Skip Navigation
Expert Brief

School Surveillance Zone

School districts are spending more on social media monitoring technology, but there is little evidence it is keeping students safer.

Published: April 30, 2019

School District Social Media Monit­or­ing, 2012–2018

As repor­ted in Smart­Pro­cure

 

 

  students (approx.)

Purchases

Year
Value
2012
 
2013
 
2014
 
2015
 
2016
 
2017
 
2018
 
Total
$654,726

Click the areas for detailed data

Year
Company
Value
2012
 
 
2013
 
 
2014
 
 
2015
 
 
2016
 
 
2017
 
 
2018
 
 
Total
 

Click the areas for detailed data

Down­load purchase records

Back to U.S. map

Hover on circles on the map to see the data5W Infographic

While stat­ist­ic­ally rare, school shoot­ings are trau­matic for those directly affected, their loved ones, and soci­ety at large. A recent spate of campus shoot­ings has led to renewed calls for gun control, often student-led. Legis­latures and schools are facing mount­ing pres­sure to address the issue, but gun control appears out of reach in the current polit­ical climate. In the mean­time, school offi­cials and poli­cy­makers have turned to a broad range of surveil­lance tech­no­lo­gies in an attempt to prevent attacks on schools.

Social media monit­or­ing is one such form of student surveil­lance. A number of compan­ies, many of which have sprung up in the last five years, are selling soft­ware that can allegedly identify signs of viol­ence or other concern­ing beha­vior by trawl­ing chil­dren’s social media posts and other online activ­ity.

In an attempt to quantify expendit­ures on social media monit­or­ing soft­ware by school districts, the Bren­nan Center examined contracts for such soft­ware using Smart­Pro­cure, a data­base of govern­ment purchase orders. Our review is based on self-repor­ted procure­ment orders in the data­base, and thus likely depicts only a portion of school spend­ing on these tools. Accord­ing to these data, school spend­ing on social media monit­or­ing soft­ware has grown in recent years. As the graph below indic­ates, the data­base shows 63 school districts across the coun­try purchas­ing social media monit­or­ing soft­ware in 2018, up from just six in 2013 — more than a tenfold increase. This is, to be sure, still a small number compared to the over 13,000 school districts across the coun­try. But it likely does not tell the whole story. Social Sentinel itself has touted the fact that it serves “thou­sands of schools in more than 35 states” – a far larger number of states, and thus more school districts, than are reflec­ted in Smart­Pro­cure’s data­base. And some school districts known to have purchased social media monit­or­ing tech­no­logy repres­ent tens or even hundreds of thou­sands of students; Orange County, FL, which purchased these tools in 2015, 2016, and 2018, has more students than Little Rock, Arkan­sas, has people.

Aside from anec­dotes promoted by the compan­ies that sell this soft­ware, there is no proof that these surveil­lance tools work. But there are plenty of risks. In any context, social media is ripe for misin­ter­pret­a­tion and misuse. But the possib­il­ity of misin­ter­pret­a­tion is partic­u­larly high for middle school and high school students, who are more likely to use slang and quotes from pop culture, and who may be espe­cially motiv­ated to evade adults’ prying eyes. Diffi­culties in inter­pret­a­tion mean that social media monit­or­ing of students is likely to lead to false posit­ives. Moreover, monit­or­ing programs are partic­u­larly bad at correctly under­stand­ing languages other than English and even non-stand­ard English, which may be used by minor­ity students.  And it is well known that school discip­line falls more heav­ily on chil­dren of color. These factors suggest that social media monit­or­ing tools are likely to dispro­por­tion­ately tag students of color as danger­ous and that those students will be punished more severely than white students who are simil­arly iden­ti­fied. Over­all, research shows that as school secur­ity meas­ures prolif­er­ate, students often feel less safe. And over­broad and unne­ces­sary surveil­lance is likely to have a detri­mental impact on students’ privacy and chill their abil­ity to express them­selves.

School District Social Media Monit­or­ing Spend­ing, 2012–2018

As repor­ted in Smart­Pro­cure

 

Hover on circles in the graph to see the data5W Infographic

Despite these consid­er­able risks, dozens of compan­ies new and old are step­ping in to take advant­age of schools’ and parents’ fears about the possib­il­ity of viol­ence at school. Law enforce­ment agen­cies have used tech­no­logy from some of these compan­ies, like Digit­alStakeout and MediaSonar, to monitor for “threats to public safety.” But in 2016, public records requests revealed that police were using these tools to surveil lawful protest­ors, lead­ing major social media plat­forms to block data access by compan­ies build­ing surveil­lance tools for law enforce­ment. Smart­Pro­cure’s purchas­ing data suggests that in recent years, these compan­ies’ share of the school surveil­lance market has been surpassed by compan­ies that focus solely on schools, such as Geo Listen­ing and Social Sentinel. These compan­ies provide an extens­ive suite of services, includ­ing track­ing social media, blogs, and forums. They also couch their services in the language of student well­ness and mental health, claim­ing they can detect bully­ing and suicidal tend­en­cies as well as acts or threats of viol­ence.

Social media monit­or­ing soft­ware is relat­ively afford­able; accord­ing to our data, the median annual expendit­ure was $8,417 in 2018. But even a moder­ate expendit­ure can enable fairly extens­ive surveil­lance of students, with little guar­an­tee that the money is making students and schools safer.

In addi­tion, Smart­Pro­cure records show that the largest outlays are clustered within a small number of states (although this may be a result of the self-repor­ted nature of the data). While districts in at least 20 states across the coun­try have docu­mented their purchases on Smart­Pro­cure, four states stand out among the purchases repor­ted to the data­base: Texas ($654K), Illinois ($321K), Flor­ida ($258K), and Cali­for­nia ($197K). In 2018, there was also a notable increase in purchases for Social Sentinel through a district services center or a regional organ­iz­a­tion rather than directly from the company. While this consor­tium pricing gives small districts purchas­ing power for educa­tional tech­no­logy that would other­wise be out of reach, it also magni­fies the risk that local parents and other stake­hold­ers will be in the dark about surveil­lance tools being imple­men­ted in their schools.

Social media and inter­net monit­or­ing are only part of the burgeon­ing school surveil­lance system. One school offi­cial has stated that “every school district across the state of Texas has been given some sort of direc­tion to make these type[s] of moves.” Many school districts also have contracts with compan­ies that integ­rate online monit­or­ing of students with other “digital safety solu­tions.” Gaggle and Securly, for example, both monitor school-provided Gmail and affil­i­ated applic­a­tions (such as Google Drive, Google Docs, Google Calen­dar, and Contacts). These compan­ies monitor content within the school’s network, includ­ing emails, texts, and networked docu­ments. They are not included on the map because their focus is on school devices and networks rather than chil­dren’s social media use (though they can include social media posts origin­at­ing from a school-provided email address). The contracts in Smart­Pro­cure suggest that purchases of these products are also increas­ing substan­tially: the total amount spent by school districts has more than doubled between 2013 ($3.9 million) and 2018 ($8.2 million).

2018 Purchases of Social Media Monit­or­ing Soft­ware by Company

As repor­ted in Smart­Pro­cure

 

The import­ance of keep­ing chil­dren safe in schools cannot be over­stated. But despite the surge in interest from schools, states, and parents, and the claims of its purvey­ors, there is little evid­ence that social media monit­or­ing soft­ware keeps students safer. And there are real risks that chil­dren, partic­u­larly those from communit­ies of color, will be wrongly tagged as dangers and that the free learn­ing envir­on­ment we all want from schools will be comprom­ised. Schools and school districts should not adopt tech­no­lo­gies that prom­ise far more than they can deliver while impos­ing real costs on the chil­dren they are meant to keep safe.