Comparative Analysis of Hedonic Wage and Discrete Choice Models in Valuing Job Safety

Category	Variable	Definition	Mean	Min	Max
Individual (Obs. 4035)	wage	Yearly wage in 2014 value (unit: 1000 TWD)	537. 2	28.5	5587.7
	eduyear	Education years	14.1	0.0	22.0
	feduyear	Father’s education years	9.2	0.0	22.0
	female	1 if female, 0 otherwise	0.4	0.0	1.0
	age	age	35.1	25.0	79.0
	marriage	1 if married, 0 otherwise	0.5	0.0	1.0
	scale	Number of employees in the firm	212.2	2.0	500.0
	wexp	Working experience years	13.6	1.0	71.0
	healthlevel	1 for very good health, 5 for very bad health	3.6	1.0	5.0
Industry (seven industries)	risk	industry mortality rates, unit: 1/1,000,000	23.6	0.0	124.3
	workhour_alt	industry average monthly total work hours, unit: hour	169.8	159.8	176.0
	unem_alt	industry unemployment rates, unit: %	3.2	2.6	4.6
Town (218 towns)	TJAN	township-level average January temperature (Celsius) from 1981 to 2005	16.4	13.1	20.8
Town (218 towns)	TJUL	township-level average July temperature (Celsius) from 1981 to 2005	28.8	26.4	30.3
Regional dummy	NORTH	1 if respondent lives in Northern Taiwan, 0 otherwise	0.5	0.0	1.0
	CENTER	1 if respondent lives in central Taiwan, 0 otherwise	0.2	0.0	1.0
	SOUTH	1 if respondent lives in southern Taiwan, 0 otherwise	0.2	0.0	1.0
	EAST	1 if respondent lives in eastern Taiwan, 0 otherwise	0.0	0.0	1.0

[i] Source: PSFD 2016, 2018; Bureau of Labour Insurance, Ministry of Labour. Taiwan Climate Change Projection and Information Platform (TCCIP) project

Table 2.

Regression Results-Discrete Choice Model

	CL		ML		PML
	Model 1	Model 2	Model 3	Model 4	Model 5	Model 6
wage_alt	0.0017^***	0.0013^***	0.0022^***	0.0018^***	0.0018^***	0.0016^***
risk_alt	–0.1091^***	–0.0986^***	–0.1244^***	–0.1162^***	–0.1632^***	–0.1422^***
workhour_alt	0.0251^**	0.0096	0.1633^***	0.1039^***	0.1529^***	0.1085^***
unem_alt	−0.7097^***	−0.5777^***	0.3434	0.1138	−0.1729	−0.2892
sd(risk_alt)			0.3826^***	0.3186^***	0.2546^***	0.2323^***
sd(workhour_alt)			0.4681^***	0.4270^***	0.4601^***	0.4771^***
sd(umem_alt)			0.0026	0.0041	2.3677^***	2.7743^***
Construction
eduyear		−0.2003^***		−1.0500^**		−1.0161^***
age		−0.0134		0.0941		0.1238^**
female		−1.0217^***		−15.3042^***		−11.4380^***
_cons	9.0968^***	11.2592^***	−42.8469^***	−19.2340	−19.0045^***	−5.0814
Catering
eduyear		−0.1084^***		−0.1538^*		−0.1549^**
age		−0.0296^***		−0.0946^***		−0.0997^***
female		0.6325^***		3.2816^***		3.2010^***
_cons	−2.0309^***	0.7968	−1.3451^***	2.2099	−1.3025^***	2.5388^*
Transportation
eduyear		−0.0522		−0.1804		−0.1984^***
age		−0.0162		−0.0123		0.0004
female		−0.3839^*		−3.2871^***		−2.4575^***
_cons	−2.9876	1.8339^**	−8.5789^***	−2.9876	−5.4116^***	0.1871
Finance
eduyear		0.2072^***		0.2016^**		0.2588^***
age		0.0058		−0.0370		−0.0454^**
female		0.6392^***		3.1610^***		2.9912^***
_cons	−3.7770^***	−7.1861^***	−8.5789^***	−9.0142^***		−8.6711^***
Science
eduyear		0.0321		0.0385		0.0385
age		0.0105		−0.0011		0.0315
female		0.4214^***		0.6866^***		1.1440^***
_cons	−0.6070^***	−1.8486^***	−1.9667^***	−2.4552^***	−2.9669^***	−5.4924^***
Service
eduyear		0.1624^***		0.1287		0.2505^***
age		−0.0345^***		−0.1047^***		−0.1237^***
female		1.1372^***		3.9960^***		4.1265^***
_cons	−2.2029^***	−4.0223^***	−2.1834^***	−2.8356	−2.4155^***	−4.1438^**
N	32,347	32,347	32,347	32,347	32,347	32,347
Pseudo R2	1.41%	1.32%	30.93%	31.52%	27.36%	28.19%
Log-likelihood	−7388.89	−7085.43	−7127.41	−6923.56	−6107.19	−5918.74
AIC	14797.78	14226.87	14280.81	13909.11	12750.25	11899.47

Note:

* p < 0.1,

** p < 0.05,

*** p < 0.01.

Table 3.

The Hedonic Wage Model Results

	(1)	(2)	(3)	(4)	(5)
	year2016	year2018	pool_1	pool_2	FE
risk_alt	0.0006^* (0.0003)	0.0011^*** (0.0004)	0.0008^*** (0.0002)	0.0012^*** (0.0003)	0.0012^*** (0.0003)
eduyear	0.0660^*** (0.0052)	0.0765^*** (0.0054)	0.0710^*** (0.0038)	0.0711^*** (0.0038)	0.0706^*** (0.0038)
feduyear	0.0095^*** (0.0029)	0.0063^** (0.0030)	0.0082^*** (0.0021)	0.0079^*** (0.0021)	0.0078^*** (0.0021)
female	−0.2149^*** (0.0166)	−0.2370^*** (0.0190)	−0.2248^*** (0.0125)	−0.2301^*** (0.0128)	−0.2299^*** (0.0128)
age	0.0158^*** (0.0039)	0.0113^*** (0.0035)	0.0142^*** (0.0026)	0.0141^*** (0.0026)	0.0138^*** (0.0026)
wexp	−0.0017 (0.0034)	0.0011 (0.0031)	−0.0005 (0.0023)	−0.0003 (0.0023)	−0.0003 (0.0023)
marry	0.1471^*** (0.0170)	0.1156^*** (0.0198)	0.1351^*** (0.0130)	0.1349^*** (0.0129)	0.1346^*** (0.0129)
scale	0.0004^*** (0.0000)	0.0004^*** (0.0000)	0.0004^*** (0.0000)	0.0005^*** (0.0000)	0.0005^*** (0.0000)
TJAN	−0.0503^*** (0.0066)	−0.0519^*** (0.0078)	−0.0502^*** (0.0050)	−0.0503^*** (0.0050)	−0.0507^*** (0.0050)
TJUL	0.0874^*** (0.0159)	0.1035^*** (0.0166)	0.0933^*** (0.0115)	0.0899^*** (0.0115)	0.0907^*** (0.0115)
workhour_alt				−0.0045^*** (0.0011)	−0.0042^*** (0.0011)
unem_alt				−0.0238^* (0.0137)	−0.0170 (0.0137)
year					0.0158^** (0.0061)
_cons	2.7904^*** (0.4185)	2.3943^*** (0.4370)	2.6150^*** (0.3026)	3.5419^*** (0.3666)	−28.3320^** (12.4421)
N	2293	1742	4035	4035	4035
R²	0.349	0.361	0.357	0.359	0.360

Note: Standard errors in parentheses,

* p < 0.1,

** p < 0.05,

*** p < 0.01.

Table 4.

MWTP for risk reduction (safety) by the magnitude of 1/1000,000 (unit: USD)

		MWTP
		Mean	l1	u1
ML	Model 1	1885	1537	2233
ML	Model 2	2152	1798	2506
PML	Model 3	3022	2739	3305
PML	Model 4	2963	2672	3253
HWM	FE	25	13	36

Table A1.

Wage Regression by Industry (2016)

	Manufactory	Construction	Catering	Transportation	Finance	Science	Service
female	−0.272^*** (0.0232)	−0.198^** (0.0785)	−0.201^*** (0.0337)	−0.242^*** (0.0798)	−0.0981 (0.0703)	−0.161^*** (0.0575)	−0.179^*** (0.0575)
eduyear	0.100^*** (0.0059)	0.0347^** (0.0169)	0.0769^*** (0.0101)	0.0495^** (0.0216)	0.113^*** (0.0193)	0.0915^*** (0.0140)	0.0501^*** (0.0130)
lnwexp	0.0990^*** (0.0212)	0.119^* (0.0633)	0.0923^** (0.0439)	0.0393 (0.0621)	0.193^** (0.0838)	0.150^*** (0.0423)	0.0162 (0.0430)
healthlevel	0.00211 (0.0144)	0.0684^* (0.0395)	0.0174 (0.0204)	0.0403 (0.0420)	−0.0068 (0.0445)	−0.0373 (0.0448)	0.0357 (0.0279)
age	0.0088^*** (0.0028)	0.0008 (0.0054)	0.0058 (0.0046)	0.0163^* (0.0089)	0.0132^* (0.0078)	0.0001 (0.0050)	0.0150^** (0.006)
_cons	4.27^*** (0.120)	5.21^*** (0.368)	4.60^*** (0.188)	4.82^*** (0.532)	3.94^*** (0.391)	4.66^*** (0.326)	4.83^*** (0.252)
N	1008	129	495	117	194	167	459
R²	0.390	0.122	0.206	0.186	0.283	0.301	0.117

Notes: Standard errors in parentheses.

* p < 0.1,

** p < 0.05,

*** p < 0.01.ù

Table A2.

Wage Regression by Industry (2018)

	Manufactory	Construction	Catering	Transportation	Finance	Science	Service
female	−0.243^*** (0.0282)	−0.175^** (0.0828)	−0.240^*** (0.0399)	−0.211^** (0.0985)	−0.102 (0.0740)	−0.105^* (0.0548)	−0.231^*** (0.0535)
eduyear	0.110^*** (0.0066)	0.0622^*** (0.0155)	0.0788^*** (0.0110)	0.0730^*** (0.0271)	0.096^*** (0.0197)	0.110^*** (0.0152)	0.0611^*** (0.0115)
lnwexp	0.0804^** (0.0335)	0.256^*** (0.0704)	0.211^*** (0.0567)	0.313^*** (0.101)	0.178^** (0.0885)	0.189^*** (0.0573)	−0.007 (0.0503)
healthlevel	0.0183 (0.0155)	0.00363 (0.0405)	0.0606^** (0.0278)	0.159^*** (0.0564)	0.0279 (0.0408)	0.0310 (0.0413)	0.0277 (0.0255)
age	0.0117^*** (0.0031)	−0.0049 (0.0058)	−0.0017 (0.006)	−0.0134 (0.0092)	0.0099 (0.0079)	0.0012 (0.0041)	0.0098 (0.0062)
_cons	4.07^*** (0.140)	4.96^*** (0.291)	4.40^*** (0.246)	4.53^*** (0.621)	4.19^*** (0.425)	4.03^*** (0.363)	4.95^*** (0.254)
N	937	125	394	96	177	158	395
R²	0.363	0.188	0.248	0.215	0.173	0.371	0.142

Notes: Standard errors in parentheses.

* p < 0.1,

** p < 0.05,

*** p < 0.01.

Table A3.

Summary Statistics of Job Attributes and Worker Attributes by Industry

Variables	Definition	Manufacture Obs. 2,257	Construction Obs. 315	Catering Obs. 1,055	Transportation Obs. 263	Finance Obs. 418	Science Obs. 413	Service Obs. 1,204
wage_alt	yearly wage income, unit: 1000 TWD in 2014 values	500.53 (264.22)	545.57 (154.25)	454.33 (147.79)	519.98 (167.09)	607.97 (228.02)	492.48 (175.29)	517.53 (204.61)
risk_alt	industry mortality rates, unit: 1/1,000,000	23.70 (0.00)	114.50 (8.57)	11.32 (1.65)	48.47 (0.20)	4.03 (1.32)	25.17 (3.64)	8.75 (2.66)
workhour_ alt	industry average monthly total work hours, unit: hour	175.74 (0.30)	166.13 (0.15)	162.34 (1.54)	172.74 (0.30)	167.05 (1.09)	172.26 (0.30)	161.10 (1.49)
umem_alt	industry unemployment rates, unit: %	2.84 (0.02)	4.58 (0.00)	3.53 (0.09)	3.19 (0.00)	2.83 (0.32)	4.31 (0.20)	2.69 (0.14)

[i] Note: standard deviation in parentheses.

[ii] wage_alt is constructed as follows: for the industries that are chosen by workers, wage_alt is their real wage reported in the PSFD, for industries not chosen, wage_alt is predicted from the first stage of wage regression by industry.

References

Authors

Metrics

Articles in this issue

DOI: https://doi.org/10.2478/ceej-2025-0001 | Journal eISSN: 2543-6821 | Journal ISSN: 2544-9001

Journal RSS Feed

Language: English

Page range: 1 - 16

Published on: Jan 7, 2025

Published by: Faculty of Economic Sciences, University of Warsaw

In partnership with: Paradigm Publishing Services

Keywords:

Nonmarket valuation,

Hedonic wage model,

Random utility model,

Value of safety,

VSL

JEL:

J17,

J28,

Q51

Related subjects:

Business and economics,

Political economics,

Economic theory, systems and structures,

Microeconomics,

Macroeconomics,

Economic policy

© 2025 Nan Zhang, published by Faculty of Economic Sciences, University of Warsaw
This work is licensed under the Creative Commons Attribution 4.0 License.

Volume 12 (2025): Issue 59 (January 2025)