Four Essays on University Economics - KOPS - Universität Konstanz
Four Essays on University Economics - KOPS - Universität Konstanz
Four Essays on University Economics - KOPS - Universität Konstanz
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
Return of foreign students<br />
Table 3.4 we summarize which individual and cultural variables have significant effects <strong>on</strong><br />
the return decisi<strong>on</strong>.<br />
The baseline hazard is initially decreasing. Graduates who spent more time in Germany<br />
after they finished their studies have a lower hazard rate: Their propensity to return in<br />
a given m<strong>on</strong>th is lower. The baseline hazard reaches a minimum at 56 m<strong>on</strong>ths after<br />
graduati<strong>on</strong>, then rises and peaks at 110 m<strong>on</strong>ths after graduati<strong>on</strong>. 10<br />
Graduates who came to Germany at an older age and graduates who visited home<br />
during their study period in Germany have a higher hazard rate. In a given m<strong>on</strong>th, they<br />
are more likely to return and are, thus, expected to return earlier than their peers. Time<br />
spent in Germany has a significantly negative but n<strong>on</strong>-linear effect <strong>on</strong> the return hazard.<br />
Students having spent more time in Germany are presumably better integrated and less<br />
likely to return. But the reducti<strong>on</strong> in the hazard rate becomes smaller with increasing<br />
time spent in Germany. 11 Students of agricultural sciences, of German and of medicine<br />
have a significantly higher hazard rate than students of other disciplines.<br />
The propensity to return is significantly higher for participants in the two invitati<strong>on</strong><br />
programs (Program 1 and Program 3) than for participants in Program 2 who were re-<br />
cruited in Germany. The hazard rate of catholic students is not significantly different<br />
from n<strong>on</strong>-catholic students. However, graduates from countries with a higher proporti<strong>on</strong><br />
of catholics in the populati<strong>on</strong> have a significantly lower hazard rate.<br />
Also cultural proximity has significant effects <strong>on</strong> the return decisi<strong>on</strong>. Students from<br />
countries in which a similar language is spoken have a significantly lower hazard rate.<br />
They have a higher propensity to stay and they can be expected to stay for a l<strong>on</strong>ger time.<br />
Graduates from Africa and from the Middle East are less likely to return than their peers<br />
from Latin America.<br />
Students from wealthier countries have a higher hazard rate and are, thus, more likely to<br />
return. In c<strong>on</strong>trast to our expectati<strong>on</strong>, graduates from countries with closer trade relati<strong>on</strong>s<br />
with Germany have a higher propensity to stay. If their home countries are important<br />
10 If the polynomial of the baseline hazard is: c1t + c2t 2 + c3t 3 the formula for the computati<strong>on</strong> of the<br />
trough and peak is: −c2± √ c2 2−3c1c3 .<br />
3c3<br />
√ 0.0025 2 −3(−0.1862)(−0.00001)<br />
Here: times of trough and peak = −0.0025±<br />
.<br />
3(−0.00001)<br />
11<br />
Only for students who spent more than 147.5 m<strong>on</strong>ths (computati<strong>on</strong>: |−0.0295/(2∗0.0001)|) in Germany<br />
at the end of the study period, the propensity to return starts to increase with time. Nobody stayed<br />
in Germany for such a l<strong>on</strong>g time.<br />
65