sons such as illness makes restart more unlikely, while OTHERREAS, i.e., unspecified
other reasons, is insignificant, thus having no impact. If an argument within the entre-
preneurial team was a reason for business closure, this seems - for logical reasons - to
encourage entrepreneurs to venture anew; the effect of QUARREL is significantly positive.
Table 3: Probit estimation results
Independent variable: Occurrence of restart (yes/no) |
Specification I |
Specification II | ||
Marginal effect a |
P-value | |||
Marginal effect a |
P-value | |||
Main covariates I: Closure reasons b c | ||||
DEBTS |
-0,082 ** |
0,041 |
-0,061 |
0,143 |
NOTWORTH |
0,007 |
0,859 |
-0,001 |
0,988 |
ILLIQUIDITY |
0,012 |
0,723 |
0,027 |
0,463 |
WAGEEMPL |
0,036 |
0,634 |
0,036 |
0,639 |
RETIRE |
-0,247 *** |
0,000 |
-0,254 *** |
0,000 |
PRIVATE |
-0,106 ** |
0,044 |
-0,107 ** |
0,040 |
OTHERREAS |
0,038 |
0,374 |
0,038 |
0,372 |
QUARREL____________________________ |
0,193 *** |
0,000 |
0,192 *** |
0,000 |
Main covariates II: Bearer of financial losses b | ||||
Sharehol |
- |
- |
-0,017 |
0,645 |
BANKS |
- |
- |
-0,112 ** |
0,013 |
Otherstake________________________ |
- |
- |
0,024_________ |
0,573 |
Control variables | ||||
CAPITAL |
0,000 |
0,482 |
0,000 |
0,412 |
CAPITALM |
-0,010 |
0,771 |
-0,008 |
0,830 |
TEAM |
0,134 *** |
0,000 |
0,135 *** |
0,000 |
TEAMM |
0,005 |
0,946 |
-0,005 |
0,941 |
LIMITED |
0,161 *** |
0,000 |
0,158 *** |
0,000 |
SIZE |
0,003 *** |
0,001 |
0,003 *** |
0,001 |
SIZEM |
0,043 |
0,443 |
0,043 |
0,433 |
AGE |
-0,023 *** |
0,001 |
-0,022 *** |
0,001 |
AGEM |
-0,144 |
0,129 |
-0,154 |
0,108 |
EAST |
-0,037 |
0,237 |
-0,035 |
0,264 |
HTIND |
-0,006 |
0,961 |
-0,008 |
0,943 |
OTHERIND |
0,020 |
0,749 |
0,025 |
0,681 |
TECHSERV |
0,230 *** |
0,001 |
0,232 *** |
0,001 |
OTHERSERV |
0,131 * |
0,054 |
0,132 * |
0,051 |
CONSTRUCT |
0,107 ** |
0,034 |
0,111 ** |
0,027 |
TRADE______________________________ |
0,080 *_______ |
0,074 |
0,083 *________ |
0,063 |
# of observations |
853 |
853 | ||
Wald χ2 (df) |
146.820 (20) |
159.330 (27) | ||
Pseudo R2 |
0.161_________ |
0.168_________ |
a Average marginal effect.
b No reference category because multiple statements were possible.
c Actually, there is another closure reason indicator available namely “change into a different self-employment
business”. Since this indicator would completely determine the outcome in the regression, 13 observations
which exclusively refer to this closure reason are not used.
* Significant at the 10 % level. ** Significant at the 5 % level. *** Significant at the 1 % level.
Note: Both indictors variance inflation factor and condition number do not indicate a multicollinearity problem.
Source: Own regression based on the ZEW Entrepreneurship Study and the ZEW Foundation Panel.