txt}Iteration 4:   log likelihood = {res}-51.339021

{txt}Fitting full model:

Iteration 0:   log likelihood = {res}-51.339021{txt}  
Iteration 1:   log likelihood = {res}-36.383223{txt}  
Iteration 2:   log likelihood = {res}-31.890671{txt}  
Iteration 3:   log likelihood = {res}-31.485739{txt}  
Iteration 4:   log likelihood = {res}-31.460983{txt}  
Iteration 5:   log likelihood = {res} -31.45641{txt}  
Iteration 6:   log likelihood = {res}-31.455838{txt}  
Iteration 7:   log likelihood = {res}-31.455811{txt}  
Iteration 8:   log likelihood = {res}-31.455809{txt}  

Weibull regression -- log relative-hazard form

No. of subjects = {res}          82                     {txt}Number of obs   ={res}      1277
{txt}No. of failures = {res}          16
{txt}Time at risk    = {res}        1277
                                                   {txt}Wald chi2({res}5{txt})    ={res}   1897.26
{txt}Log likelihood  =   {res}-31.455809                     {txt}Prob > chi2     ={res}    0.0000

                            {txt}(standard errors adjusted for clustering on ccode)
{hline 13}{c TT}{hline 64}
             {c |}               Robust
          _t {c |}      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
{hline 13}{c +}{hline 64}
          iw {c |}  {res}-20.74407   .6538103   -31.73   0.000    -22.02552   -19.46263
    {txt}dregfrc6 {c |}  {res} 4.388851   3.335301     1.32   0.188    -2.148219    10.92592
      {txt}lnrgdp {c |}  {res} -1.89001   .3629545    -5.21   0.000    -2.601387   -1.178632
      {txt}growth {c |}  {res}-.4707213   1.204923    -0.39   0.696    -2.832327    1.890885
       {txt}islam {c |}  {res} .0190177   .0076708     2.48   0.013     .0039832    .0340522
       {txt}_cons {c |}  {res} 8.125682   2.006607     4.05   0.000     4.192804    12.05856
{txt}{hline 13}{c +}{hline 64}
       /ln_p {c |}  {res} .1919317   .1332162     1.44   0.150{col 58}-.0691673    .4530307
{txt}{hline 13}{c +}{hline 64}
           p {c |}  {res} 1.211588   .1614032{col 58} .9331705    1.573073
         {txt}1/p {c |}  {res} .8253633   .1099518{col 58} .6356986    1.071616
{txt}{hline 13}{c BT}{hline 64}

{com}. *That replicates Model 8 from Table 2.
. log close
       {txt}log:  {res}E:\intdemdata\JOP Aug 01 replication.smcl
  {txt}log type:  {res}smcl
 {txt}closed on:  {res}21 Mar 2003, 15:55:18
{txt}{.-}
{smcl}
{txt}{sf}{ul off}