More detailed breakdown on impacts of shocks and efficient measuring of mental health

Abstract

Descriptives on shocks, mental health dynamics and correlation between measures

Shocks

Impact of Shocks on Mental Health

Note

A negative shock is either a natural disaster or a livestock death. The N (positive) row in the table below indicate how many of each shock kind we’ve had in the study. Additionally, more descriptives are available below

The regressions below show the impact of having a shock within a study period on participants’ mental health. Overall, funerals improve mental health, which is not unexpected since they are a large ceremony in rural Ghana according to Damien. Additionally, the questionnaires don’t necessarily specify who’s funeral it is and so the most salient part of the funeral might be the ceremony. However, death of livestock and natural disasters also improve mental health. And, religious events seem to decrease it.

All the regressions included household and period fixed effects. The results are robust to survey month FEs, as well as no time FEs at all. They are also directionally consistent when using different measures of mental health or when only considering the control group, though magnitudes and significance vary (see appendix for more results)

Outcome: Mental Health Index

	Religious Event	Relative Death	Livestock Death	Natural Disaster	Negative shocks
Shock	0.0940**	-0.1497***	-0.0922**	-0.1742**	-0.0979**
	(0.0406)	(0.0385)	(0.0440)	(0.0868)	(0.0408)
Household FE	Yes	Yes	Yes	Yes	Yes
Period FE	Yes	Yes	Yes	Yes	Yes
N	14,666	14,666	14,666	14,666	14,666
R² (within)	0.0018	0.0005	0.0002	0.0003	0.0003
N (positive)	773	667	440	81	513
*p<0.1; **p<0.05; ***p<0.01

Impacts of Shocks - Dynamics

One potential story is the shocks aren’t instant but continuous stressors. For example, the death of a relative could be preceeded by periods of illness, large health care costs and so on. Thus, the actual shock itself might not make things worse.

In such a scenario, shocks would have impact mental health in the period before the actual shock itself. However, regression results don’t support this idea.

Funeral

Religious Event

Livestock death

Negative shocks

Cost of shocks

The impact of a shock does not seem to be affected by the costs associated with it. Below, the coefficients associated with the shock costs are per 100 cedis

Outcome: Mental Health Index

	Religious Event	Relative Death	Livestock Death	Natural Disaster	Negative shocks
Shock	0.1116**	-0.1333***	-0.0801	-0.1140	-0.0878*
	(0.0492)	(0.0417)	(0.0501)	(0.1102)	(0.0462)
Shock Cost	-0.00487060	-0.00341268	-0.00135938	-0.01536153	-0.00120463
	(0.00846744)	(0.00346892)	(0.00179716)	(0.01557997)	(0.00181409)
Household FE	Yes	Yes	Yes	Yes	Yes
Period FE	Yes	Yes	Yes	Yes	Yes
N	14,666	14,666	14,666	14,666	14,666
R² (within)	0.0018	0.0006	0.0003	0.0003	0.0004
N (positive)	773	667	440	81	513
*p<0.1; **p<0.05; ***p<0.01

Hetereogeneity by household characteristics

There seems to be some heterogeneity in the shock impacts. Notably, funerals improve MH for HHs with good mental health at baseline. And, religious events have negative impacts on those who are below median on the census wealth index (could be because associated costs are more burdensome for poorer households).

Mental Health

Outcome: Mental Health Index — Above-Median Baseline MH

	Religious Event	Relative Death	Livestock Death	Natural Disaster	Negative shocks
Shock	0.1108*	-0.2275***	-0.0748	-0.1597	-0.0855
	(0.0578)	(0.0526)	(0.0645)	(0.1178)	(0.0595)
Household FE	Yes	Yes	Yes	Yes	Yes
Period FE	Yes	Yes	Yes	Yes	Yes
N	7,402	7,402	7,402	7,402	7,402
R² (within)	0.0020	0.0013	-0.0000	0.0002	0.0001
N (positive)	383	347	224	42	260
*p<0.1; **p<0.05; ***p<0.01

Outcome: Mental Health Index — Below-Median Baseline MH

	Religious Event	Relative Death	Livestock Death	Natural Disaster	Negative shocks
Shock	0.0767	-0.0646	-0.1107*	-0.1863	-0.1111**
	(0.0570)	(0.0564)	(0.0597)	(0.1272)	(0.0558)
Household FE	Yes	Yes	Yes	Yes	Yes
Period FE	Yes	Yes	Yes	Yes	Yes
N	7,264	7,264	7,264	7,264	7,264
R² (within)	0.0015	0.0000	0.0006	0.0003	0.0007
N (positive)	390	320	216	39	253
*p<0.1; **p<0.05; ***p<0.01

Wealth

Outcome: Mental Health Index — Above-Median Baseline Wealth

	Religious Event	Relative Death	Livestock Death	Natural Disaster	Negative shocks
Shock	0.0218	-0.1663***	-0.0557	-0.1253	-0.0501
	(0.0529)	(0.0538)	(0.0596)	(0.1184)	(0.0558)
Household FE	Yes	Yes	Yes	Yes	Yes
Period FE	Yes	Yes	Yes	Yes	Yes
N	7,425	7,425	7,425	7,425	7,425
R² (within)	0.0004	0.0007	0.0001	0.0001	0.0001
N (positive)	430	321	246	40	280
*p<0.1; **p<0.05; ***p<0.01

Outcome: Mental Health Index — Below-Median Baseline Wealth

	Religious Event	Relative Death	Livestock Death	Natural Disaster	Negative shocks
Shock	0.1805***	-0.1337**	-0.1417**	-0.2271*	-0.1597***
	(0.0627)	(0.0549)	(0.0653)	(0.1274)	(0.0598)
Household FE	Yes	Yes	Yes	Yes	Yes
Period FE	Yes	Yes	Yes	Yes	Yes
N	7,241	7,241	7,241	7,241	7,241
R² (within)	0.0036	0.0003	0.0005	0.0004	0.0008
N (positive)	343	346	194	41	233
*p<0.1; **p<0.05; ***p<0.01

Selection Effects

Shocked households seem to be somewhat richer and larger. However, these should not bias the regression results since we are controlling for household fixed effects.

Baseline Balance of shocked and unshocked households

Balance test: Baseline covariates by endline non-religious shock status (N shocked=347, N not shocked=1925)

Variable	Shocked (mean)	Not shocked (mean)	Diff	Norm. diff	p-value
PHQ-8 (depression, z)	0.058	-0.010	0.068	0.069	0.236
GAD-7 (anxiety, z)	0.043	-0.008	0.051	0.051	0.382
PSS-10 (stress, z)	0.046	-0.008	0.055	0.055	0.336
PSWQ-16 (worry, z)	-0.033	0.006	-0.040	-0.040	0.479
Diener-5 (wellbeing, z)	0.085	-0.015	0.100	0.101	0.078
MH index (average)	0.038	-0.007	0.044	0.066	0.251
Cantril (total)	10.948	11.222	-0.274	-0.072	0.217
Wealth index (census)	-0.179	-0.268	0.089	0.139	0.019
Total consumption (AE, win.)	345.771	344.722	1.049	0.005	0.935
Food consumption (AE, win.)	264.772	253.476	11.296	0.061	0.309
Total income (AE, win.)	91.269	83.142	8.127	0.046	0.420
Current savings (win.)	307.045	264.818	42.227	0.090	0.132
Livestock owned (win.)	4000.285	2708.229	1292.056	0.164	0.011
Durable goods value (win.)	9543.798	8335.584	1208.214	0.134	0.030
Loans taken	0.608	0.549	0.059	0.059	0.309
Household size	8.043	7.655	0.389	0.146	0.012
Participant age	32.931	32.671	0.260	0.029	0.612
Education (years)	1.366	1.498	-0.133	-0.039	0.497
Number of children	4.542	4.299	0.243	0.109	0.067
Number of adults	3.527	3.367	0.161	0.154	0.007
Dependency ratio	1.415	1.393	0.021	0.028	0.641
Body weight (kg)	57.001	58.103	-1.102	-0.076	0.124
Dietary diversity (HDDS)	5.144	5.045	0.099	0.052	0.367
Food insecurity index	0.001	-0.003	0.004	0.006	0.919
Medical expenses	84.324	69.527	14.797	0.040	0.559
Welch's t-test (unequal variance). Norm. diff = difference / sqrt((s²_shocked + s²_not) / 2).

Selection into who reports a shock

It seems that people with worse mental health at time of survey are more likely to report a shock. However, this isn’t perfectly consistent across measures.

Mental Health Index

Outcome: Mental health index (anderson) at Endline

	Relative Death	Religious Event	Negative Shock	Livestock Death	Natural Disaster
Shock	0.0918*	0.0142	0.0093	-0.0071	0.0721
	(0.0476)	(0.0900)	(0.0463)	(0.0457)	(0.0725)
Community FE	Yes	Yes	Yes	Yes	Yes
N	2,186	2,186	2,186	2,186	2,186
R²	0.1777	0.1763	0.1763	0.1763	0.1767
N Reported Shocks	1,228	1,723	1,198	1,059	290
*p<0.1; **p<0.05; ***p<0.01

PHQ-2 (Depression)

Outcome: PHQ-2 (Depression) at Endline

	Relative Death	Religious Event	Negative Shock	Livestock Death	Natural Disaster
Shock	0.1506***	0.0507	0.0901*	0.0590	0.1081
	(0.0486)	(0.0923)	(0.0464)	(0.0459)	(0.0725)
Community FE	Yes	Yes	Yes	Yes	Yes
N	2,186	2,186	2,186	2,186	2,186
R²	0.0886	0.0845	0.0860	0.0850	0.0853
N Reported Shocks	1,228	1,723	1,198	1,059	290
*p<0.1; **p<0.05; ***p<0.01

GAD-2 (Anxiety)

Outcome: GAD-2 (Anxiety) at Endline

	Relative Death	Religious Event	Negative Shock	Livestock Death	Natural Disaster
Shock	0.0994*	-0.0330	0.0528	0.0190	0.1742**
	(0.0522)	(0.0979)	(0.0492)	(0.0492)	(0.0788)
Community FE	Yes	Yes	Yes	Yes	Yes
N	2,186	2,186	2,186	2,186	2,186
R²	0.1021	0.1006	0.1010	0.1006	0.1028
N Reported Shocks	1,228	1,723	1,198	1,059	290
*p<0.1; **p<0.05; ***p<0.01

PSS-4 (Stress)

Outcome: PSS-4 (Stress) at Endline

	Relative Death	Religious Event	Negative Shock	Livestock Death	Natural Disaster
Shock	0.1413**	-0.1806*	-0.0074	-0.0070	0.0307
	(0.0556)	(0.0991)	(0.0544)	(0.0532)	(0.0826)
Community FE	Yes	Yes	Yes	Yes	Yes
N	2,186	2,186	2,186	2,186	2,186
R²	0.1156	0.1143	0.1128	0.1128	0.1129
N Reported Shocks	1,228	1,723	1,198	1,059	290
*p<0.1; **p<0.05; ***p<0.01

PSWQ-3 (Worry)

Outcome: PSWQ-3 (Worry) at Endline

	Relative Death	Religious Event	Negative Shock	Livestock Death	Natural Disaster
Shock	0.1070**	0.1253	0.0206	0.0050	0.0697
	(0.0452)	(0.0831)	(0.0439)	(0.0431)	(0.0668)
Community FE	Yes	Yes	Yes	Yes	Yes
N	2,186	2,186	2,186	2,186	2,186
R²	0.1728	0.1715	0.1706	0.1705	0.1710
N Reported Shocks	1,228	1,723	1,198	1,059	290
*p<0.1; **p<0.05; ***p<0.01

Diener 3-item (Wellbeing)

Outcome: Diener 3-item (Wellbeing) at Endline

	Relative Death	Religious Event	Negative Shock	Livestock Death	Natural Disaster
Shock	0.0998**	-0.0412	0.0735*	0.0617	0.0684
	(0.0439)	(0.0865)	(0.0436)	(0.0432)	(0.0681)
Community FE	Yes	Yes	Yes	Yes	Yes
N	2,186	2,186	2,186	2,186	2,186
R²	0.1539	0.1520	0.1530	0.1527	0.1523
N Reported Shocks	1,228	1,723	1,198	1,059	290
*p<0.1; **p<0.05; ***p<0.01

Community Effects

Below, we take a look at how being a in community that is shocked affects a household. Note that being in a shocked is a reasonably good predictor of experiencing a shock (see table in appendix)

Being in a community with more shocks

Below, we define number of shocks per community member in each period as a measure of shock density. Community size is defined as both the number of participants in a community. Shocks per participant calculation excludes the household’s own shock, so a household isn’t “predicting itself” by reporting a shock. We find that being in a community with more shocks is associated with better mental health.

Outcome: Mental Health Index. Columns indicate type of shock

	Non-religious shocks	Negative shocks	Funeral / Death of relative
Community shocks / participant	-0.3032***	-0.3492**	-0.2831*
	(0.1070)	(0.1540)	(0.1557)
Household FE	Yes	Yes	Yes
Period FE	Yes	Yes	Yes
N	14,666	14,666	14,666
R² (within)	-0.0019	-0.0002	-0.0014
*p<0.1; **p<0.05; ***p<0.01

Interaction with Community Effects

It doesn’t seem that there is much interaction between experiencing a shock and being a community that is shocked a lot. As seen in previous regressions, shocks and being a in a shocked community seemingly improves mental health, but the interaction between the two is generally insignificant, though notable in magnitude.

MH Index

Outcome: Mental health index (anderson). Columns indicate type of shock

	Non-religious shocks	Non-religious shocks w/ int.	Negative Shocks	Negative shocks w/ int	Funeral / Death of relative	Funeral / Death of relative w/ int
Shocked (individual)	-0.1208***	-0.0688	-0.0921**	-0.0832	-0.1454***	-0.1447***
	(0.0304)	(0.0473)	(0.0409)	(0.0579)	(0.0385)	(0.0539)
Community shocks / participant	-0.2696**	-0.2161*	-0.3280**	-0.3194**	-0.2470	-0.2463
	(0.1067)	(0.1146)	(0.1542)	(0.1604)	(0.1555)	(0.1619)
Shocked × community shocks / participant		-0.3507		-0.1256		-0.0074
		(0.2487)		(0.5082)		(0.4307)
Household FE	Yes	Yes	Yes	Yes	Yes	Yes
Period FE	Yes	Yes	Yes	Yes	Yes	Yes
N	14,666	14,666	14,666	14,666	14,666	14,666
R² (within)	-0.0013	-0.0012	0.0001	0.0001	-0.0009	-0.0009
*p<0.1; **p<0.05; ***p<0.01

PHQ-2 (Depression)

Outcome: PHQ-2 (Depression). Columns indicate type of shock

	Non-religious shocks	Non-religious shocks w/ int.	Negative Shocks	Negative shocks w/ int	Funeral / Death of relative	Funeral / Death of relative w/ int
Shocked (individual)	-0.0943***	-0.0783	-0.0581	-0.0282	-0.1300***	-0.1380**
	(0.0315)	(0.0483)	(0.0432)	(0.0608)	(0.0394)	(0.0569)
Community shocks / participant	-0.2731**	-0.2566**	-0.4107***	-0.3816**	-0.1817	-0.1902
	(0.1112)	(0.1174)	(0.1556)	(0.1614)	(0.1588)	(0.1656)
Shocked × community shocks / participant		-0.1082		-0.4211		0.0847
		(0.2713)		(0.5853)		(0.4664)
Household FE	Yes	Yes	Yes	Yes	Yes	Yes
Period FE	Yes	Yes	Yes	Yes	Yes	Yes
N	14,666	14,666	14,666	14,666	14,666	14,666
R² (within)	-0.0051	-0.0051	-0.0016	-0.0014	-0.0027	-0.0027
*p<0.1; **p<0.05; ***p<0.01

GAD-2 (Anxiety)

Outcome: GAD-2 (Anxiety). Columns indicate type of shock

	Non-religious shocks	Non-religious shocks w/ int.	Negative Shocks	Negative shocks w/ int	Funeral / Death of relative	Funeral / Death of relative w/ int
Shocked (individual)	-0.0930***	-0.0341	-0.0342	-0.0254	-0.1438***	-0.1343**
	(0.0348)	(0.0538)	(0.0498)	(0.0664)	(0.0434)	(0.0630)
Community shocks / participant	-0.1861	-0.1256	-0.2757	-0.2672	-0.1241	-0.1140
	(0.1204)	(0.1288)	(0.1695)	(0.1766)	(0.1773)	(0.1836)
Shocked × community shocks / participant		-0.3969		-0.1240		-0.1009
		(0.3040)		(0.6272)		(0.5168)
Household FE	Yes	Yes	Yes	Yes	Yes	Yes
Period FE	Yes	Yes	Yes	Yes	Yes	Yes
N	14,666	14,666	14,666	14,666	14,666	14,666
R² (within)	-0.0043	-0.0039	-0.0015	-0.0014	-0.0024	-0.0024
*p<0.1; **p<0.05; ***p<0.01

PSS-4 (Stress)

Outcome: PSS-4 (Stress). Columns indicate type of shock

	Non-religious shocks	Non-religious shocks w/ int.	Negative Shocks	Negative shocks w/ int	Funeral / Death of relative	Funeral / Death of relative w/ int
Shocked (individual)	-0.0611	-0.0515	-0.0776	-0.1320*	-0.0598	-0.0580
	(0.0391)	(0.0607)	(0.0532)	(0.0733)	(0.0503)	(0.0725)
Community shocks / participant	-0.1195	-0.1096	0.0382	-0.0145	-0.2492	-0.2473
	(0.1302)	(0.1394)	(0.1891)	(0.1973)	(0.1844)	(0.1908)
Shocked × community shocks / participant		-0.0652		0.7644		-0.0190
		(0.3082)		(0.6542)		(0.5277)
Household FE	Yes	Yes	Yes	Yes	Yes	Yes
Period FE	Yes	Yes	Yes	Yes	Yes	Yes
N	14,666	14,666	14,666	14,666	14,666	14,666
R² (within)	0.0005	0.0004	0.0002	0.0004	0.0003	0.0003
*p<0.1; **p<0.05; ***p<0.01

PSWQ-3 (Worry)

Outcome: PSWQ-3 (Worry). Columns indicate type of shock

	Non-religious shocks	Non-religious shocks w/ int.	Negative Shocks	Negative shocks w/ int	Funeral / Death of relative	Funeral / Death of relative w/ int
Shocked (individual)	-0.0840**	-0.0389	-0.0646	-0.0643	-0.0989**	-0.1302**
	(0.0359)	(0.0568)	(0.0504)	(0.0721)	(0.0448)	(0.0629)
Community shocks / participant	-0.2975**	-0.2511**	-0.1881	-0.1877	-0.4252**	-0.4584**
	(0.1172)	(0.1262)	(0.1882)	(0.1935)	(0.1709)	(0.1803)
Shocked × community shocks / participant		-0.3040		-0.0048		0.3297
		(0.3031)		(0.6736)		(0.5043)
Household FE	Yes	Yes	Yes	Yes	Yes	Yes
Period FE	Yes	Yes	Yes	Yes	Yes	Yes
N	14,666	14,666	14,666	14,666	14,666	14,666
R² (within)	-0.0030	-0.0030	-0.0004	-0.0004	-0.0031	-0.0031
*p<0.1; **p<0.05; ***p<0.01

Diener 3-item (Wellbeing)

Outcome: Diener 3-item (Wellbeing). Columns indicate type of shock

	Non-religious shocks	Non-religious shocks w/ int.	Negative Shocks	Negative shocks w/ int	Funeral / Death of relative	Funeral / Death of relative w/ int
Shocked (individual)	0.0657**	0.0291	0.0602	0.0316	0.0607	0.0395
	(0.0329)	(0.0525)	(0.0475)	(0.0681)	(0.0413)	(0.0616)
Community shocks / participant	0.0557	0.0180	0.2191	0.1913	-0.0673	-0.0899
	(0.1096)	(0.1168)	(0.1654)	(0.1692)	(0.1647)	(0.1702)
Shocked × community shocks / participant		0.2469		0.4022		0.2239
		(0.2844)		(0.6706)		(0.4999)
Household FE	Yes	Yes	Yes	Yes	Yes	Yes
Period FE	Yes	Yes	Yes	Yes	Yes	Yes
N	14,666	14,666	14,666	14,666	14,666	14,666
R² (within)	0.0019	0.0019	0.0014	0.0014	0.0002	0.0002
*p<0.1; **p<0.05; ***p<0.01

Descriptives on Shocks

Timing of Shocks

Reported shock dates in baseline and endline surveys. Point shape indicates whether the shock was reported in period 0 or period 6; counts are capped at 100 households on the y-axis

Types of shock we had and counts

Note

Transparent colors indicate endline data

Count of households reporting each shock type at baseline (period 0)

Shocks and Communities - Descriptives

All non-religious shocks per study participant

Natural disaster shocks per study participant

Livestock death shocks per study participant

Relative death shocks per study participant

Non-religious shocks geographic distribution

/var/folders/rj/c4rjx52d217gsm3ccfx5q20r0000gr/T/ipykernel_99518/1357806734.py:33: DeprecationWarning: *scattermapbox* is deprecated! Use *scattermap* instead. Learn more at: https://plotly.com/python/mapbox-to-maplibre/

Natural disasters distribution

/var/folders/rj/c4rjx52d217gsm3ccfx5q20r0000gr/T/ipykernel_99518/1357806734.py:33: DeprecationWarning: *scattermapbox* is deprecated! Use *scattermap* instead. Learn more at: https://plotly.com/python/mapbox-to-maplibre/

Livestock death distribution

/var/folders/rj/c4rjx52d217gsm3ccfx5q20r0000gr/T/ipykernel_99518/1357806734.py:33: DeprecationWarning: *scattermapbox* is deprecated! Use *scattermap* instead. Learn more at: https://plotly.com/python/mapbox-to-maplibre/

Appendix

Regressions - Shocks on Mental Health

The below are alternate specs for the main shock regressions with different sets of fixed effects and impact on individual mental health measures.

Control Group Only

Outcome: Mental Health Index

	Religious Event	Relative Death	Livestock Death	Natural Disaster	Negative shocks
Shock	0.1170	-0.1235	-0.0968	-0.2061	-0.1018
	(0.0858)	(0.0824)	(0.1080)	(0.3937)	(0.1047)
Household FE	Yes	Yes	Yes	Yes	Yes
Period FE	Yes	Yes	Yes	Yes	Yes
N	2,508	2,508	2,508	2,508	2,508
R² (within)	0.0009	0.0007	0.0004	0.0001	0.0005
N (positive)	122	95	85	4	89
*p<0.1; **p<0.05; ***p<0.01

Survey Month FEs

Outcome: Mental Health Index

	Religious Event	Relative Death	Livestock Death	Natural Disaster	Negative shocks
Shock	0.1308***	-0.1077***	-0.0743*	-0.1845**	-0.0850**
	(0.0346)	(0.0355)	(0.0407)	(0.0792)	(0.0376)
Household FE	Yes	Yes	Yes	Yes	Yes
Period FE	Yes	Yes	Yes	Yes	Yes
N	14,666	14,666	14,666	14,666	14,666
R² (within)	0.0078	0.0073	0.0068	0.0069	0.0070
N (positive)	773	667	440	81	513
*p<0.1; **p<0.05; ***p<0.01

No Time FEs

Outcome: Mental Health Index

	Religious Event	Relative Death	Livestock Death	Natural Disaster	Negative shocks
Shock	0.1725***	-0.1003***	-0.0742*	-0.1822**	-0.0838**
	(0.0332)	(0.0345)	(0.0400)	(0.0788)	(0.0369)
Household FE	Yes	Yes	Yes	Yes	Yes
Period FE	Yes	Yes	Yes	Yes	Yes
N	14,666	14,666	14,666	14,666	14,666
R² (within)	0.0022	0.0007	0.0002	0.0003	0.0003
N (positive)	773	667	440	81	513
*p<0.1; **p<0.05; ***p<0.01

PHQ-2 (Depression)

Outcome: PHQ-2 (Depression)

	Religious Event	Relative Death	Livestock Death	Natural Disaster	Negative shocks
Shock	0.0556	-0.1332***	-0.0545	-0.1499	-0.0654
	(0.0409)	(0.0395)	(0.0468)	(0.1022)	(0.0432)
Household FE	Yes	Yes	Yes	Yes	Yes
Period FE	Yes	Yes	Yes	Yes	Yes
N	14,666	14,666	14,666	14,666	14,666
R² (within)	0.0006	-0.0006	-0.0001	0.0002	-0.0001
N (positive)	773	667	440	81	513
*p<0.1; **p<0.05; ***p<0.01

GAD-2 (Anxiety)

Outcome: GAD-2 (Anxiety)

	Religious Event	Relative Death	Livestock Death	Natural Disaster	Negative shocks
Shock	-0.0029	-0.1460***	-0.0216	-0.0703	-0.0391
	(0.0434)	(0.0433)	(0.0539)	(0.1214)	(0.0497)
Household FE	Yes	Yes	Yes	Yes	Yes
Period FE	Yes	Yes	Yes	Yes	Yes
N	14,666	14,666	14,666	14,666	14,666
R² (within)	-0.0001	-0.0009	-0.0001	0.0000	-0.0001
N (positive)	773	667	440	81	513
*p<0.1; **p<0.05; ***p<0.01

PSS-4 (Stress)

Outcome: PSS-4 (Stress)

	Religious Event	Relative Death	Livestock Death	Natural Disaster	Negative shocks
Shock	0.1109**	-0.0642	-0.0612	-0.1662	-0.0770
	(0.0466)	(0.0500)	(0.0576)	(0.1176)	(0.0531)
Household FE	Yes	Yes	Yes	Yes	Yes
Period FE	Yes	Yes	Yes	Yes	Yes
N	14,666	14,666	14,666	14,666	14,666
R² (within)	0.0004	0.0002	0.0002	0.0002	0.0003
N (positive)	773	667	440	81	513
*p<0.1; **p<0.05; ***p<0.01

PSWQ-3 (Worry)

Outcome: PSWQ-3 (Worry)

	Religious Event	Relative Death	Livestock Death	Natural Disaster	Negative shocks
Shock	0.0785*	-0.1063**	-0.0864	0.0307	-0.0680
	(0.0461)	(0.0448)	(0.0534)	(0.1302)	(0.0503)
Household FE	Yes	Yes	Yes	Yes	Yes
Period FE	Yes	Yes	Yes	Yes	Yes
N	14,666	14,666	14,666	14,666	14,666
R² (within)	0.0011	-0.0003	0.0000	0.0000	0.0000
N (positive)	773	667	440	81	513
*p<0.1; **p<0.05; ***p<0.01

Diener 3-item (Wellbeing)

Outcome: Diener 3-item (Wellbeing)

	Religious Event	Relative Death	Livestock Death	Natural Disaster	Negative shocks
Shock	-0.0575	0.0595	0.0645	0.1784	0.0641
	(0.0441)	(0.0412)	(0.0508)	(0.1183)	(0.0474)
Household FE	Yes	Yes	Yes	Yes	Yes
Period FE	Yes	Yes	Yes	Yes	Yes
N	14,666	14,666	14,666	14,666	14,666
R² (within)	0.0004	0.0008	0.0004	0.0002	0.0005
N (positive)	773	667	440	81	513
*p<0.1; **p<0.05; ***p<0.01

Community Shocks Predicting Individual Shocks

Outcome: Individual shock (LPM). Columns indicate type of shock

	Non-religious shocks	Negative shocks	Death of relative
Shock in Community (Yes / No)	0.0905***	0.0734***	0.0756***
	(0.0052)	(0.0039)	(0.0042)
Household FE	Yes	Yes	Yes
Period FE	Yes	Yes	Yes
N	14,666	14,666	14,666
R² (within)	0.0425	0.0452	0.0451
*p<0.1; **p<0.05; ***p<0.01

Baseline Shock Reports

shock_type_label	Livestock Death	Natural Disaster	Other	Relative Death	Religious Event
treatment
Control	141	41	8	137	283
Predictable	145	49	7	154	294
Stable	139	38	5	150	273
Unpredictable	387	100	18	405	780

Endline Shock Reports

shock_type_label	Livestock Death	Natural Disaster	Other	Relative Death	Religious Event
treatment
Control	96	8	17	121	168
Predictable	92	15	31	125	159
Stable	75	21	27	142	171
Unpredictable	236	51	74	363	438