U.S. Parks Research

NSF Science, Engineering and Education for Sustainability (SEES) Postdoctoral Research

Kelly Research and Outreach Lab, Department of Environmental Science, Policy and Management, University of California, Berkeley

Mount Rainier National Park (WA)

Multi-Disciplinary Study of Crime in US National Parks

US National Park units see crimes that are both characteristic of small-town or urban policing (e.g. domestic violence, assault, rape, theft, and even murder), as well as crimes that are unique to the stunning landscapes and cultural resources these areas protect (e.g. stolen moss, redwood burls, artifacts, and cacti as well as poached wildlife, defaced rock walls and historic structures).  Understanding where these crimes are happening, and how well they may be detected is integral to service-wide considerations of resource allocation and future planning.

This Political Ecology (PE) research stands at the intersection of critical geography, sociology, political economy, and GIscience, using a suite of qualitative and quantitative data collection and analytic methods to best understand park crime statistics.  

This work was done in partnership with the National Park Service's Pacific West Regional Office and the UC Berkeley Kelly Research and Outreach Lab.  While I was the lead researcher on this project, I worked closely with Maggi Kelly (no relation) and Jenny Palomino on data collection and research design.  

A Note on Crime and Criminals

Crime and criminals are not fixed, or pre-determined categories.  Instead, what is a crime and who is a criminal is often defined by power-dynamics, politics, and economics.  While my work in northern Cameroon (see Past Research) grapples with these definitions extensively, my work on US National Parks uses Federal and State Law uncritically for the purposes of this primary analysis.  In future research, I hope to deal with the ways in which crimes and criminals are and have been defined in relationship to protected areas in the United States.  I also hope to examine how and in what ways laws may be enforced differently under different circumstances.  


Haleakala National Park (HI) 

Iterative & Participatory Research

To best understand what factors influence park crime statistics we interviewed Chief Rangers and field Rangers in parks across California.  Through these conversations and ride-alongs on Ranger patrols we compiled 90 possible factors that influence recorded crime in parks ranging from geophysical factors, to officer well-being and safety, to park relationships with adjacent agencies and internally.  

We shared these 90 factors back with the Rangers we surveyed, as well as with Rangers across the country in a variety of different parks and positions in the National Park Service.  We asked 30 Rangers to rank the importance of each metric in influencing crime statistics in US National Parks.  Twenty-five of these rangers responded, and on average each of the metrics was ranked as important to very important.  Only one metric was deemed unimportant.  Some Rangers suggested additional metrics or ways to expand on existing metrics.  

While time consuming, we believe that taking the time to understand the constraints on policing National Parks first-hand and to incorporate Ranger input into our final group of factors to be studied has made this research more robust, and will make its results ultimately more useful to the National Park Service.  

Santa Monica National Recreation Area (CA)

Understanding Park Crime Statistics

We have grouped the factors we are examining into three different groups--ranger capacities, issues internal to the park, and issues external to the park.  While these factors are currently based on anecdotal evidence from Ranger interviews and surveys, we are currently collecting data on all of these factors to test how and in what ways they influence park crime statistics.  To check out some visualizations of the data we have collected so far, go to our Data Collection Page.  To check out collaborators on this project, go to our Collaborators page. 

Ranger Capacities

Charged with visitor and resource protection in areas as disparate as Yosemite National Park, Golden Gate National Recreation Area, and Lake Mead, National Park Service Law Enforcement Rangers (LE Rangers) must:

  • Adapt their work to the natural, social, and political landscape, visitor-types, and park infrastructure at hand, and
  • Balance their law enforcement duties with those of Emergency Medical Services (EMS), Search and Rescue (SAR), Wild-land and Structural Fire, as well as their collateral duties and trainings to maintain industry standards in all of these activities.

Thus, the degree to which LE Rangers are able to actively patrol, how secure they feel in those patrols (due to communications or back-up availability), how they handle or record offences, how arrests are handled by the US District Attorney's office, and how much state, county, or local law enforcement agencies operate in their parks determines how crimes in different parks are recorded.  

Park-Specific Factors

What types of crime, and how many crimes are recorded in parks can also be contingent park's cultural histories as well as their geographical, and biophysical environments.  Some parks have highly attractive cultural resources like Native American and early-settler artifacts that are prime targets for theft and may attract people to the park for these purposes.  Other parks may have attractive natural targets like prized tree species, moss, salal, or commercially valuable minerals and metals.  Parks' roads and trail networks, their infrastructure, concession operations, as well as the activities they allow or forbid visitors to partake in also influence their visitation and use which may also contribute to different levels and types of crime.  

Neighborhood Effects

Park crime statistics are influenced by their surrounding areas as well.  The types of landownership adjacent to the parks, the number of land owners, and the uses on those lands can either act as a buffer for park landscapes, or invite illicit use.  The presence of alternative diversions like recreational lakes, ATV parks, or lands available for plant collection or mining may relieve some pressure on park landscapes.  The populations around parks make a difference too--the crime rates within these communities, the socio-economic make-up of residents, the presence of gangs or extremist groups can all make a difference in the rates and types of crimes parks experience.  


It is, of course, not a single factor, or a single group of factors that influence park crime statistics.  Rather, we seek to understand how these factors interact in different ways with varying results.  Just because a National Park is directly adjacent to a large population center does not necessarily mean that park will have higher rates of recorded crime.  Who is actually visiting the park and/or living directly adjacent to it, and what attractive targets the park may have all influence the rates of crime within the park landscape.  

Yosemite National Park (CA)

Yosemite National Park (CA)

Historical Analyses

Because I am somewhat of a newcomer to research in US National Parks, I have enriched my understanding of law enforcement in US National Parks by delving into its historical roots.  Interestingly, there has been no rigorous documentation of the history of National Park Service Law Enforcement so I have written a paper on this topic.  The paper outlines the uneven progression of professionalization and militarization of US Park Rangers over time, seeking to explain why Park Rangers continue to be the most assaulted federal law enforcement officers, and why visitors are continually surprised when they see a Park Ranger wearing a gun-belt.  

Contemporary Analyses

Policing National Parks is an incredibly difficult task due to these parks' sometimes extreme environments, rugged landscapes, and vast expanses of land.  Interested in news articles and discussions around the use of drones and remote sensing for the purposes of policing, Maggi Kelly and I undertook a review of the current state of the literature on this topic.  In particular, we were interested in the validation of "crimes" sighted using remote-sensing technologies.  We were surprised to find that validation of these observed offenses was rarely discussed in detail, and often not mentioned at all.  In reaction to this disturbing revelation, Maggi and I propose some methods of validating remotely sensed crime, and admonish users of these remote sensing technologies to practice caution.  

Whiskeytown National Recreation Area (CA)