Does unemployment have asymmetric effects on suicide rates? Evidence from the United States: 1928–2013

Figures & data

Table 1. Descriptive statistics^a.

	Mean	S.D.	Min	Max
Age-adjusted suicide rate	13.244	2.514	10.391	22.092
Age 15–24 years suicide rate	8.560	3.186	3.829	13.671
Age 25–34 years suicide rate	12.959	2.495	8.196	17.139
Age 35–44 years suicide rate	15.960	2.420	12.099	23.070
Age 45–54 years suicide rate	19.995	5.192	14.189	36.862
Age 55–64 years suicide rate	22.170	8.084	12.061	48.780
Unemployment rate (%)	7.163	4.762	1.200	24.900

^a The whole sample period going from 1928 to 2013, resulting in a total of 86 annual observations in level. Suicide rates are defined as the number of suicides/100,000 population.

Source: Please see 2.2 Data sources in the main text.

Figure 1. Unemployment and suicide rates. Source: Plotted by the authors.

Table 2. Results for the Phillips and Perron unit roots test.

	Level		Difference
	De-mean	De-trend	De-mean	De-trend
Age-adjusted suicide rate	−2.046	−1.840	−5.726**	−5.675**
Age 15–24 years suicide rate	−0.885	−1.508	−8.270**	−8.238**
Age 25–34 years suicide rate	−1.242	−1.506	−8.818**	−8.815**
Age 35–44 years suicide rate	−2.045	−1.851	−8.002**	−8.080**
Age 45–54 years suicide rate	−1.946	−1.572	−6.120**	−6.136**
Age 55–64 years suicide rate	−1.543	−1.739	−5.398**	−6.005**
Unemployment rate (%)	−2.589	−2.822	−5.558**	−5.545**

** represents 1% significance level.

Sources: Calculated by the authors.

Table 3. Results for testing causality using bootstrap simulations.^a

	Null hypothesis	Testing statistics	Bootstrap C.V. at 5%	Bootstrap C.V. at 1%	Lag
Age-adjusted suicide rate (overall rate)	UM≠ > SR	1.490	4.055	6.976	1
	UM^+≠ > SR^+	0.586	3.972	7.209	1
	UM^−≠ > SR^−	7.596*	4.126	8.261	1
Age 15–24 years	UM≠ > SR	0.103	3.966	6.925	1
	UM^+≠ > SR^+	0.242	3.991	6.890	1
	UM^−≠ > SR^−	4.239*	3.983	7.757	1
Age 25–34 years	UM≠ > SR	0.079	4.067	7.417	1
	UM^+≠ > SR^+	0.345	3.898	6.815	1
	UM^−≠ > SR^−	9.817**	4.212	8.822	1
Age 35–44 years	UM≠ > SR	1.048	4.174	7.396	1
	UM^+≠ > SR^+	2.543	3.991	7.008	1
	UM^−≠ > SR^−	12.155**	4.131	8.405	1
Age 45–54 years	UM≠ > SR	2.578	4.068	7.052	1
	UM^+≠ > SR^+	1.061	4.000	7.206	1
	UM^−≠ > SR^−	5.203*	4.041	7.931	1
Age 55–64 years	UM≠ > SR	3.632	3.988	7.079	1
	UM^+≠ > SR^+	4.661*	4.055	7.575	1
	UM^−≠ > SR^−	1.536	4.128	8.097	1

^a Lags used in the symmetric and asymmetric causality tests were chosen using the Hatemi-J information criteria, respectively. As suggested by Toda and Yamamoto (1995), an additional unrestricted lag was included in the V.A.R. model in order to accommodate the unit root property of the unemployment and suicide rates. ** and * represent 1% and 5% significance levels, respectively. UM ≠ > SR means unemployment does not Granger cause suicide rates.

Sources: Calculated by the authors.

Figure 2. Standard generalised responses of suicide rate to unemployment rate. Sources: Plotted by the authors.

†Red-dot lines represent 95% confidence intervals using a Monte-Carlo simulation method with 10,000 repetitions.

Toda, H. Y., & Yamamoto, T. (1995). Statistical inference in vector autoregressions with possibly integrated processes. Journal of Econometrics, 66(1–2), 225–250.

 Web of Science ®Google Scholar