WLS RESULTS

The original data set consists of all the EP elections ever held (64). The old member countries have held 5 elections, Spain and Portugal 4 elections and the three newcomers only 2 elections.In practise this means that the old member countries have 2,5 times more weight in the analysis than the newcomers. Basically, this is ok since we are analysing elections, not countries. However, the difference in weights may have some bearing on the empirical results. The results below show that is not the case. The results are from weighted leasts squares (WLS) analysis, where the weights are such that each country has equal weight in the analysis (1 for the old countries, 1.25 for Spain and Portugal, and 2.5 for three latest members).

Summary:

Compulsory voting         ++++++
Simultanous elections     ++++++
Weekend voting            ++++++
Country net payer         -----
Country net receiver      ++
Share of seats            ++
Strict party lists        -
Multiple constituencies   ++
New member country        ++

MODEL 1A


--> REGRESS;Lhs=EPELEC;Rhs=ONE, compulso, arrival, concurr, seatshar, restday...
    ,constitu, netpayer, netrecei
    ;wts=weight$

+-----------------------------------------------------------------------+
| Ordinary    least squares regression    Weighting variable = WEIGHT   |
| Dep. var. = EPELEC   Mean=   58.70166667    , S.D.=   18.59343724     |
| Model size: Observations =      64, Parameters =  10, Deg.Fr.=     54 |
| Residuals:  Sum of squares= 5056.057207    , Std.Dev.=        9.67630 |
| Fit:        R-squared=  .767859, Adjusted R-squared =          .72917 |
| Model test: F[  9,     54] =   19.85,    Prob value =          .00000 |
| Diagnostic: Log-L =   -232.0687, Restricted(b=0) Log-L =    -278.8018 |
|             LogAmemiyaPrCrt.=    4.685, Akaike Info. Crt.=      7.565 |
| Autocorrel: Durbin-Watson Statistic =   1.14714,   Rho =       .42643 |
+-----------------------------------------------------------------------+
+---------+--------------+----------------+--------+---------+----------+
|Variable | Coefficient  | Standard Error |t-ratio |P[|T|>t] | Mean of X|
+---------+--------------+----------------+--------+---------+----------+
 Constant  40.03455424      3.3275405       12.031   .0000
 COMPULSO  28.75725047      3.8572011        7.455   .0000  .24000000
 ARRIVAL   4.117893439      4.0533721        1.016   .3142  .14666667
 CONCURR   9.772786597      3.2656862        2.993   .0042  .25666667
 SEATSHAR  1.111202075      .39714737        2.798   .0071  7.6670991
 RESTDAY   8.202811906      3.5979736        2.280   .0266  .73333333
 LIST     -14.67521185      4.9926436       -2.939   .0048  .40000000
 CONSTITU -3.474726518      4.2181767        -.824   .4137  .26666667
 NETPAYER -5.438879648      3.3966010       -1.601   .1152  .20000000
 NETRECEI  7.509107255      4.3866188        1.712   .0927  .26666667

MODEL 1B


--> REGRESS;Lhs=EPELEC;Rhs=ONE, compulso, arrival, concurr, seatshar, restday...
    ,constitu, netpayer, netrecei, y1999, italcomp,uk, finland, sweden, france
    ;wts=weight$

+-----------------------------------------------------------------------+
| Ordinary    least squares regression    Weighting variable = WEIGHT   |
| Dep. var. = EPELEC   Mean=   58.70166667    , S.D.=   18.59343724     |
| Model size: Observations =      64, Parameters =  16, Deg.Fr.=     48 |
| Residuals:  Sum of squares= 1331.270708    , Std.Dev.=        5.26638 |
| Fit:        R-squared=  .938877, Adjusted R-squared =          .91978 |
| Model test: F[ 15,     48] =   49.15,    Prob value =          .00000 |
| Diagnostic: Log-L =   -189.3662, Restricted(b=0) Log-L =    -278.8018 |
|             LogAmemiyaPrCrt.=    3.546, Akaike Info. Crt.=      6.418 |
| Autocorrel: Durbin-Watson Statistic =   2.05410,   Rho =      -.02705 |
+-----------------------------------------------------------------------+
+---------+--------------+----------------+--------+---------+----------+
|Variable | Coefficient  | Standard Error |t-ratio |P[|T|>t] | Mean of X|
+---------+--------------+----------------+--------+---------+----------+
 Constant  46.26026345      2.5400071       18.213   .0000
 COMPULSO  23.05822641      2.9689112        7.767   .0000  .24000000
 ARRIVAL   4.816880720      2.7359257        1.761   .0847  .14666667
 CONCURR   11.52538956      2.0046825        5.749   .0000  .25666667
 SEATSHAR  .9150792064      .37168490        2.462   .0175  7.6670991
 RESTDAY   11.14183131      2.8702801        3.882   .0003  .73333333
 LIST     -7.649597506      5.7850456       -1.322   .1923  .40000000
 CONSTITU  5.299130667      3.4376645        1.541   .1298  .26666667
 NETPAYER -8.733692955      2.7544506       -3.171   .0026  .20000000
 NETRECEI -5.142435279      4.4026426       -1.168   .2486  .26666667
 Y1999    -6.921439310      1.8470970       -3.747   .0005  .26666667
 ITALCOMP -22.98998666      5.7790996       -3.978   .0002  .40000000E-01
 UK       -25.47044219      6.7856977       -3.754   .0005  .66666667E-01
 FINLAND  -19.40137083      3.3902004       -5.723   .0000  .66666667E-01
 SWEDEN   -10.88205582      4.3245889       -2.516   .0153  .66666667E-01
 FRANCE   -10.45314284      3.9353118       -2.656   .0107  .66666667E-01
 

MODEL 2A

--> REGRESS;Lhs=differen;Rhs=ONE, compulso, arrival, concurr, seatshar, restd...
    ,constitu, netpayer, netrecei
    ;wts=weight$

+-----------------------------------------------------------------------+
| Ordinary    least squares regression    Weighting variable = WEIGHT   |
| Dep. var. = DIFFEREN Mean=  -20.99073333    , S.D.=   15.37090991     |
| Model size: Observations =      64, Parameters =  10, Deg.Fr.=     54 |
| Residuals:  Sum of squares= 3139.816796    , Std.Dev.=        7.62527 |
| Fit:        R-squared=  .789057, Adjusted R-squared =          .75390 |
| Model test: F[  9,     54] =   22.44,    Prob value =          .00000 |
| Diagnostic: Log-L =   -216.8232, Restricted(b=0) Log-L =    -266.6205 |
|             LogAmemiyaPrCrt.=    4.208, Akaike Info. Crt.=      7.088 |
| Autocorrel: Durbin-Watson Statistic =   1.41877,   Rho =       .29061 |
+-----------------------------------------------------------------------+
+---------+--------------+----------------+--------+---------+----------+
|Variable | Coefficient  | Standard Error |t-ratio |P[|T|>t] | Mean of X|
+---------+--------------+----------------+--------+---------+----------+
 Constant -38.50327652      2.6222223      -14.683   .0000
 COMPULSO  14.99815535      3.0396141        4.934   .0000  .24000000
 ARRIVAL   1.646990326      3.1942039         .516   .6082  .14666667
 CONCURR   13.76536362      2.5734789        5.349   .0000  .25666667
 SEATSHAR  .6067475030      .31296650        1.939   .0578  7.6670991
 RESTDAY   10.83976892      2.8353334        3.823   .0003  .73333333
 LIST     -6.993120484      3.9343838       -1.777   .0811  .40000000
 CONSTITU -.3561474328      3.3240759        -.107   .9151  .26666667
 NETPAYER -9.648479284      2.6766445       -3.605   .0007  .20000000
 NETRECEI  8.846538391      3.4568144        2.559   .0133  .26666667

MODEL 2B

--> REGRESS;Lhs=differen;Rhs=ONE, compulso, arrival, concurr, seatshar, restd...
    ,constitu, netpayer, netrecei, y1999,uk, luxembou, italcomp, sweden
    ;wts=weight$

+-----------------------------------------------------------------------+
| Ordinary    least squares regression    Weighting variable = WEIGHT   |
| Dep. var. = DIFFEREN Mean=  -20.99073333    , S.D.=   15.37090991     |
| Model size: Observations =      64, Parameters =  15, Deg.Fr.=     49 |
| Residuals:  Sum of squares= 1167.807475    , Std.Dev.=        4.88189 |
| Fit:        R-squared=  .921543, Adjusted R-squared =          .89913 |
| Model test: F[ 14,     49] =   41.11,    Prob value =          .00000 |
| Diagnostic: Log-L =   -185.1740, Restricted(b=0) Log-L =    -266.6205 |
|             LogAmemiyaPrCrt.=    3.382, Akaike Info. Crt.=      6.255 |
| Autocorrel: Durbin-Watson Statistic =   2.10252,   Rho =      -.05126 |
+-----------------------------------------------------------------------+
+---------+--------------+----------------+--------+---------+----------+
|Variable | Coefficient  | Standard Error |t-ratio |P[|T|>t] | Mean of X|
+---------+--------------+----------------+--------+---------+----------+
 Constant -36.30189097      2.2512930      -16.125   .0000
 COMPULSO  11.64069963      2.7340010        4.258   .0001  .24000000
 ARRIVAL   7.416148732      2.7890859        2.659   .0106  .14666667
 CONCURR   9.108110211      2.0333295        4.479   .0000  .25666667
 SEATSHAR  .4335544342      .34142395        1.270   .2101  7.6670991
 RESTDAY   7.612218153      2.6166946        2.909   .0054  .73333333
 LIST      2.432630093      5.3821971         .452   .6533  .40000000
 CONSTITU  11.86198643      3.5493892        3.342   .0016  .26666667
 NETPAYER -7.515052045      2.2978571       -3.270   .0020  .20000000
 NETRECEI  2.582692085      3.6801433         .702   .4861  .26666667
 Y1999    -3.375698605      1.7615951       -1.916   .0612  .26666667
 UK       -26.86046714      6.2952054       -4.267   .0001  .66666667E-01
 LUXEMBOU  12.46295880      4.4524774        2.799   .0073  .66666667E-01
 ITALCOMP -12.76123230      5.4135253       -2.357   .0224  .40000000E-01
 SWEDEN   -11.48917393      3.7471953       -3.066   .0035  .66666667E-01

MODEL3A

--> REGRESS;Lhs=EPELEC;Rhs=ONE, compulso, arrival, concurr, seatshar, restday...
    ,constitu, netpayer, netrecei, npelec
    ;wts=weight$

+-----------------------------------------------------------------------+
| Ordinary    least squares regression    Weighting variable = WEIGHT   |
| Dep. var. = EPELEC   Mean=   58.70166667    , S.D.=   18.59343724     |
| Model size: Observations =      64, Parameters =  11, Deg.Fr.=     53 |
| Residuals:  Sum of squares= 3138.785917    , Std.Dev.=        7.69561 |
| Fit:        R-squared=  .855887, Adjusted R-squared =          .82870 |
| Model test: F[ 10,     53] =   31.48,    Prob value =          .00000 |
| Diagnostic: Log-L =   -216.8126, Restricted(b=0) Log-L =    -278.8018 |
|             LogAmemiyaPrCrt.=    4.240, Akaike Info. Crt.=      7.119 |
| Autocorrel: Durbin-Watson Statistic =   1.42464,   Rho =       .28768 |
+-----------------------------------------------------------------------+
+---------+--------------+----------------+--------+---------+----------+
|Variable | Coefficient  | Standard Error |t-ratio |P[|T|>t] | Mean of X|
+---------+--------------+----------------+--------+---------+----------+
 Constant -40.36763694      14.376546       -2.808   .0070
 COMPULSO  14.67153680      3.9419627        3.722   .0005  .24000000
 ARRIVAL   1.588335103      3.2541791         .488   .6275  .14666667
 CONCURR   13.86014091      2.6947320        5.143   .0000  .25666667
 SEATSHAR  .5947725717      .32863565        1.810   .0760  7.6670991
 RESTDAY   10.90236600      2.9005540        3.759   .0004  .73333333
 LIST     -6.810760124      4.2043700       -1.620   .1112  .40000000
 CONSTITU -.2821174356      3.4013392        -.083   .9342  .26666667
 NETPAYER -9.748408335      2.8055083       -3.475   .0010  .20000000
 NETRECEI  8.878286831      3.4969898        2.539   .0141  .26666667
 NPELEC    1.023738374      .17992442        5.690   .0000  79.692400

MODEL3B

--> REGRESS;Lhs=EPELEC;Rhs=ONE, compulso, arrival, concurr, seatshar, restday...
    ,constitu, netpayer, netrecei, npelec, y1999,uk, italcomp, luxembou, sweden
    ;wts=weight$

+-----------------------------------------------------------------------+
| Ordinary    least squares regression    Weighting variable = WEIGHT   |
| Dep. var. = EPELEC   Mean=   58.70166667    , S.D.=   18.59343724     |
| Model size: Observations =      64, Parameters =  16, Deg.Fr.=     48 |
| Residuals:  Sum of squares= 1135.852558    , Std.Dev.=        4.86452 |
| Fit:        R-squared=  .947849, Adjusted R-squared =          .93155 |
| Model test: F[ 15,     48] =   58.16,    Prob value =          .00000 |
| Diagnostic: Log-L =   -184.2862, Restricted(b=0) Log-L =    -278.8018 |
|             LogAmemiyaPrCrt.=    3.387, Akaike Info. Crt.=      6.259 |
| Autocorrel: Durbin-Watson Statistic =   2.08448,   Rho =      -.04224 |
+-----------------------------------------------------------------------+
+---------+--------------+----------------+--------+---------+----------+
|Variable | Coefficient  | Standard Error |t-ratio |P[|T|>t] | Mean of X|
+---------+--------------+----------------+--------+---------+----------+
 Constant -25.21363888      9.8020449       -2.572   .0133
 COMPULSO  13.90122309      3.3475033        4.153   .0001  .24000000
 ARRIVAL   7.496478219      2.7800267        2.697   .0096  .14666667
 CONCURR   8.760967485      2.0480026        4.278   .0001  .25666667
 SEATSHAR  .5404496973      .35242648        1.534   .1317  7.6670991
 RESTDAY   7.196146579      2.6318576        2.734   .0087  .73333333
 LIST      1.076440399      5.4885696         .196   .8453  .40000000
 CONSTITU  11.33981511      3.5651975        3.181   .0026  .26666667
 NETPAYER -7.172482290      2.3085846       -3.107   .0032  .20000000
 NETRECEI  2.364444853      3.6718620         .644   .5227  .26666667
 NPELEC    .8591545719      .12120325        7.089   .0000  79.692400
 Y1999    -3.707214370      1.7783619       -2.085   .0424  .26666667
 UK       -27.27501088      6.2829533       -4.341   .0001  .66666667E-01
 ITALCOMP -14.27793022      5.5499262       -2.573   .0132  .40000000E-01
 LUXEMBOU  11.95031472      4.4585218        2.680   .0100  .66666667E-01
 SWEDEN   -10.90890081      3.7671114       -2.896   .0057  .66666667E-01