Spatialities of Dog Theft: A Critical Perspective

Simple Summary Dogs are considered property under U.K. law, while owners generally regard their canine companions as family. Reports that the number of stolen dogs in England and Wales rose from 1788 in 2016 to 1909 in 2017 led to public calls to change the law. Recognising that a more robust analysis of dog theft crime statistics is required, we gathered dog theft data for 2015, 2016, and 2017 from 41 of 44 police forces. This paper examines how dog theft crime statistics are constructed, assesses the strengths and weaknesses of these data, and categorises, maps, and measures dog theft changes temporally per police force in England and Wales. Our findings reveal there has been an increase in dog theft crimes, with 1559 thefts in 2015, 1653 in 2016 (+6.03%), and 1842 in 2017 (+11.43%), and a decrease in court charges related to dog theft crimes, with 64 (3.97%) in 2015, 51 (3.08%) in 2016, and 39 (2.11%) in 2017. The actual number of dog theft crimes will be higher as three forces could not supply useable data. There is a need for a qualitative study to understand dog theft crime in different parts of the country, and a standardised approach to recording dog theft by all police forces in England and Wales. We recommend classifying dog theft (or pet theft more generally) as a crime in itself under the Sentencing Guidelines associated with the Theft Act 1968. Abstract Dogs are considered property under U.K. law, while current discourses of pet ownership place canine companions as part of an extended family. This means sentences for those who steal dogs are not reflective of a dogs’ sentience and agency, rather in line with charges for those who steal a laptop or wallet. This is particularly problematic as dog theft is currently on the rise in England and Wales, leading to public calls to change the law. Recognising that a more robust analysis of dog theft crime statistics is required, we gathered dog theft data for 2015, 2016, and 2017 from 41 of 44 police forces through Freedom of Information (FOI) requests. This paper uses these data to examine how dog theft crime statistics are constructed, assesses the strengths and weaknesses of these data, and categorises, maps, and measures dog theft changes temporally per police force in England and Wales. Our findings reveal there has been an increase in dog theft crimes, with 1559 in 2015, 1653 in 2016 (+6.03%), and 1842 in 2017 (+11.43%), and a decrease in court charges related to dog theft crimes, with 64 (3.97%) in 2015, 51 (3.08%) in 2016, and 39 (2.11%) in 2017. There were police force inconsistencies in recording dog theft crime, which meant some data were unusable or could not be accessed or analysed. We recommend a qualitative study to understand stakeholder perspectives of dog theft crime in different areas, and a standardised and transparent approach to recording the theft of a dog by all forces across England and Wales. This could be achieved by classifying dog theft (or pet theft more generally) as a crime in itself under the Sentencing Guidelines associated with the Theft Act 1968.


Introduction
Under U.K. law, pets including dogs are regarded as "property" and pet theft is not classified as a specific crime in itself. Sentencing within the Theft Act 1968 is dependent on the monetary value of the stolen animal (under or above £500), and the crime is treated as a category three (fine to two years in custody) or category four offence (fine to 36 weeks in custody) in magistrates' court [1,2]. Despite this legal status, in social terms dogs are generally recognised as members of the family [3][4][5]. They are loved, cared for, and accepted as individuals with unique personalities and emotional significance [6]. As such, a significant tension between the social and political dimensions of pet ownership continues to be exacerbated, whereby the sensual experiences of pet engagement are poorly represented through the legalities of U.K. law. Such tension has influenced respective campaigns by the Stolen and Missing Pets Alliance (Sampa) [7] and Dogs Trust [8] to reform the Theft Act 1968 [1] and its associated Sentencing Guidelines [2].
Scholars of human-animal studies maintain that dogs actively shape relations between family members who change everyday practices to incorporate the needs of their dogs [9][10][11]. The rise in positive training, for example, equates to the recognition of sentience and mindedness of dogs [12], and the emotional, affective, and caring relationships they facilitate. Fox argues further that in shaping everyday familial practices animal companions are intrinsic to our "sense" of home and belonging [3]. Practices such as grooming, walking, and playing not only show the embedding of "doggy-ness" into family life but are also practices of human care for their animals [13]. Human care can similarly be expressed through a greater diversity and commercialization of pet-related commodities including "doggy spas" and groomers, pet boutiques, pet hotels, high-end nutritional food markets, pet cemeteries, and even pet activities such as "dog yoga" or "doggy dancing" [3,14,15]. While some of these commodities seem eccentric, they signify an emerging culture of care and the importance of pets to human lives in contemporary society as animals of love and affection. Furthermore, dogs also care for humans through actively embedding physical activity into their lives, providing security and safety, emotional support, and the ability to navigate safely [16][17][18]. Other studies point to how dogs also facilitate social interaction by acting as a social stimulus through making people more approachable, being a subject for idle chat of shared interests and helping provide a sense of community [19,20].
The human-dog relationship is characterised by affectionate caring practices that work relationally; both human and dog share a unique emotional bond with one another. Self-identification as "pet parents" has become a dominant discourse in western society, as has regarding dogs as "furry children" [4,5,10]. Ascribing names, feelings, and personalities may be forms of anthropomorphism, but it is a process that allows humans to relate to animals [3]. Experienced and embodied through trust, bodily gestures, and emotional investment, the ability to affect and be affected extends beyond and between human and animal bodies [21]. This can be expressed through owner's recognition of their pets as minded individuals capable of empathizing and comforting them-a form of mutual communication [22]. Haraway maintains that humans and dogs are bonded by "significant otherness", in "varied webs of interspecies dependence" [23,24]. Significantly, many people talk about the emotional difficulties of a pet's death and how it is comparable to the loss of a loved one within the family [15,25]. Framing pets as both irreplaceable and grievable [14,25] in this way shows a deeply embodied and emotional relationship of care and companionship [6].
Dog theft is a crime that exploits these relationships [26], and is on the rise in the United Kingdom. Freedom of Information (FOI) research conducted by Direct Line Insurance revealed that the number of dogs reported stolen in England and Wales rose from 1788 in 2016 to 1909 in 2017 [27]. As can be seen in Figure 1, these data are useful for identifying the "most-stolen dog breeds", information which can inform dog insurance policies and help raise public awareness through media campaigns. A similar study by Emporium Insurance showed 1712 thefts in 2015, 1803 in 2016 and 1977 in 2017 [28]. Media headlines included: "Lincolnshire dog theft capital of Britain" [29] and "Dog thefts increase by more than 200% in Dyfed-Powys area in just one year" [30]. Public support for pet theft reform grew in 2018. A petition to "Reclassify the theft of a pet to a specific crime in its own right" gained 107,353 signatures in six months and was debated in Parliament on 2 July 2018 [31]. Despite cross-party support, George Eustice MP concluded that "at the moment the Government are not convinced that we need to change the law", but stressed "that the Government interpret the latest guidance from the Sentencing Council that the theft of a pet should generally be treated as a category two or three offence". Alongside this, Eustice acknowledged "the need for statistics" [32]. The following day, Ross Thomson MP presented the first reading of the Pets (Theft) Bill in the House of Commons; its aim was "to amend the Animal Welfare Act 2006 and the Animal Health and Welfare (Scotland) Act 2006 to make the theft of pets an offence" [33].
Recognising that a more robust analysis of dog theft crime statistics is required, this paper (1) examines how dog theft statistics are constructed; (2) assesses the strengths and weaknesses of these data; and (3) categorises, maps and measures dog theft changes temporally per police force in England and Wales for 2015, 2016, and 2017.

Materials and Methods
The analyses presented within this paper are the product of a two-step process. The former involved a data search for any information on dog theft across England and Wales, both qualitative and, predominantly, quantitative sources of information. Objectively, this allowed key objectives to be answered: (1) How are dog theft statistics being constructed? and (2) What are the strengths and weaknesses of these data? The second step involved attempting to collect these data ourselves from the 44 police precincts across England and Wales, via the Freedom of Information Act, in order gain a spatial representation of the prevalence of these crimes across the date range 2015-2017. This was completed using ArcGIS software (Esri, Collin TX, USA). Maps included in this paper were created using ArcGIS ® software by Esri. ArcGIS ® and ArcMapTM are the intellectual property of Esri and are used herein under licence. Copyright © Esri. All rights reserved. For more information about Esri ® software, please visit www.esri.com. The process of each of these steps is outlined below:  [28]. Media headlines included: "Lincolnshire dog theft capital of Britain" [29] and "Dog thefts increase by more than 200% in Dyfed-Powys area in just one year" [30]. Public support for pet theft reform grew in 2018. A petition to "Reclassify the theft of a pet to a specific crime in its own right" gained 107,353 signatures in six months and was debated in Parliament on 2 July 2018 [31]. Despite cross-party support, George Eustice MP concluded that "at the moment the Government are not convinced that we need to change the law", but stressed "that the Government interpret the latest guidance from the Sentencing Council that the theft of a pet should generally be treated as a category two or three offence". Alongside this, Eustice acknowledged "the need for statistics" [32]. The following day, Ross Thomson MP presented the first reading of the Pets (Theft) Bill in the House of Commons; its aim was "to amend the Animal Welfare Act 2006 and the Animal Health and Welfare (Scotland) Act 2006 to make the theft of pets an offence" [33].
Recognising that a more robust analysis of dog theft crime statistics is required, this paper (1) examines how dog theft statistics are constructed; (2) assesses the strengths and weaknesses of these data; and (3) categorises, maps and measures dog theft changes temporally per police force in England and Wales for 2015, 2016, and 2017.

Materials and Methods
The analyses presented within this paper are the product of a two-step process. The former involved a data search for any information on dog theft across England and Wales, both qualitative and, predominantly, quantitative sources of information. Objectively, this allowed key objectives to be answered: (1) How are dog theft statistics being constructed? and (2) What are the strengths and weaknesses of these data? The second step involved attempting to collect these data ourselves from the 44 police precincts across England and Wales, via the Freedom of Information Act, in order gain a spatial representation of the prevalence of these crimes across the date range 2015-2017. This was completed using ArcGIS software (Esri, Collin TX, USA). Maps included in this paper were created using ArcGIS ® software by Esri. ArcGIS ® and ArcMapTM are the intellectual property of Esri and are used herein under licence. Copyright © Esri. All rights reserved. For more information about Esri ® software, please visit www.esri.com. The process of each of these steps is outlined below:

Data Search for Dog Theft Data
Despite the increasing prevalence of dog theft across England and Wales [26,27], our own research demonstrated that data pertaining to such thefts are either particularly coarse/vague and missing large chunks of information-or altogether lacking. An initial data search brought two key sources of dog theft datum to our attention: one compiled by Direct Line Insurance and one compiled by Emporium Insurance. Both data sets were explored in order to try and deconstruct how dog theft statistics are calculated on the national scale and, indeed, to explore the relative strengths and limitations of each data set.
The data supplied by Direct Line and Emporium had followed similar methods of data collection to our own. Via the Freedom of Information Act the company had sent requests to each of the 44 police forces across the England and Wales. In the case of Direct Line, 41/44 forces responded to their requests for the years 2016-2017. The data provide a useful picture of the number of dogs stolen by region and further provided a list of the top ten breeds of dogs which are stolen. However, there were a number of issues. Firstly, the information sheet which is attached to these data groups the data on a regional basis, despite the geography being provided via each force. Similarly, no graphical or cartographical information is provided. More problematically, however, the data focus on the number of dogs stolen and not the number of crimes, failing to provide a breakdown as to the crime rate for each force. The number of dogs taken cannot be representative of the crime level as any number of dogs could be taken at one time, particularly given the increase in crime targeting dog breeders where multiple dogs can be taken during one burglary.
Comparatively, the data provided by Emporium Insurance included 38/44 police forces across England and Wales for number of dogs stolen. Only 26/44 police forces provided data for number of crimes, with three of those 26 sources being particularly coarse. There was greater detail in the different breeds which had been taken and the data did provide a figure as to the number of crimes that had occurred, as well as the number of dogs that were taken. This again illustrates the above point regarding using the number of dogs stolen as an indicator; their data show that, for example, in 2015 in the West Yorkshire police force jurisdiction, 184 dogs were stolen, but only 164 crimes were recorded as having occurred, misrepresenting the crime rate.
Thus, such an analysis of existing data reveals a three-fold problem with the ways in which national statistics related to dog theft are being calculated. On the one hand, there is data for some police forces missing entirely, meaning the statistics are not fully representative of the dog theft situation across England and Wales. Of more prominent concern, however, is significant discrepancy between the number of forces' data collected. Given each company devises their own statistics based on their respective data, then it becomes increasingly problematic to track a true, definitive, national picture of this issue. Finally, the use of the number of dogs stolen as a representation of the amount of crime is also problematic, as using data of this type neglects to consider the quantities of dogs that could be taken. The number of dogs is not equivalent to the number of crimes being committed. Given we have used data from two key insurance companies, if other companies are generating similar statistics with varying levels of information, then there is a clear need to group these data together in order to develop a more detailed, singular and national focus on dog theft. Indeed, this comprises the focus of stage two of the methods.

Data Collection and Geographic Information System (GIS) Analyses
In order to begin to further deconstruct the notion of "pet theft"-its respective characteristics and spatial prominence within England and Wales-we sent our own FOI requests to each of the 44 police forces across England and Wales. We asked two distinctive questions in relation to dog crime: What was the total number of dog theft crimes in 2015/2016/2017? What was the outcome (charge/summons, community resolution, active investigation, evidential difficulties, no suspect identified) for each dog theft crime in 2015/2016/2017?
We received 41/44 responses for the police forces across England and Wales. Some of the data from different forces were more detailed than others, whereby some forces would provide just a yearly figure as to how many thefts the system had returned upon using the key word "dog" in the forces' system search, whilst others provided a breakdown of, for instance, where the crimes had been committed (i.e., in a residential building, in a car, or in a public setting). However, this level of detail was seldom given by many police forces. Hampshire, Sussex, and Wiltshire police forces were unable to provide any data or did not reply to requests under freedom of information legislation. Initially, Humberside Police force was able to provide some information; however, they admitted that these data were limited as the search had been stopped due to the cost/time it would take for these forces to gather the data: "Although excess cost removes the forces obligations under the Freedom of Information Act, as a gesture of goodwill, I have supplied information, relative to your request, retrieved or available before it was realised that the fees limit would be exceeded. I trust this is helpful, but it does not affect our legal right to rely on the fees regulations for the remainder of your request." A later request, however, led to a more detailed response. Thus, we were able to compile data for 41/44 forces for dog theft crime.
The data from each force were compiled in to one larger database-a lengthy process given the different ways the data had been presented to us. This process entailed sorting the data in to categories, through which some categories, those in Table 1, were created and represent an amalgamation of different responses (individually described below). We deemed this exercise important because, as previously described, the data had some discrepancies. For example, the data we were given were entered under the sub-heading "total claims of dog related crime" for each of the three years. However, this was not taken as the absolute value as there were instances where a "dog crime" was provided, but was not technically a theft. For instance, a lack of detail in data from West Yorkshire in 2017 prevents breaking the data down and therefore the assumption is that these were all thefts, whereas data from the West Midlands in 2017 indicate one recording of dog-related crime was fear/ provocation of violence, but this is not technically theft and therefore is not included in the total thefts category. Thus, the data were sorted as best as possible in order to "clean" them. We do, however, acknowledge that despite our best efforts the data are by no means perfect. Table 1. Any "merged" categories that were made during the data sorting process. These were rational decisions made to manage the data given the vast differences between how the data are recorded or how they were presented to us.

Theft in public:
Any instance where the pet was taken in the public, such as at an ATM or in a store. This does not include thefts taking place in vehicles (separate classification, below) or thefts taking place within buildings deemed as businesses.
Theft of/from a vehicle: Any instance where a pet was taken from a car, or was within a car when the vehicle was stolen.
Evidential difficulties: This includes instances where either the police deemed there not to be enough evidence to proceed, thus closing the case, or where the public were willing to cooperate but could not provide evidence.
Withdrawn support/unwilling to assist: Instances where the public could not or did not cooperate with the police investigation, therefore closing the case. This is separated from evidential difficulties as it could demonstrate false reporting, such as a domestic dispute over pet ownership etc.
Not in public interest (police decision)/prevented: This includes instances where the crime was classified as "prevented"-the assumption is that the crime did not happen. It also includes instances where the police deemed the prosecution as not being worth pursuing, usually because the dispute was settled. Instances where "no further action" was listed in the data, where the case was cancelled or transferred by the force, or where the case was falsely recorded.
Penalty notice/caution/other: Any instance where a penalty notice or fine was implemented as punishment, where a police caution was given instead of an arrest, or where a court summons is recorded but the outcome of which is not described further. This also includes categories such as youth restorative programmes, extended professional opinions, situations where complaints were made, or where another force had primacy jurisdiction. It also includes court disposal of cases or instances where prosecution time was marked as "expired".

GIS Spatial Mapping
Once the data had been managed we then categorised them using seven classes chosen due to the best number for representing colour categories ( Table 2 below). This measured the number of reported crimes in quantities of 39. Maintaining these classes enabled comparisons of how the number of dog thefts has risen or fallen for each police force over the years of interest to the study. The objective here was not to compare the data between forces, as despite sorting the data there were unavoidable biases relating to how the data had been managed or input in to the system (owing to the lack of treating pet theft as a crime in its own right), but to show the spatial variation registered dog thefts by different forces. Once classes had been assigned, the database was imported in to ArcMap and joined with the Police Force Areas (December 2016) Shapefile provided by the Office for National Statistics [34]. The same colour symbology was adopted consistently for each of the categories across each of the years.

Crime Rates
For crime rates to be calculated, the estimated population of police force areas were identified from the Office for National Statistics for 2015, 2016, and 2017. The annual number of dog theft crimes per force were divided by the estimated population then multiplied by 100,000.

Theft in England and Wales
Dog theft crime is not recognised by the Office for National Statistics on their "Crime in England and Wales: Police force area data tables". Therefore, FOI requests are the only way to access these data. Table 3 Table 3. It is important to situate our research with broader theft offences within the United Kingdom.

Dog Theft in England and Wales
In 2015, the forces with the highest numbers of dog theft crimes (DTCs) were the Metropolitan police (167) and the West Yorkshire (166), Greater Manchester (120), Kent (102), and Essex (74) police. The lowest number of recorded dog theft crimes was in Cheshire (6). This is indicated on the map in Figure 2. The dark red colour specifies 160+ DTCs (i.e., Metropolitan and West Yorkshire police), whereas the pale orange specifies 0-39 DTCs (i.e., Cheshire police). The City of London had no recorded dog theft crimes-this was not mapped as the force only covers a 2.8 km 2 area. The blue shows the police force areas in which we had no data (i.e. Wiltshire police). Overall, 25 police forces recorded between 0 and 39 dog theft crimes. Furthermore, Figure 2 Table 4.
The rates of dog theft crime charges also varied. In Greater Manchester there were six charges from 120 dog theft crimes (5%) whereas in Northumbria there were 44 dog theft crimes with six charges (13.64%). The FOI requests do not, however, make it clear whether charges for dog theft crime relate to multiple dog thefts or one. This is important as this can misconstrue the crime rate. FOI requests were made to the Ministry of Justice to access information on sentences-their response stated "centrally held information cannot identify dog theft from other theft. Therefore any request for this information is likely to require a manual search of all sentenced cases related to theft". With "over 77,000 sentences handed down by courts in England and Wales cases related to theft", accessing these data is near impossible [36].
Exploring DTCs per 100,000 people (DTC rates), we found that West Yorkshire ranked highest (7.27), followed by Kent (5.66), Gwent (5.15), Staffordshire (4.93), South Wales (4.74), and Greater Manchester (4.35). The lowest rates of DTCs per 100,000 were in Cheshire (0.57) and City of London (0). The Metropolitan police (1.92) had a relatively low rate of DTCs per 100,000 compared to the number of dog theft crimes (167). While the Metropolitan police had highest DTC crimes in 2015 and currently the most police officers per 100,000, the DTC rate per 100,000 was low due to the Metropolitan area having the greatest population in England and Wales. On the other hand, Staffordshire (4.93) had a relatively high rate of DTCs per 100,000 compared to the number of dog theft crimes (55). They are ranked 41st for police per 100,000 people. Overall, the police force data available to us through FOI requests showed 1559 dog thefts nationally in 2015. The outcome of 853 cases was "no further action" (54.71%), and there were 64 charges (3.97%).    In Northumbria there were seven charges from 61 DTCs (11.48%), compared to Greater Manchester with three charges from 132 DTCs (2.27%). Again a similar issue arises as FOI requests do not, however, make it clear whether charges for dog theft crime relate to multiple dog thefts or one. This is important as this can misconstrue the crime rate.
In 2016 West Yorkshire retained the highest-ranking DTC rate per 100,000 (8.58), followed by Lancashire    Our findings, in Table 5  The numbers of people charged also varied. While Kent recorded 130 dog theft crimes in 2017, there were only four charges (3.07%). In Bedfordshire there were 20 dog theft crimes, with an outcome of four charges (20%). There is a vast difference between the number of dog theft crimes and the number of people charged for these Kent and Bedfordshire police forces. However, there is little difference between the number of police officers per 100,000 people (17th and 22nd, respectively) -as shown on Table 4.
The highest dog theft crime rates in 2017 per 100,000 were West Yorkshire (7.45), Kent (6.72), Lancashire (6.23), Staffordshire (6.21), and Dyfed-Powys (5.99). The lowest recorded dog theft crime rates per 100,000 people were in Cheshire (0.37) and the City of London (0). West Yorkshire had the greatest number of DTCs in 2017 as well as previously in 2016, and also had the highest DTC rate per 100,000 people from 2015 to 2017. Aside from City of London, which had an estimated population of 7700 and no recorded DTCs, Cheshire had both the lowest DTC in 2017 and the lowest DTCs per 100,000 people. Staffordshire's DTC rate per 100,000 increased from 2016 (6.07) to 6.21 as it recorded two more DTCs in 2017 (70). The Metropolitan police still has a relatively low DTC per 100,000 (1.91) compared to its second highest number of DTCs of 169. Overall, there were 1842 dog thefts nationally in 2017, an annual increase of 11.43%. Of these, the outcome of 1013 was no further action (61.28%) and there were 39 charges (2.11%). Whilst there was an increase in dog theft crimes annually, there was also a decrease in the number of charges annually.
Our findings, in Table 5  Our research also revealed wider discrepancies with the data provided from FOI requests by Emporium and Direct Line. Emporium, for example, stated 152 dogs had been stolen from Lincolnshire in 2017, compared to the 27 stolen dogs stated by Direct Line. Our own FOI data, which were the last to be requested, revealed only 24 dog theft crimes (Table 6). While we acknowledge these are "live" systems and "recording these figures are not generic, nor are the procedures used locally in capturing We found both strengths and weaknesses within the data we used. Firstly, although our dog theft crime database provides the most robust dog theft crime data available for England and Wales in 2015, 2016 and 2017, the dataset remains incomplete. There were police force inconsistencies in recording dog theft crime, which meant some data was unusable or could not be accessed or analysed. This has implications on the accuracy of any spatial analysis. Secondly, FOI requests are the only way to access this data. This led to some of the data from some police forces not being provided due to the cost and time requirements of searching and providing the data. As a result, three of 44 police forces could not supply useable data.
Other issues arose due to dog theft not being classified as a crime in its own right. This showed through a lack of universal recording and the data being very vague or coarse. The quantitative analysis of police force data we have presented only provides a superficial understanding of dog theft crime in the areas included-the figures do not reveal why dog theft crime is increasing or decreasing. Furthermore, our approach has not taken into consideration the experiences of those involved in dog theft. Pet ownership is a highly emotional, affective, and caring practice, and pets are important in shaping the lives of humans [3,6,13]; stakeholder perspectives would provide more detailed insights.
Consequently, we have two key recommendations. First, there is a need for a qualitative study to understand dog theft crime in different parts of the country. While our approach was important, it did not tell us why dog theft crime and related charges increased or decreased annually from 2015 to 2017. A qualitative study with victims of dog theft, theft support organizations, police officers, and convicted dog thieves would help understand why dog theft crime increased/decreased in certain areas. Furthermore, as we have shown, some police forces have a higher percentage rate of people charged for dog theft crimes, and others have seen a decrease in dog theft crimes. A qualitative study is required to understand stakeholder perspectives of dog theft crime in specific areas. This would take into account police force strategies, external factors such as the media engagement and organizational collaboration, and the experiences of victims of dog theft crime. Alongside this, the databases of stolen and missing pet organisations could also be analysed to gain a better quantitative understanding of the spatialities of dog theft.
Second, there is a need for a standardised approach to recording dog theft to help provide greater transparency with respect to dog theft crimes in England and Wales. This could be achieved by classifying dog theft (or pet theft more generally) as a crime in itself under the Sentencing Guidelines associated with the Theft Act 1968. Classifying dog theft in this way would also help compare and contrast the spatiality of dog theft crime, providing details from police force jurisdictions with greater problems with dog theft crime. The classification of dog theft as a specific crime would also reflect the greater public discourse around dog ownership and help situate them as emotional sentient beings rather than disposable inanimate objects.