“Fine...I’ll do it myself”: Lessons from self-employment grants in a long recession period

Year	# Grants	# New unlim. liab. firms	# New lim. liab. Firms	Mean amount	Median amount	SD of amount	Total amount
2010	282	6,539	19,074	4,061	3,983	491	1,145,272
2011	770	7,532	19,947	3,851	3,876	744	2,965,571
2012	838	7,372	19,984	2,614	2,467	304	2,190,132
2013	4,409	9,641	30,416	2,074	1,930	1,175	9,145,960
2014	2,938	7,816	22,168	3,321	3,263	1,038	9,757,121
2015	2,649	7,235	21,119	3,398	3,274	643	9,000,336
2016	2,162	7,652	23,337	3,415	3,308	809	7,384,163
2017	2,526	9,905	6,060^*	4,749	4,658	1,269	11,995,409

Note: CES grants, unlimited and limited liability firms database. The monetary units are converted to Euros from Croatian Kuna based on the medium exchange rate on the last day of a particular year given by the Croatian National Bank. The period 2009–2014 was a period of recession in the Republic of Croatia.

* The firm-level database is truncated with March 29, 2017.

CES, Croatian Employment Service.

Table 2

Descriptive statistics of unemployment spells which ended in employment

Individual-level variables	Full sample				Entrepreneurial sample
	Grant		No grant		Grant		No grant
	M	SD	M	SD	M	SD	M	SD
Female	0.426	0.494	0.523	0.499	0.428	0.495	0.458	0.498
*Age (years)*
≤20	0.027	0.163	0.104	0.305	0.027	0.163	0.051	0.221
20–30	0.305	0.46	0.386	0.487	0.302	0.459	0.354	0.478
30–40	0.376	0.484	0.236	0.425	0.377	0.485	0.315	0.464
40–50	0.203	0.402	0.173	0.378	0.204	0.403	0.192	0.394
≥50	0.089	0.284	0.101	0.302	0.089	0.285	0.088	0.283
*Education*
Elementary or less	0.057	0.232	0.152	0.359	0.058	0.233	0.088	0.284
High school	0.685	0.464	0.671	0.47	0.688	0.463	0.708	0.455
University or more	0.258	0.437	0.177	0.382	0.254	0.435	0.204	0.403
*Tenure*
≤2 years	0.098	0.297	0.179	0.383	0.097	0.296	0.128	0.334
2–5 years	0.108	0.311	0.134	0.34	0.107	0.309	0.131	0.337
5–10 years	0.206	0.404	0.179	0.383	0.204	0.403	0.211	0.408
10–20 years	0.365	0.482	0.237	0.425	0.369	0.483	0.315	0.465
≥20 years	0.14	0.347	0.155	0.362	0.142	0.349	0.135	0.342
*Unemployment episode variables Duration of unemployment episode*
≤30 days	0.016	0.126	0.108	0.31	0.016	0.125	0.092	0.289
30–90 days	0.223	0.416	0.252	0.434	0.224	0.417	0.247	0.431
90–180 days	0.294	0.456	0.254	0.435	0.3	0.458	0.263	0.44
180 days–1 year	0.245	0.43	0.228	0.42	0.243	0.429	0.235	0.424
1–2 years	0.146	0.353	0.102	0.302	0.143	0.35	0.11	0.313
2–3 years	0.043	0.203	0.031	0.173	0.043	0.203	0.032	0.175
≥3 years	0.032	0.176	0.025	0.158	0.032	0.175	0.021	0.145
*Status before unemployment*
Employment	0.702	0.457	0.757	0.429	0.704	0.456	0.763	0.425
Education	0.036	0.185	0.083	0.275	0.034	0.182	0.047	0.211
Inactivity	0.241	0.428	0.128	0.334	0.239	0.427	0.17	0.376
Vocational training	0.008	0.092	0.02	0.138	0.009	0.092	0.008	0.09
Other	0.013	0.113	0.013	0.112	0.013	0.115	0.012	0.108
*Employment exit reason*
Not applicable	0.299	0.458	0.245	0.43	0.296	0.457	0.239	0.426
Expiration of contract	0.226	0.418	0.446	0.497	0.225	0.418	0.309	0.462
Expiration of seasonal contract	0.012	0.11	0.068	0.251	0.012	0.11	0.044	0.205
Dismissal (econ., tech., and org. reasons)	0.309	0.462	0.16	0.367	0.311	0.463	0.272	0.445
Dismissal	0.022	0.147	0.012	0.107	0.023	0.149	0.018	0.133
Worker resigned	0.03	0.171	0.016	0.125	0.03	0.171	0.017	0.131
Consensual resignation	0.085	0.279	0.043	0.203	0.084	0.278	0.054	0.227
Firm closure	0.015	0.121	0.009	0.097	0.016	0.124	0.043	0.203
Other	0.003	0.054	0.002	0.043	0.003	0.056	0.003	0.051
*Unemployment exit reason*
Job in-country	0.069	0.254	0.953	0.211	0.001	0.025	0.532	0.499
Job abroad	0	0	0.017	0.128	0	0	0.007	0.085
Opening of limited liability firm	0.337	0.473	0.003	0.056	0.362	0.481	0.135	0.341
Opening of unlimited liability firm	0.593	0.491	0.007	0.084	0.637	0.481	0.31	0.462
Other	0	0.016	0.02	0.139	0	0	0.016	0.126
*Individual-based spell variables*
Unemployment reentry	0.129	0.336	0.661	0.473	0.13	0.336	0.602	0.489
Days to unemployment reentry	976.68	621.54	555.73	681.740	962.59	612.40	699.48	718.49
*Sample size*	15,129		1,514,542		14,086		35,125

[i] Note: Full sample consists of all unemployment episodes from January 1, 2009, until November 23, 2017, and finished in employment. The entrepreneurial sample is a subset of the full sample containing all episodes of individuals who exited unemployment due to entrepreneurial reasons at least once. Year of entry and exit, NACE sectors of entry are omitted for brevity reasons.

Kaplan–Meier estimate of employment survival.

Table 3

Results, unemployment reentry: linear models

	Sample mean	Entrepreneurial sample				N
	Sample mean	(1)	(2)	(3)	(4)	N
Not unemployed after
6 months	0.768	0.197^*** (0.004)	0.191^*** (0.004)	0.023^*** (0.003)	0.020 (0.015)	48,259
12 months	0.652	0.335^*** (0.004)	0.326^*** (0.004)	0.066^*** (0.004)	0.086^*** (0.017)	45,692
18 months	0.577	0.353^*** (0.005)	0.343^*** (0.005)	0.052^*** (0.005)	0.064^*** (0.020)	43,406
24 months	0.515	0.381^*** (0.005)	0.371^*** (0.005)	0.057^*** (0.005)	0.080^*** (0.022)	40,508
36 months	0.416	0.354^*** (0.006)	0.342^*** (0.006)	0.020^*** (0.006)	0.073^*** (0.024)	34,530
Individual-based covariates		No	Yes	Yes	Yes	–
Episode based covariates		No	No	Yes	Yes	–
Individual fixed-effects		No	No	No	Yes	–

Note: Entrepreneurial sample is a subset of a full sample containing all episodes of individuals who exited unemployment due to entrepreneurial reasons at least once. Individual-based covariates are gender, age, education, tenure, all discretized according to Table 2. Episode-based covariates include duration of the episode, reason for unemployment entry, reason for unemployment exit, reason for employment exit, NACE 1-digit industry sector, year of unemployment entry, and year of unemployment exit, all dummies discretized according to Table 2. Standard errors clustered at the individual level are in parentheses.

* p < 0.10.

** p < 0.05.

*** p < 0.01.

Table 4

Results, unemployment reentry: survival models

	Entrepreneurial sample
	COX proportional hazard		Parametric Weibull		COX mixed effects
Coefficient hazard ratio
Survival	−0.443^*** (0.030)	0.642 [0.606, 0.680]	0.501^*** (0.033)	0.640 [0.599, 0.683]	−0.447^*** (0.030)	0.640 [0.603, 0.679]
Observations	49,211		49,211		49,211
Ind. covariates	Yes		Yes		Yes
Epis. covariates	Yes		Yes		Yes

Note: Entrepreneurial sample is a subset of a full sample containing all episodes of individuals who exited unemployment due to entrepreneurial reasons at least once. All estimates include individual-based covariates—gender, age, education, tenure, and episode-based covariates include duration of the episode, reason for unemployment entry, reason for unemployment exit, reason for employment exit, NACE 1-digit industry sector, year of unemployment entry, and year of unemployment exit, all dummies discretized according to Table 2. Parentheses contain robust standard errors, while brackets contain a 95% confidence interval of the hazard ratio.

* p < 0.10.

** p < 0.05.

*** p < 0.01.

Kaplan–Meier estimate of employment survival by legal form.

Heterogeneous effects: time spent at the unemployment office.

Heterogeneous effects: potentially bogus employment.

Kaplan–Meier firm survival estimates by years.

Table 5

Regression results for limited liability firms

	Employees (at least)					N
	Sales	Employment	One	Two	Survival
	(1)	(2)	(3)	(4)	(5)
Time horizon
End of firm registration year (t)	−9,467^*** (1,173)	−0.023 (0.070)	0.332^*** (0.006)	−0.040^*** (0.006)		46,336
1 year later (t+1)	−36,214^*** (5,217)	−0.436^*** (0.097)	0.217^*** (0.007)	−0.074^*** (0.010)	−0.018^*** (0.006)	34,300
2 years later (t+2)	−35,921^*** (5,946)	−0.579^*** (0.131)	0.136^*** (0.010)	−0.045^*** (0.012)	−0.037^*** (0.009)	24,804

Note: Models 1 and 2 are OLS regression estimates, while Models 3–5 are marginal effects from probit regressions. All estimates include NACE 2-digit industry dummies, county of firm headquarters, year of firm registration, type of a firm's legal form, number of months a firm was officially open during the year (all dummies), the value of capital owner registered, and other types of grants as covariates [used in Srhoj et al. (2021, 2019)]. Standard errors clustered at NACE 2-digit level are in the parenthesis. Sample sizes of probit regressions on survival are 37,625 (t+1) and 30,330 (t+2).

* p < 0.10.

** p < 0.05.

*** p < 0.01.

Number of appearances of individuals in the entrepreneurial sample.

Kaplan–Meier estimates of employment survival: full sample.

Table A1

Results, unemployment re-entry: full sample

	Sample mean	Full sample				N
	Sample mean	(1)	(2)	(3)	(4)	N
Not unemployed after
6 months	0.617	0.297^*** (0.002)	0.273^*** (0.002)	0.052^*** (0.003)	0.013 (0.013)	1,497,186
12 months	0.428	0.474^*** (0.003)	0.448^*** (0.003)	0.112^*** (0.004)	0.095^*** (0.015)	1,424,800
18 months	0.371	0.471^*** (0.003)	0.445^*** (0.003)	0.103^*** (0.005)	0.077^*** (0.017)	1,338,826
24 months	0.314	0.490^*** (0.004)	0.464^*** (0.004)	0.111^*** (0.005)	0.104^*** (0.018)	1,255,829
36 months	0.251	0.442^*** (0.005)	0.419^*** (0.005)	0.087^*** (0.006)	0.112^*** (0.020)	1,066,772
Individual-based covariates		No	Yes	Yes	Yes	–
Episode-based covariates		No	No	Yes	Yes	–
Individual fixed-effects		No	No	No	Yes	–

Note: The full sample consists of all unemployment episodes from January 1, 2009, until November 23, 2017, and finishing in employment. Individual-based covariates are gender, age, education, tenure, all discretized according to Table 2. Episode-based covariates include duration of the episode, reason for entering unemployment, reason of exiting unemployment, reason of employment exit, NACE 1-digit industry sector, year of entering unemployment, and year of exiting unemployment, all dummies discretized according to Table 2. Standard errors clustered at the individual level are in the parenthesis.

* p < 0.10.

** p < 0.05.

*** p < 0.01.

Overlap of propensity scores before and after nearest neighbor matching.

Table A2

Nearest neighbor matching and linear model

	Matching			Linear model
	Sample mean	(1)	N	Sample mean	(2)	N
Not employed after
6 months	0.877	0.010^** (0.005)	21,404	0.768	0.023^*** (0.003)	48,259
12 months	0.775	0.055^*** (0.006)	21,404	0.652	0.066^*** (0.004)	45,692
18 months	0.673	0.066^*** (0.007)	21,404	0.577	0.052^*** (0.005)	43,406
24 months	0.576	0.095^*** (0.007)	21,404	0.515	0.057^*** (0.005)	40,508
36 months	0.405	0.058^*** (0.007)	21,404	0.416	0.020^*** (0.006)	34,530

Notes: Estimates in the column “Matching” represent the average treatment effect on the treated obtained by the nearest neighbor matching. The selection equation is solved with a probit model containing covariates as in Table 3. Individual-based covariates are gender, age, education, tenure, all discretized according to Table 2. Episode-based covariates include duration of the episode, reason of unemployment entry, reason of unemployment exit, reason of employment exit, NACE 1-digit industry sector, year of unemployment entry, and year of unemployment exit. “Linear model” estimates are obtained from Model 3 given in Table 3.

* p < 0.10.

** p < 0.05.

*** p < 0.01.

Table A3

Results, unemployment re-entry: episodes by individuals with at least two appearances in the dataset

	Sample mean	(1)	(2)	(3)	(4)	N
Panel A: Full sample
Not unemployed after
6 months	0.550	0.348^*** (0.004)	0.319^*** (0.004)	0.087^*** (0.006)	0.013 (0.012)	1,127,780
12 months	0.325	0.541^*** (0.005)	0.513^*** (0.005)	0.166^*** (0.007)	0.095^*** (0.013)	1,073,749
18 months	0.262	0.497^*** (0.006)	0.471^*** (0.006)	0.135^*** (0.009)	0.077^*** (0.015)	1,011,130
24 months	0.200	0.494^*** (0.007)	0.471^*** (0.007)	0.143^*** (0.010)	0.104^*** (0.016)	946,819
36 months	0.135	0.393^*** (0.009)	0.375^*** (0.009)	0.111^*** (0.011)	0.112^*** (0.018)	802,001
Individual-based covariates		No	Yes	Yes	Yes	–
Episode-based covariates		No	No	Yes	Yes	–
Individual fixed-effects		No	No	No	Yes	–
Panel B: Entrepreneurial sample
Not unemployed after
6 months	0.671	0.269^*** (0.006)	0.261^*** (0.006)	0.065^*** (0.006)	0.020^* (0.012)	30,436
12 months	0.503	0.424^*** (0.006)	0.417^*** (0.006)	0.132^*** (0.007)	0.086^*** (0.014)	29,021
18 months	0.407	0.410^*** (0.007)	0.401^*** (0.008)	0.100^*** (0.009)	0.064^*** (0.016)	27,565
24 months	0.330	0.421^*** (0.008)	0.414^*** (0.008)	0.107^*** (0.010)	0.080^*** (0.017)	25,855
36 months	0.217	0.348^*** (0.010)	0.343^*** (0.010)	0.076^*** (0.012)	0.073^*** (0.019)	22,228
Individual-based covariates		No	Yes	Yes	Yes	–
Episode-based covariates		No	No	Yes	Yes	–
Individual fixed-effects		No	No	No	Yes	–

Note: The full sample consists of all unemployment episodes from January 1, 2009, until November 23, 2017, and finishing in employment. The entrepreneurial sample is a subset of a full sample containing all episodes of individuals who exited unemployment due to entrepreneurial reasons at least once. Individual-based covariates are gender, age, education, tenure, all discretized according to Table 2. Episode-based covariates include duration of the episode, reason of entering unemployment, reason of exiting unemployment, reason of employment exit, NACE 1-digit industry sector, year of entering unemployment, and year of exiting unemployment, all dummies discretized according to Table 2. Standard errors clustered at the individual level are in the parenthesis.

* p < 0.10.

** p < 0.05.

*** p < 0.01.

Table A4

Heterogeneous effects: COX PH model with the most comprehensive specification

	Full sample			Entrepreneurial sample
	Hazard ratio	Hazard ratio low	Hazard ratio high	Hazard ratio	Hazard ratio low	Hazard ratio high
	(1)	(2)	(3)	(4)	(5)	(6)
Age (years)
≤20	0.645	0.518	0.803	0.870	0.672	1.127
20–30	0.426	0.389	0.466	0.611	0.550	0.679
30–40	0.434	0.397	0.475	0.662	0.596	0.737
40–50	0.450	0.405	0.501	0.594	0.524	0.673
≥50	0.533	0.459	0.619	0.718	0.599	0.860
Gender
Female	0.474	0.440	0.510	0.670	0.615	0.730
Male	0.438	0.409	0.469	0.621	0.573	0.673
Education
Elementary or less	0.608	0.515	0.717	0.742	0.613	0.899
High school	0.450	0.424	0.477	0.622	0.581	0.666
University or more	0.399	0.354	0.449	0.655	0.568	0.755
Potential tenure
≤2 years	0.431	0.368	0.504	0.635	0.529	0.764
2–5 years	0.430	0.369	0.502	0.654	0.545	0.783
5–10 years	0.429	0.383	0.480	0.651	0.569	0.744
10–20 years	0.431	0.394	0.471	0.604	0.544	0.670
≥20 years	0.510	0.454	0.573	0.670	0.584	0.769
Not applicable	0.532	0.454	0.624	0.751	0.620	0.909
Unemployment duration
≤30 days	0.150	0.071	0.315	0.260	0.122	0.553
30–90 days	0.356	0.314	0.404	0.514	0.444	0.594
90–180 days	0.382	0.342	0.426	0.513	0.452	0.581
180 days–1 year	0.455	0.415	0.499	0.649	0.582	0.725
1–2 years	0.551	0.494	0.614	0.796	0.696	0.911
2–3 years	0.636	0.523	0.773	0.853	0.665	1.095
≥3 years	0.576	0.443	0.748	0.820	0.574	1.173
Status before unemployment
Working	0.446	0.419	0.473	0.641	0.597	0.688
Education	0.613	0.507	0.741	0.853	0.674	1.078
Inactivity	0.400	0.361	0.443	0.583	0.515	0.660
Vocational training	0.592	0.324	1.083	1.390	0.592	3.260
Other	0.416	0.269	0.641	0.423	0.260	0.688
Employment exit reason
Not applicable	0.440	0.403	0.480	0.614	0.553	0.682
Expiration of contract	0.472	0.427	0.521	0.659	0.586	0.741
Expiration of seasonal contract	0.458	0.327	0.640	0.703	0.480	1.029
Dismissal (econ., tech., and org. reasons)	0.501	0.456	0.551	0.684	0.612	0.764
Dismissal	0.687	0.486	0.971	0.989	0.640	1.529
Worker resigned	0.407	0.286	0.580	0.539	0.343	0.845
Consensual resignation	0.339	0.271	0.424	0.516	0.395	0.674
Firm closure	0.358	0.249	0.515	0.366	0.246	0.545
Other	0.387	0.106	1.414	0.358	0.022	5.809
Year of exit
2010	0.535	0.415	0.691	1.079	0.780	1.494
2011	0.606	0.527	0.697	0.847	0.717	1.000
2012	0.500	0.430	0.581	0.692	0.582	0.822
2013	0.461	0.422	0.503	0.705	0.628	0.791
2014	0.508	0.448	0.577	0.691	0.598	0.798
2015	0.451	0.386	0.526	0.577	0.483	0.688
2016	0.189	0.144	0.248	0.292	0.217	0.393
2017	0.023	0.006	0.091	0.037	0.009	0.151

Note: The full sample consists of all unemployment episodes from January 1, 2009, until November 23, 2017, and finishing in employment. The entrepreneurial sample is a subset of a full sample containing all episodes of individuals who exited unemployment due to entrepreneurial reasons at least once. All estimates include individual-based covariates are gender, age, education, tenure; and episode-based covariates include duration of the episode, reason of entering unemployment, reason exiting unemployment, reason of employment exit, NACE 1-digit industry sector, year of entering unemployment, and year of exiting unemployment, all dummies discretized according to Table 2 (excluding a covariate we use to partition the dataset). Hazard ratio low and hazard ratio high represents the 95% confidence interval.

* p < 0.10

** p < 0.05

*** p < 0.01

Kaplan–Meier estimates of employment survival by years.

Table A5

OLS and Cox proportional hazard estimates of firm survival (unlimited liability firms)

Dependent variable (method)
	Firms survive (OLS)					Survival (COX PH)
	6 months	12 months	18 months	24 months	36 months	Coefficient	Hazard ratio
Grant received
	0.089^*** (0.006)	0.148^*** (0.007)	0.069^*** (0.009)	0.049^*** (0.010)	0.015 (0.013)	−0.272^*** (0.021)	0.762 [0.696, 0.835]
Observations	61,241	56,384	51,758	48,029	40,997	61,953
R²	0.030	0.045	0.036	0.042	0.047	0.054

Note: The first five columns present the OLS estimate of the received grant effect on the corresponding indicator of firm survival (with an associated sample truncation). The sixth and seventh columns present results of the estimation of the Cox proportional hazard model of the received grant effect on survival probability (the sixth column is coefficient, while the seventh is hazard ratio). All estimates include dummies for the NACE 2-digit industry sector of the firm, dummies for the county of the firm's headquarters, and dummies for the year of the firm opening. Parentheses contain standard errors clustered at the NACE 2-digit level, while brackets contain a 95% confidence interval of hazard ratio (last column).

* p < 0.10.

** p < 0.05.

*** p < 0.01.

Table A6

Descriptive statistics of limited liability firms

	Grant			No grant
	M	Median	SD	M	Median	SD
Registered capita (in euro)	1,444	1.333	7,178	6,166	2,667	29,202
Sales (in euro)
t	13,225	5,245	27,698	26,134	5,383	56,868
t+1	35,641	16,234	71,016	103,409	24,615	522,825
t+2	45,859	19,620	108,306	137,826	28,229	816,707
Number of employees
t	1.321	1.000	1.230	1.341	1.000	2.689
t+1	1.529	1.000	1.629	2.201	1.000	6.154
t+2	1.633	1.000	1.784	2.618	1.000	9.596
At least one employee
t		0.917			0.542
t+1		0.906			0.649
t+2		0.843			0.664
At least two employees
t		0.234			0.259
t+1		0.301			0.369
t+2		0.363			0.406
Firm survives
t+1		0.883			0.915
t+2		0.759			0.824
Year of firm registration
2010		0.005			0.106
2011		0.023			0.108
2012		0.033			0.115
2013		0.369			0.171
2014		0.137			0.164
2015		0.223			0.151
2016		0.210			0.186
Sample size		4,154			42,182

[i] Note: The sample consists of all new limited liability firms registered from January 1, 2010, until December 31, 2016. For brevity, we do not show NACE 2-digit industry dummies, dummies for the county of the firm's headquarters, dummies for types of firm legal forms, and dummies for the number of months that a firm was officially open.

References

Authors

Metrics

Articles in this issue

DOI: https://doi.org/10.2478/izajolp-2021-0006 | Journal eISSN: 2193-9004

Journal RSS Feed

Language: English

Accepted on: Apr 23, 2021

Published on: Jun 19, 2021

Published by: Sciendo

In partnership with: Paradigm Publishing Services

Keywords: