Is Automation Coming for Your Criminal Justice or Public Safety Job? Think Again.

Robotics and Automation

When most of us think about automation, we think of technology advancements, assembly lines "manned" by robots, or even chatbots… Things that take the place of human beings and still accomplish basic — often repetitive — job tasks. 

Most of us would not think of automation affecting roles in the Justice arena.  But Dr. Johannes Moenius shared with us some interesting statistics that might imply that some of us might want to reconsider our assumptions.

 
Justice Clearinghouse Editors: Tell us about your organization, The Institute for Spatial Economic Analysis at the University of Redlands, and the work you do regarding understanding the impacts on automation and jobs. 

Dr. Johannes Moenius : The Institute for Spatial Economic Analysis (ISEA, pronounced as I-see) was born out of the need during the Great Recession to provide localized information for decision makers. For example, city-level house price indices did not really capture what was going on in the market on the neighborhood level: we found price swings in some neighborhoods to be almost double the average price increase from 2000 to the peak in 2007. In other words, while the average house price doubled during that period, we found some neighborhoods where house prices actually almost quadrupled. Most of those then crashed as much as they increased before, leading to highly localized negative equity.

We did topical research on these phenomena from 2009 onward. Starting in 2011, we provided systematic analysis of issues related to employment, wages, job turnover, income inequality, industry concentration and economic risks on the national level, and retail developments as well as house prices on the California and Southern California level, respectively. We always try to break down the data to the smallest geographic unit that we can reasonably provide estimates for, which is usually the zip code level. Most of this analysis is free to the public on our website. We have a map-based interface where people can find out, for example, how large job growth was in their zip code, what their largest sector is, and how this sector is positioned in their zip code relative to the rest of the nation. They can even create individualized reports for any collection of zip codes they are interested in.

Robotics and AutomationDr. Jess Chen, the lead author of our automation study, and I got interested in the effects of automation on jobs when we realized that there was nothing out there that addressed what automation really would mean to communities. We then found out that we could break down the analysis that was generally done on the national level to the geographic and demographic detail that decision makers need: How much will my region be affected? What are the demographic groups most at risk? Where within my region will the effects of automation be felt the most? 

The more we dug into this topic, the more we understood that the effects of automation, despite all the progress in the past, were hardly tangible today, but had the potential to – and we actually believe it will – to disrupt professional life, socio-economic systems worldwide, and could possibly even turn the world economic system the way we know it upside down. We can think of it almost like a tsunami: you may barely feel the earthquake preceding it as a slight shaking under your soles, but when the wave hits, it may take all your possessions away, if not more.

 

~~~~~

We can think of [automation] almost like a tsunami:

you may barely feel the earthquake preceding it as a slight shaking under your soles,

but when the wave hits, it may take all your possessions away, if not more.

 

~~~~~

 

JCH: When most of us think about automation, we think about manufacturing or extremely repetitive jobs. Is that necessarily a complete picture of what automation is or where it’s headed?

Johannes: Historically, it is correct that automation was largely limited to highly repetitive jobs. This is about to change due to the huge progress made in Artificial Intelligence and sensor technology. Virtually every smartphone has artificial intelligence and a multitude of sensors loaded onto it.

With billions of them now in the hands of consumers, this allows for increasingly large investments specifically into AI applications. Think household robots, which are probably the most similar to the robots used in the service sector: according to the 2016 World Robotics Report from the International Federation of Robotics, in 2015 3.7 million household robots were sold. This number is expected to jump to almost 31 million by 2019.

This means household robot sales will grow at 70% per year during that period. Even if we just assume growth rates after 2019 similar to those we currently see in industrial robots, we can expect annual sales of about 220 million units per year in 2035.

So the key to understanding these phenomena is that the recent advances in robotics and AI not only make those technologies continuously more affordable. The key is that those advances no longer restrict robots to the manufacturing sector.

Since 90% of employees in the US are employed outside of manufacturing, the potential for growth of robotics and AI in these sectors is even larger, leading to even faster cost reductions for robots both inside and outside of manufacturing.

Additionally, hardware in many AI applications is less costly than entire robots. Informational AI software such as Siri, Google Assistant, Alexa or Cortana all use hardware not exclusively designed for their use, such as cell phones, tablets, or computers. Self-driving cars and trucks don’t need much more than steering units as many of the actuators are standard equipment in cars and trucks already. Vacuum cleaners and lawn mowers only need mobility and added intelligence. Given the expected high speed of this transformation, it will be, to say the least, a gigantic challenge for the economy to create additional or entirely new jobs at the same pace as robots can replace existing jobs. 

 

~~~~~

Since 90% of employees in the US are employed outside of manufacturing,

the potential for growth of robotics and AI in these sectors is even larger,

leading to even faster cost reductions for robots both inside and outside of manufacturing.

 

~~~~~

 

JCH: Our members are part of the justice community – public safety. So, we have everyone from police officers to probation, prison guards to campus safety officers; to the courtroom personnel who help manage court proceedings and judges; to prosecutors, attorneys, victim’s advocates… How could automation really impact what our members are doing?

Johannes:  A whole lot: some legal and public safety jobs are actually highly susceptible to automation, although most of them are not.

Robotics and AutomationFor example, according to Frey and Osborne (2013), paralegals face a 94% chance that it is technically feasible to automate their jobs in the next 20 years. This probability is less than 1% for Forensic Science Technicians, police officers, and their supervisors. But even parking enforcement workers face an 84% probability of automation, correctional officers and jailers a 60% probability, and detectives and criminal investigators a 34% probability.

And technology will assist those working in the professions at risk mentioned in such a way that each one of them can do the work of many. Parking enforcement officers may deploy robots in streets to scan location and specifics of vehicles and only report back vehicles in violation of codes, together with a ticket ready to be printed. The officer only needs to verify the information and, if she cannot reach the owner electronically, put the ticket under the windshield wiper. Detectives will receive intelligently filtered and linked up information to follow only the more promising routes of investigation.

So two main channels will be assisting and replacing, but there is a third channel that might play out substantially if politics, as well as society as a whole, does not react: income inequality and unemployment.

As robots take over jobs, those who tell robots what to do will make more money, and those who are getting told by machines what to do will either lose their jobs or at least make less money. This will lead to unemployment and increases in income inequality. Previous research has shown that increases in unemployment and income inequality both have increased crime rates substantially. So I would expect certain public safety and legal profession to see increased demand for their services.

 

~~~~~

Paralegals face a 94% chance of automation in the next 20 years.

Parking enforcement workers face an 84% probability of automation,

correctional officers and jailers a 60% probability.

Detectives and criminal investigators a 34% probability.

 

~~~~~

 

JCH: Knowing that your research shows how automation impacts specific communities – is there a way you could look at your data a little differently? Could you look at the justice profession for us? See how automation will affect our industry/sector? 

The bureau of labor statistics lists a little more than one million people that work in legal professions directly, which includes judges, lawyers, judicial law clerks, arbitrators, paralegals and so on. Across legal professions, the average automation probability (weighted by employment numbers) is a relatively low 36%, with Title Examiners, Abstractors and Searchers facing the highest risk with an almost sure possibility of having their jobs automated away, and Lawyers as well as arbitrators facing almost no risk at all.

These numbers cover all legal professionals, regardless whether hired by the government or the private sector. Of course, legal services rely on many other professions, from IT services to managers. The automation probability for the legal services industry (NAICS code 5411) across all jobs is about 58%, with roughly 35% of the total wages in this industry being affected.

If we add protective service occupations, community and social service occupations as well as some other occupations directly linked to the justice system, one might come up with a list of occupational groups as follows in the table below. We included the 2016 national employment in those occupations, their average wage as well as their probability of being susceptible to automation.

Occupational Group Name

Employment in Occupation, National

Annual Wage in Occupation, National

Probability of Automation

Forensic Science Technicians

14,800

$60,690

1.0%

Substance Abuse and Behavioral Disorder Counselors

91,040

$44,160

3.3%

Educational, Guidance, School, and Vocational Counselors

260,670

$57,620

0.9%

Mental Health Counselors

139,820

$46,050

0.5%

Child, Family, and School Social Workers

298,840

$47,510

2.8%

Healthcare Social Workers

159,310

$55,510

0.4%

Mental Health and Substance Abuse Social Workers

114,040

$47,880

0.3%

Health Educators

57,570

$57,900

4.5%

Probation Officers and Correctional Treatment Specialists

87,500

$55,380

25.0%

Social and Human Service Assistants

360,650

$34,120

13.0%

Clergy

49,320

$49,450

0.8%

Directors, Religious Activities and Education

20,590

$44,840

2.5%

Lawyers

619,530

$139,880

3.5%

Judicial Law Clerks

13,410

$59,840

41.0%

Administrative Law Judges, Adjudicators, and Hearing Officers

14,540

$95,240

64.0%

Arbitrators, Mediators, and Conciliators

6,300

$72,730

6.0%

Judges, Magistrate Judges, and Magistrates

27,210

$115,460

40.0%

Paralegals and Legal Assistants

277,310

$53,180

94.0%

Court Reporters

17,700

$56,940

50.0%

First-Line Supervisors of Correctional Officers

43,230

$65,100

2.5%

First-Line Supervisors of Police and Detectives

100,200

$88,400

0.4%

First-Line Supervisors of Fire Fighting and Prevention Workers

57,170

$77,050

0.4%

Firefighters

315,910

$50,520

17.0%

Fire Inspectors and Investigators

11,910

$61,660

48.0%

Forest Fire Inspectors and Prevention Specialists

1,650

$44,300

4.8%

Bailiffs

17,880

$45,740

36.0%

Correctional Officers and Jailers

431,600

$46,750

60.0%

Detectives and Criminal Investigators

104,980

$81,490

34.0%

Fish and Game Wardens

6,610

$54,760

8.0%

Parking Enforcement Workers

8,920

$39,650

84.0%

Police and Sheriff's Patrol Officers

657,690

$62,760

9.8%

Transit and Railroad Police

4,810

$67,850

57.0%

Animal Control Workers

12,970

$36,600

21.0%

Private Detectives and Investigators

28,490

$53,530

31.0%

Gaming Surveillance Officers and Gaming Investigators

10,460

$35,280

95.0%

Security Guards

1,103,120

$29,730

84.0%

Lifeguards, Ski Patrol, and Other Recreational Protective Service Workers

145,100

$22,640

67.0%

Police, Fire, and Ambulance Dispatchers

95,170

$41,070

49.0%

Dispatchers, Except Police, Fire, and Ambulance

197,910

$41,190

96.0%

Sources: Bureau of Labor Statistics, Frey and Osborne (2013)

 

Averaging those out, this implies that the legal and public safety professions face a 35% automation risk, with 25% of the wage bill going away – assuming that no extra police officers and legal workers are needed. Considering the larger socioeconomic challenges, this really seems unlikely to me. Moreover, those numbers may seem large, but compared to the national averages across industries, these are actually quite low and close to the bottom of the distribution.

 

~~~~~

The good news is, many jobs in the justice community are safe,

and I would even expect increased hiring in many of those fields for the reasons given above.

~~~~~

 
JCH: This might be a bit surprising for some of our members: how should our members think about this news? 

Johannes: The good news is, many jobs in the justice community are safe, and I would even expect increased hiring in many of those fields for the reasons given above.

For those that will likely see their jobs going away, it may help to think about it in the following manner: there is still time till the major waves of automation will hit the legal profession. So this allows for getting additional education or even retraining for jobs with lower automation risk. But it is essential to start preparing for it today, and I mean this literally.

Five years is a long time in the development of AI – Siri, the assistant on the iPhone, has not been in existence much longer than that. Imagine what Siri will be able to do 5 years from now. Will it be able to make your job easier … or even do much of it? If I was working as a paralegal, I could imagine the answer to be closer to “yes” than to “no”. So start preparing now.

If you think your job is safe, think again. The automation probabilities used in our work originally were determined by Oxford Professors based on expected technological feasibility in 2013. Since then, new technologies have emerged, smartphone sales increased by a staggering 50%, substantially reducing the per unit cost of AI applications.

So everyone in their 20s and 30s, regardless of profession, should start thinking of what makes his work unique – or how it could be made unique, hard to replace by robots, and valuable to society.   

 

 

Additional Resources
Join the Justice Clearinghouse Community of over 23,309 Justice Practitioners!

Join the Justice Clearinghouse Community of over 23,309 Justice Practitioners!

3-5 times per week we will send you updates on free upcoming webinars, custom created infographics and interviews with our presenters

You have Successfully Subscribed!

X