nature .

human behaviour



LETTERS



PUBLISHED: 30 JANUARY 2017 | VOLUME: 1| ARTICLE NUMBER: 0040



Economic insecurity and the rise in gun violence



at US schools



A. R. Pah’, J. Hagan?4, A. L. Jennings?, A. Jain°®, K. Albrecht3, A. J. Hockenberry® and L. A. N. Amaral'*7”*



Frequent school shootings are a unique US phenomenon that

has defied understanding"?. Uncovering the aetiology of this

problem is hampered by the lack of an established dataset?“.

Here we assemble a carefully curated dataset for the period

1990-2013 that is built upon an exhaustive review of exist-

ing data and original sources. Using this dataset, we find

that the rate of gun violence is time-dependent and that this

rate is heightened from 2007 to 2013. We further find that

periods of increased shooting rates are significantly correlated

with increases in the unemployment rate across different

geographic aggregation levels (national, regional and city).

Consistent with the hypothesis that increasing uncertainty

in the school-to-work transition contributes to school shoot-

ings, we find that multiple indicators of economic distress

significantly correlate with increases in the rate of gun vio-

lence when events at both K12 and post-secondary schools

are considered.



Although there is extensive work on school shootings'**"", the

result is a patchwork of contradictory claims. Some studies report

an insignificant increase in the rate of school shooting incidents

or homicides over time?*°, while others point to copycat effects

and an increasing frequency in certain population segments and

settings*'®. Going beyond these basic insights muddies the waters

further, as the identification of sociological and risk factors related

to being a shooter!" are of doubtful predictive value because of

their lack of specificity to individuals that commit gun violence

at schools*!*'8,



Resolving the differing perspectives on gun violence at schools is

challenging for several reasons. Not least is the lack of a definitive

data source with clear event inclusion criteria. Event counts of gun

violence at schools vary dramatically across sources (Fig. 1a), which

creates concern about the reliability of previous quantitative analy-

ses. Even simple questions—such as whether the rate of shootings is

increasing—are impossible to answer without valid data.



Multiple datasets with different inclusion criteria were used in

previous research on mass killings, shootings and gun violence at

schools (Fig. 1). To create a consistent dataset to investigate the

phenomenon of gun violence at schools, we advance the following

criteria for event inclusion: (1) the shooting must involve a firearm

being discharged, even if by accident; (2) it must occur on a school

campus; and (3) it must involve students or school employees, either

as perpetrators, bystanders or victims. As an example, gang violence

on a playground at night during the summer months would not be

included since it violates the last criterion, while a student being

shot at the school’s baseball field after a game is included.



To build our dataset, we merged events from six original datasets

pertaining to school violence (Fig. 1), resulting in 535 events. We

cleaned the merged dataset and corrected dates using primary news

sources, resulting in 529 events for potential inclusion. Three coders

then independently evaluated each individual event against the

defined criteria. If at least two coders agreed that an event should be

included, then the event was added to the final consensus dataset.

This process yielded 379 events meeting our strict criteria and two

additional events found during the discovery process that were not

present in any of the original six datasets.



Our consensus dataset is positively correlated with all of the

original datasets—demonstrating that our criteria do not exclude

events from any one dataset at a noticeably higher rate than any of

the others (Supplementary Fig. 2). We categorized the events within

the consensus dataset to gain a more concrete understanding of

what constitutes gun violence at schools (Fig. 2). Consistent with

previous reports*”’, we find that most events are targeted, that is, the

shooter intends to harm a specific person.



In our dataset, gang-related violence constitutes 6.6% of all

incidents; this is a much smaller fraction compared with what

is observed for urban violence outside of schools”. The average

number of fatalities per event is one and the number of incidents

with three or more deaths constitutes 6.3% of included events.

Moreover, gun violence at schools has not become more deadly over

time (Fig. 3a and Supplementary Table 3).



It is also important to note that that this dataset is focused on

all gun violence at schools and is not limited to mass shootings.

Although there are notable mass shooting events on school cam-

puses, most mass shootings happen at locations other than schools.

Similarly, this dataset includes all instances of gun usage, whether

someone dies in the course of the event or not, since the discharge

of a gun is not permitted on school campuses (despite allowances ‘to

carry on some state college campuses”'). The inclusion of attempted

violence distinguishes this dataset from other measures of violence,

such as the homicide rate, since that rate is by definition only

concerned with acts of violence resulting in a death.



We next evaluate the timing of these events to determine whether

they follow a Poisson process. Since our dates cover an extended

period of time, we allow for the possibility that the rate parame-

ter, A, varies over time in a stepwise manner. We fit models with

an increasing number of change-points to the monthly time series

of events and find that the best fit has four distinct periods sepa-

rated by three change-points (see Supplementary Information for

methods, parameter values and information criteria scores; Fig. 3a

for model fit).



‘Northwestern Institute on Complex Systems (NICO), Northwestern University, Evanston, Illinois 60208, USA. *Management and Organizations, Kellogg

School of Management, Northwestern University, Evanston, Illinois 60208, USA. ?Department of Sociology, Northwestern University, Evanston, Illinois

60208, USA. “American Bar Foundation, Chicago, Illinois 60611, USA. °Department of Chemical and Biological Engineering, Northwestern University,

Evanston, Illinois 60208, USA. Department of Physics and Astronomy, Northwestern University, Evanston, Illinois 60208, USA. Department of Medicine,

Northwestern University, Evanston, Illinois 60208, USA. *e-mail: amaral@northwestern.edu



NATURE HUMAN BEHAVIOUR 1, 0040 (2017) | DOI: 10.1038/s41562-016-0040 | www.nature.com/nathumbehav 1



© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.

LETTERS NATURE HUMAN BEHAVIOUR

a

~ 404 / Brady Campaign 40 4 SAVD

Ss 2 events 236 events

£2 301 /- 304

o

3

£ 204 Lo 20 4

8

= 104 Lo 10 4

<

0-4 0+

1990 1995 2000 2005 2010 2015 1990 1995 2000 2005 2010 2015

~ 407 Shultz 40 + Slate

SF 215 events 166 events.

wn 7

€ 304 304

3 VA. VA

© 204 20 4

3 a y

= 104 10 4

<

o- o+

1990 1995 2000 2005 2010 2015 1990 1995 2000 2005 2010 2015

4 Virginia Tech 4 Wikipedia

S 40 aan 40 174 events

2 304 / 304

vo

3

€ 204 Lo 20 4

mi)

a

= 104 10 4

<x

o Lian —puNenten gle o-

1990 1995 2000 2005 2010 2015 1990 1995 2000 2005 2010 2015

Year Year

b

40 4

Consensus

30 4 381 events

s

wn

<

o

3

2 204

vl

2

Cc

<x

(e) T T | | 1

1990 1995 2000 2005 2010 2015



Year



Figure 1| Comparison of annual shooting event counts in school-associated datasets for 1990-2013. a, The individual datasets differ dramatically in

both the annual number (n) of events and the pattern over time. The School Associated Violent Deaths (SAVD) and Shultz datasets record a high number

of events for 1993 and 1994 and display a decreasing trend in the number of events over time. In contrast, the Brady Campaign, Slate and Wikipedia

datasets record the greatest number of events in the period 2006-2012 and display an increasing trend in the number of events over time. The Virginia

Tech report has a relatively small number of recorded events during this time period, which makes it difficult to assess any trend over time. Years not

covered in an individual dataset are shaded in grey to make it clear that the missing data is due to absence of coverage. b, Our curated consensus dataset

comprises 381 events and displays a more complex trend over time than any of the independent datasets.



Distinct periods marked by different rates suggest a possible

dependence on exogenous factors. An exogenous factor that may

be considered a plausible cause of this phenomenon is gun avail-

ability, but changes in gun ownership do not align with periods of

higher values of gun violence in schools (Supplementary Fig. 4). In

contrast, there is a visual congruence between periods of increased

unemployment and periods of elevated shooting rates at the national

level (Fig. 3b).



The unemployment rate is particularly of interest, since it

is a single aggregate statistic that captures the difficulties faced



by older students in the school-to-work transition or by students’

families”. Joblessness is related to lowered self-esteem, dimin-

ished status and detrimental behaviour’. There is also evidence

that minors may be responsive to the unemployment of their

parents**> and that the attitudes of youths have a significant

impact on their future employment prospects and earnings**°. We

posit that gun violence at schools is a response, in part, to the

breakdown in the expectation that sustained participation in

the educational system will improve economic opportunities

and outcomes.



2 NATURE HUMAN BEHAVIOUR 1, 0040 (2017) | DOI: 10.1038/s41562-016-0040 | www.nature.com/nathumbehav



© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.

NATURE HUMAN BEHAVIOUR LETTERS

a a _—

70- 50- nv Change-point model

60 + 40 | Z 10-

% 07 & 5 8

£ 404 #30 :

& 30 8 594 £ 64

oO o =

2 204 a 10 = 44

104 = 21

0s O° > 0

‘S < x &< & 4. . Le 7

se sx SF M&M & < 1990 1995 2000 2005 2010

S we ne <2 Year

; ; b 12

Figure 2 | Breakdown of urbanity and event type in the consensus 3 -

dataset. a, The urbanity of the events mirrors the population distribution Ss 10

wn oO

in the United States (6% rural, 26% suburban, 67% urban), with 5 6-4 3 3

most events occurring in urban areas. b, Events categorized based 3 cy

Cc

on the reported outcome and shooter intent (when available). 344 6 a

Targeted events (an attacker intending to shoot one or two others) = > a

. . fo} 4 x

is the primary event type. = 4s

0-5 2

National level 1990 1995 2000 2005 2010

Since we hypothesize that increased school shootings are a response Year

to increasing unemployment, we fit the data using Poisson regression: fc 4) 95% Cl

E[Sypltms Ms] = eF0tP itt Poms (1) < 6.

5 we ;

where S,, is the number of shootings per month, u,, isthe monthly % 44 ese P e208

unemployment rate, m, is a dummy variable that accounts for the = @| eee eee @| ee emmme @o ww

summer months, E is the expected value, and f,, 6,, and 8, arethe § 27 Cmnen G00 WHOMOS GO C8 © O@ OEOw

parameters being estimated. We find a significant (P<10~, pseudo- = aero © neem oman ¢ Ene © eae

: : + © SER STS 1S Tre ao~

R’?=0.074) relationship between the unemployment rate and num- 0 | —_ oe



ber of incidents per month (95% confidence intervals (CIs) shown

in Fig. 3c and Supplementary Table 8). Although the pseudo-R?

value is low, this is largely due to the inherent noise in Poisson pro-

cesses—the fit captures 53% of the maximum variance that would

be expected for this number of observations even if there was a

perfect correlation between unemployment and the number of inci-

dents per month (Supplementary Fig. 5). The unemployment rate

is still a significant predictor if we control for the change in student

population over time (Supplementary Table 9).



To further confirm the robustness of this finding, we test our

hypothesis in two additional ways. First, we model the relationship

of the average time between events to unemployment. Using this

formulation, we find again that there is a significant relationship

between increasing unemployment and decreasing time between

event incidences (P=0.011, R’=0.10; Supplementary Table 10).

Second, we normalize the unemployment rate into the range

[0,1] during the time period studied and categorize the months

based on the number of shootings within each month. Since the

period (1994-2007) with a lowered rate of shootings has an aver-

age of approximately one shooting per month, we use that number

of shootings per month as a threshold to separate the two groups.

If unemployment is a factor in school shootings then we would

expect that months with more shootings would have a significantly

larger mean normalized unemployment rate. This is indeed what

we observe. We find that the two distributions are significantly dif-

ferent and that months with two or more events have a larger mean

normalized unemployment rate (0.43 versus 0.35; Kolmogorov—

Smirnov (K-S) two-sample test, P=0.006; Fig. 3d).



Next, we test our hypothesis at different levels of spatial aggrega-

tion to assess whether this relationship is conditional on location or

might arise from an ecological fallacy*’.



Regional level

We partition the continental United States into seven regions according

to geography and socioeconomic similarity (Fig. 4a and Supplementary



T T T

2000 2005 2010



Year



T T

1990 1995



a



2.5 = @<1event

@ >1 event



Probability density



0.0 O1 02 03

Normalized unemployment rate



04 05 06 07 O08 09



Figure 3 | The rate of school shootings is time-dependent and correlated

with increasing unemployment above ‘normal’ levels. a, The monthly

number of events categorized based on number of fatalities (green O-1,

orange 2-5 and red >5). We fit Poisson process change-point models to

the monthly incident time series and find that the best fit model has four

distinct periods (see Supplementary Table 7). b, National unemployment

rate peaks (black line) qualitatively align with periods of elevated rates of

school shootings (blue bars). ¢, Confidence intervals for the fit of national

unemployment rate to monthly shootings (blue dots) (see Equation (1)).

d, The distributions of normalized unemployment in months with <1 and

>1 event differ significantly and months with >1 shooting have a larger

mean normalized unemployment.



Fig. 7). We examine the distribution of normalized unemployment

rates, with each region having its unemployment normalized into

the range [0,1] individually. Due to the lower frequency of events at a

regional scale, we partition months into those with no shootings and

those with shootings. As before, we find that months with one or more

shootings have a normalized unemployment rate distribution that sig-

nificantly differs with a larger mean normalized unemployment rate

(0.41 versus 0.37; K-S two-sample test, P=0.017; Fig. 4b).



NATURE HUMAN BEHAVIOUR 1, 0040 (2017) | DOI: 10.1038/s41562-016-0040 | www.nature.com/nathumbehav 3



© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.

a

S 12 5 Northeast 9 | S 12 5 Southeast We

¥ 104 8 ® 4 104 9 38

5 84 73 5 a 8 3

3s o 3 co b

4 < Vv 4 <

2 6 6 3 2 6 7 3 B No events

= 44 5 3 2 44 6 3 » 2556 Events

2 24 43 © 24 Ee 8

iS} x [e} & 4 2.0

= 04 3° = 04 4 ~- o

1990 1995 2000 2005 2010 1990 1995 2000 2005 2010 3 15

= 10

Se 125 Great lakes ae Se 125 Pacific Ze g

S ns + nz 9 05

4» 104 10 2 4 10-4 ic) a

5 3 92 5 a1 pe 9.0

3 6 8 g 3 6 4 3 Sg 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

= a Z 3 = a4 7 a Normalized unemployment rate

s 5 6%

e 24 48 5§ 27 5 §

= 04 3“ = 04 4 ~

1990 1995 2000 2005 2010 1990 1995 2000 2005 2010

Year Year



Figure 4 | Elevated unemployment is correlated with an increased rate of school shootings at the regional level. a, Average unemployment (black lines)

and school shooting incidents (blue bars) for the four most active regions (all eight regions shown in Supplementary Figure 7). b, When the distributions



of these normalized unemployment rates are compared, we find that there is a significant difference and months with shootings have an



over-representation of months with a high normalized unemployment rate.



City level



We analyse the six cities with the most gun violence at schools: New York

City, Detroit, Chicago, Memphis, Los Angeles and Houston (Fig. 5a). As

for the national and regional levels, we find that months with one or

more shootings have a normalized unemployment rate distribution that

significantly differs and has a larger mean normalized unemployment

rate (0.51 versus 0.41; K-S two-sample test, P=0.005; Fig. 5b).



Educational attainment

Our results strongly support the hypothesis that a breakdown in

the school-to-work transition contributes to an increase in gun



violence at schools. Taking this hypothesis a step further, we would

expect that there would be a shift in the temporal location of these

shootings during the period when post-secondary education has

increasingly supplanted high school in determining successful

school-to-work transitions*’*”. When we analyse the post-secondary

event series separately, we do find that the rate of gun violence is

elevated from November 2005 to December 2013 (Supplementary

Fig. 8 and Supplementary Table 15).



When these individual time series are fit against correspond-

ing unemployment metrics (‘less than high school’ unemploy-

ment levels for K12 schools and ‘some college’ unemployment for



4a 6, Chicago r13 4 ~ 64 Los Angeles -14 7

s 12 & s &

2°] bi wS 2°] b 12 WS

g4t plas B44 lio a3

S +9 50 3 ie}

£37 lg So = 34 lg #2

2 24 17 23 224 <3

z 1 Fé sa = 44 Le 38

L 5 0 09



ele Li bss 27) Uh Lie



1990 1995 2000 2005 2010 1990 1995 2000 2005 2010

2 65 Houston ro Ss 2 65 Detroit [ #0 > b 2.5 Noevents

So] Leg Sec] r & 2 B Events

@ 45 t7 33 @ 45 p14 38 S45

s 30S -f12 50 >

£37 ;&§ Sg 2 34 Lio 32 24



L ge



21 mimirsé 23) 4 HL fs 8" 20,



0 T T T 3 0 T T 2~ 0.0



1990 1995 2000 2005 2010 1990 1995 2000 2005 2010 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9



6 New York 10 6 M hi n Normalized unemployment rate

~ 67 ew Yor r 10 ~ 69 emphis rc A

s 5 J Lo & Ss 54 L10 &

2 lg wS 2 Lg wS

o 44 32 6 44 3

3 L782 8 re 82

23) Po es 234 b? 38



wo bt og



= 27 i5 ag £71 83

ae Laas 514 A at

2o/1 Jl {1 13% =2,| ll | IIs ¢



1990 1995 2000 2005 2010 1990 1995 2000 2005 2010



Year Year



Figure 5 | Elevated unemployment is significantly correlated with an increased rate of school shootings in the most active cities. a, The city-wide

unemployment rate (black lines) and school shooting incidents (blue bars) for the six cities with the most events. b, The distributions of normalized

unemployment rates between months with and without a shooting differ significantly and months with a shooting have a larger mean normalized

unemployment rate.



4 NATURE HUMAN BEHAVIOUR 1, 0040 (2017) | DOI: 10.1038/s41562-016-0040 | www.nature.com/nathumbehav



© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.

NATURE HUMAN BEHAVIOUR LETTERS

85 K12 25 0.20 5

e774 20 3 #

2 64 15 3 § 0154

s 5 4 1.0 g ‘S

£44 05 8 2 010+

> 34 00 2

= ny E

€ 24 054 60.054

fo} fal S

214 -10 § &

o4 15 ~ 0.004

1990 1995 2000 2005 2010

44 Post-secondary p25 _, 0.85

fo}

S L 2.0 3 = 0.7 5

¥% 34 lis 9 @& 067

c . Pat

2 Lio a § 057

£27 os 5 044

= p<? 2 2 034

N

214 r00 3 & o24

{o} g Ss

= | r --05°2 & oi

Ly 1 3

(0) } ; --10 ~ 0.0 4

1990 1995 2000 2005 2010

9 5 Al 25 0.30 5

S 7 20 3 €

2 74 16 3 5

3°] 10 &

S54 os 28

a . g 5

23 00 2 @

cS 4 \ E

6 24 -0.5 # s

= 414 -1.0 S a

04 15 ~~

1990 1995 2000 2005 2010

Year

SS



Figure 6 | A broad range of economic indicators significantly correlate to gun violence in post-secondary and all schools. Left: all economic indicators

(black lines) are normalized according to their mean and standard deviation to make the parameter coefficients comparable across indicators.



The consumer confidence is inverted so it is instead a measure of the lack of consumer confidence in the economy. The composite measure is



a naive linear combination of all three economic indicators. Monthly incidents (blue bars) in K12 schools and the entire dataset are best described



with the national unemployment rate, while the foreclosure rate best describes events in post-secondary schools. Right: with the exception of consumer

confidence with K12 schools, all of the economic indicators are comparable in magnitude within K12, post-secondary and all schools (error bars are



standard errors).



post-secondary schools), we again find unemployment as a signifi-

cant predictor of shootings (P=0.007, pseudo-R?=0.068 for K12

schools and P< 10~°, pseudo-R’ = 0.072 for post-secondary schools;

Supplementary Tables 17 and 18).



Broader economic indicators

Although unemployment is one proxy for the breakdown in the

school-to-work transition and for economic insecurity, there are

other plausible measures that could be used to model this break-

down, especially given the changes in the US economy over the last

25 years. Two other metrics that may provide a proxy for economic

insecurity of families are the foreclosure rate and consumer confi-

dence. We assess the ability of these two measures, unemployment

and a composite indicator (an equal weighting of foreclosures,

inverse consumer confidence and unemployment) to describe the

incidence of gun violence events at K12 schools, post-secondary

schools and across all schools (Fig. 6). We find that correlations with

these indicators are significant at post-secondary schools and across

all schools, while only unemployment is strongly significant at K12

schools. Further, the magnitude of the parameters are comparable

for a given setting after normalizing the indicators, suggesting that

this is a generalizable phenomenon that can be robustly measured

with a variety of measures of economic insecurity.



In the last 25 years, there have been two periods of elevated gun

violence at schools in the United States and the timing of these



periods significantly correlates with increased economic insecu-

rity. With the unemployment rate as an indicator, this effect persists

at the national, regional and city levels of geographic aggregation.

Further, we find that this effect is measurable across several eco-

nomic indicators, which underscores the robustness of our findings.



In accordance with the theory that gun violence is associated witha

breakdown in the school-to-work transition, we find that where these

events occur has shifted in the last 20 years. The 2007-2013 period

of elevated gun violence is largely due to events at post-secondary

schools, while the 1992-1994 period more often involves events

at K12 schools. Given the nature of the school-to-work transition,

it is predictable that more violence would occur closer to the last

link in the chain from education to employment. An implication of

our findings is that as economic prospects improve, the frequency

of shootings in K12 schools should remain relatively stable, with

declines at post-secondary schools.



The far-reaching impact of job loss, short-term unemployment,

under-employment and long-term idleness is well established in the

research literature*, with results including numerous detrimental

behaviours, such as drug and alcohol consumption. This literature

explains that job loss and unemployment negatively impact well-

being, self-esteem and sense of control, resulting in diminished socio-

economic status and societal position. This decrease in status and

position can exacerbate already apparent differences between stu-

dents and further increase feelings of ostracism, isolation and failure.



NATURE HUMAN BEHAVIOUR 1, 0040 (2017) | DOI: 10.1038/s41562-016-0040 | www.nature.com/nathumbehav 5



© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.

LETTERS NATURE HUMAN BEHAVIOUR



This suggests that other factors that also increase the isolation of

individuals should be considered as research continues in this area.

Although the increasing fragility in the school-to-work transition can

explain, at least in part, an increase in the frequency of school shoot-

ings, it does not explain why such a large basal rate of gun violence

should exist at all in the United States.



Materials



Data. The Shultz’ dataset was obtained from the corresponding

author. The official reports from the Virginia Tech Review Panel”

and the National School Safety Council on School Associated

Violent Deaths** were used. Events were downloaded from Slate*,

the Brady Campaign to Prevent Gun Violence** and Wikipedia’’.

When constructing the consensus dataset, the inter-coder agree-

ment across the three coders was 88%, 89% and 84%.



School population data were collected from the National Center

for Education Statistics (https://nces.ed.gov/) using the ELSi tool

for K12 schools and the Integrated Postsecondary Education Data

System (IEPDS) tool for post-secondary schools.



National, regional and educational attainment unemployment

rates were obtained from the St. Louis Federal Reserve (https://

fred.stlouisfed.org/). City unemployment rates were obtained

from the Bureau of Labor Statistic’s Local Area Unemployment

Statistics (www.bls.gov/lau). Foreclosure rates were obtained from

the Mortgage Banker’s Association (www.mba.org) and consumer

confidence was obtained from the Organisation for Economic

Co-operation and Development*.



Data availability. The dataset presented in this manuscript is avail-

able at https://amaral.northwestern.edu/school_gun_violence/.



Received 27 July 2016; accepted 15 December 2016;

published 30 January 2017



References



1. Moore, M. H., Petrie, C. V., Braga, A. A. & Mclaughlin, B. L. Deadly Lessons:

Understanding Lethal School Violence (National Academies Press, 2003).



2. Borum, R., Cornell, D. G., Modzeleski, W. & Jimerson, S. R. What can be

done about school shootings? A review of the evidence. Educ. Researcher 39,

27-37 (2010).



3. Shultz, J. M., Cohen, A. M., Muschert, G. W. & Flores de Apodaca, R.



Fatal school shootings and the epidemiological context of firearm mortality

in the United States. Disaster Health 1, 84-101 (2013).



4. Towers, S., Gomez-Lievano, A., Khan, M., Mubayi, A. & Castillo-Chavez, C.

Contagion in mass killings and school shootings. PLoS ONE 10, e0117259 (2015).



5. Nekvasil, E. K., Cornell, D. G. & Huang, F. L. Prevalence and offense

characteristics of multiple casualty homicides: are schools at higher risk

than other locations? Psychol. Viol. 5, 236-245 (2015).



6. Kachur, P. S. et al. School-associated violent deaths in the United States,

1992 to 1994. J. Am. Med. Assoc. 275, 1729-1733 (1996).



7. Kimmel, M. S. & Mahler, M. Adolescent masculinity, homophobia, and

violence: random school shootings, 1982-2001. Am. Behav. Sci. 46,

1439-1458 (2003).



8. Twemlow, S. W. et al. Premeditated mass shootings in schools: threat

assessment. J. Am. Acad. Child Psy. 41, 475-477 (2002).



9. Lickel, B., Schmader, T. & Hamilton, D. L. A case of collective responsibility:

who else was to blame for the Columbine High School shootings?



Pers. Soc. Psychol. B. 29, 194-204 (2003).



10. Garcia-Bernardo, J. et al. Social media affects the timing, location, and severity

of school shootings. Preprint at http://arxiv.org/abs/1506.06305 (2015).



11. Newman, K. S., Fox, C., Roth, W., Mehta, J. & Harding, D. Rampage:



The Social Roots of School Shootings (Basic Books, 2005).



12. Verlinden, S., Hersen, M. & Thomas, J. Risk factors in school shootings.

Clin. Psychol. Rev. 20, 3-56 (2000).



13. Leary, M. R., Kowalski, R. M., Smith, L. & Phillips, S. Teasing, rejection,

and violence: case studies of the school shootings. Aggressive Behav. 29,

202-214 (2003).



14. Kellner, D. Guys and Guns Amok: Domestic Terrorism and School Shootings

from the Oklahoma Bombing to the Virginia Tech Massacre (Routledge, 2008).



15. Donohue, E., Schiraldi, V. & Ziedenberg, J. School House Hype: School Shootings

and the Real Risks Kids Face in America (Justice Policy Institute, 1998).



16. Vossekuil, B., Fein, R. A., Reddy, M., Borum, R. & Modzeleski, W. The Final

Report and Findings of the Safe School Initiative: Implications for the



Prevention of School Attacks in the United States (US Secret Service and

US Department of Education, 2002).



17. Heck, W. P. School shooter: one community’s experience. FBI Law

Enforcement Bull. 70, 9-13 (2001).



18. Mulvey, E. P. & Cauffman, E. The inherent limits of predicting school

violence. Am. Psychol. 56, 797-802 (2001).



19. Anderson, M. et al. School-associated violent deaths in the United States,

1994-1999. J. Am. Med. Assoc. 286, 2695-2702 (2001).



20. National Youth Gang Survey Analysis (National Gang Center, 2016); https://

www.nationalgangcenter.gov/survey-analysis/measuring-the-extent-of-gang-

problems



21. Morse, A., Sisneros, L., Perez, Z. Jr & Sponsler, B. A. Guns on Campus:



The Architecture and Momentum of State Policy Action (NASPA and

Education Commission of the States, 2016).



22. Schoon, I. & Silbereisen, R. K. Transitions from School to Work: Globalization,

Individualization, and Patterns of Diversity (Cambridge Univ. Press, 2009).



23. Brand, J. E. The far-reaching impact of job loss and unemployment.



Annu. Rev. Sociol. 41, 359-375 (2015).



24. Brand, J. E. & Thomas, J. S. Job displacement among single mothers:

effects on children’s outcomes in young adulthood. Am. J. Sociol. 119,

955-1001 (2014).



25. Johnson, R. C., Kalil, A. & Dunifon, R. E. Employment patterns of less-skilled

workers: links to children’s behavior and academic progress. Demography

49, 747-772 (2012).



26. Mohanty, M. S. Effects of positive attitude on earnings: evidence from the US

longitudinal data. J. Socio-Econ. 38, 357-371 (2009).



27. Mohanty, M. S. Effects of positive attitude and optimism on employment:

evidence from the US data. J. Socio-Econ. 39, 258-270 (2010).



28. Mohanty, M. S. Effects of positive attitude and optimism on wage

and employment: a double selection approach. J. Socio-Econ. 41,



304-316 (2012).



29, Waddell, G. R. Labor-market consequences of poor attitude and low

self-esteem in youth. Econ. Ing. 44, 69-97 (2006).



30. Robinson, W. S. Ecological correlations and the behavior of individuals.

Am. Sociol. Rev. 15, 351-357 (1950).



31. The Rising Cost of Not Going to College (Pew Research Center, 2014).



32. Danziger, S. & Ratner, D. Labor market outcomes and the transition to

adulthood. Future Child. 20, 133-158 (2010).



33. Mass Shootings at Virginia Tech April 16, 2007: Report of the Virginia Tech

Review Panel Appendix L (Virginia Tech Review Panel, 2007).



34. School Associated Violent Deaths (National School Safety Center, accessed

12 October 2015); http://www.schoolsafety.us/media-resources/

school-associated-violent-deaths



35. Kirk, C. Since 1980, 297 people have been killed in school shootings. Slate

http://www.slate.com/articles/news_and_politics/map_of_the_week/2012/12/

sandy_hook_a_chart_of_all_196_fatal_school_shootings_since_1980_map.

html (2012).



36. Major School Shootings in the United States since 1997 (Brady Campaign to

Prevent Gun Violence, 2012).



37. List of School Shootings in the United States (Wikipedia, accessed 15 February

2015); https://en.wikipedia.org/wiki/List_of_school_shootings_in_the_

United_States



38. Consumer Confidence Index (CCI) (Indicator) (OECD, 2014); http://dx.doi.

org/10.1787/46434d78-en



Acknowledgements



We would like to thank D. Figlio and B. Carruthers for their thoughtful consideration

and feedback on an early version of this work. A.J.H. would like to acknowledge

financial support from the Northwestern University Presidential Fellowship.



The funder had no role in study design, data collection and analysis, decision to

publish or preparation of the manuscript.



Author contributions



A.R.P., J.H. and L.A.N.A. developed the research question. A.R.P., A.L.J., AJ., K.A. and

A,J.H. collected, validated and annotated the data. A.R.P. and A.L.J. conducted the initial

analyses. A.R.P. conducted the primary analyses and supporting analyses. A.R.P. and

L.A.N.A. wrote the manuscript. A.R.P., J.-H., AJ.H. and L.A.N.A. edited the manuscript.



Additional information



Supplementary information is available for this paper.

Reprints and permissions information is available at www.nature.com/reprints.

Correspondence and requests for materials should be addressed to L.A.N.A.



How to cite this article: Pah, A. R. et al. Economic insecurity and the rise in gun

violence at US schools. Nat. Hum. Behav. 1, 0040 (2017).



Competing interests



The authors declare no competing interests.



6 NATURE HUMAN BEHAVIOUR 1, 0040 (2017) | DOI: 10.1038/s41562-016-0040 | www.nature.com/nathumbehav



© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.