Evaluation of animal control measures on pet demographics in Santa Clara County, California, 1993–2006

Study population

The same private survey firm used in 1993 was commissioned to conduct a similar random telephone survey of 1,000 households throughout Santa Clara County except Palo Alto (which has its own small shelter and did not participate in 1993). An equal probability of selection method (EPSEM) phone list of residential landline telephone numbers for the survey was purchased from a private company.⁴ Three attempts were made at each number over successive nights. Over 7,000 calls (including disconnected lines, no answers, refusals) were attempted to reach 1000 respondents. People who agreed to be questioned were asked whether or not they owned dogs or cats, fed stray dogs or cats, whether or not the animals had been altered, if they had reproduced, if cats had been declawed, how they obtained their pets, whether or not cats were allowed outside, purebred status, city of residence, and residence type. Data was initially recorded on written interview forms, and manually entered into a Microsoft Excel 2007 (Microsoft Corporation, Redmond, WA) spreadsheet for statistical analysis.

Animal shelter entry information was provided by Santa Clara County Animal Control, Humane Society of Silicon Valley, San Jose Animal Care and Services, and for comparative purposes the Peninsula Humane Society and SPCA in neighboring San Mateo County and Los Angeles County Animal Control. Spay/neuter voucher program information, costs and statistics were obtained from the City of San Jose, and County of Santa Clara. San Mateo County was chosen for comparative purposes, as it most closely resembled Santa Clara County, as opposed to the other four more rural surrounding counties.

Statistical analysis

The 1982–1993 Santa Clara County shelter intake records (from before the launch of the spay/neuter voucher program) were used for projecting the expected numbers of shelter intakes from 1994 to 2005. US Census data was used to determine the number of county households, which was used with survey-derived estimates of the average number of dogs and cats per household and proportion of households that owned dogs and cats to estimate the number of owned dogs and cats in the county, and with survey-derived estimates of the percentage of households feeding stray dogs and the average number of unowned but fed dogs and cats per household to estimate the number of stray dogs and cats in the county.

Data from the survey was initially described using proportions. Pearson’s chi-square test was used to compare proportions; p-values < 0.05 were considered statistically significant. The observed numbers of shelter intakes between 1994 and 2005, during which the spay/neuter program was in place, were compared with projected numbers based on varying the proportion of cats in the voucher program that were owned versus unowned to assess the program’s effect (i.e., change in numbers of shelter intakes). Autoregressive integrated moving average (ARIMA (p, q, d, where p =  order of autoregression, q =  order of moving averages, and d =  order of differencing)) models were fit to the data before the launch of the spay/neuter voucher program (1982 to 1993). Autocorrelation function (ACF) and partial autocorrelation function (PACF) plots were used to select the best-fit ARIMA models and to evaluate the model fit. The selected ARIMA models were then used to estimate and project the trend in number of shelter intakes after the implementation of spay/neuter program (1994 to 2005) with corresponding 95% confidence limits of the ARIMA projections. In addition, we constructed a stochastic model to estimate the number of additional cats that would have been born and taken into the shelters between 1994 and 2005 had the spay/neuter program never been implemented. The key parameters, their corresponding probability distributions for owned and unowned cats, and the data sources are listed in Table 1. This model was also used for benefit–cost analysis of the voucher program. The software program @Risk (version 5.0.0, Palisade Corp., Ithaca, New York) was used for the simulation, using Latin Hypercube sampling and Mersenne Twister generator with a fixed initial seed of 12,345 for 10,000 iterations. Median and the 5th and 95th percentiles were reported.

Dog survey results

Twenty nine percent of responding county households reported that they owned dogs (unchanged from 1993); the average household owned 1.9 dogs, representing an increase from 1.3 dogs in 1993. Using US census data led to an estimate of 332,000 owned dogs in Santa Clara County (assuming Palo Alto has the same ownership frequency). Registered and unregistered purebred dogs were 33% and 18% (total = 51%) of the dog population, respectively; the remainder (49%) of dogs was either mixed or unknown breeds. Dogs were acquired from a variety of sources; the most common were friends or relatives (30%), breeders (25%), public or private animal shelter (15%), with the remainder (less than 10% each) coming from a breed rescue group, a newspaper advertisement, found as stray, being born at home, acquired from a pet store, and rare other sources (Fig. 1).

Seventy five percent (75%) of owners reported surgically sterilizing their dogs. Among those that declined to alter them, 28% of owners said this was a deliberate decision, and none claimed that cost was a justification for not sterilizing. Thirty three percent (33%) of the unaltered dogs were intended for breeding purposes, and 17% were puppies too young for surgery. Of the 99 unaltered dogs, 70 (70.7%) were male and 29 (29.3%) were female.

Twenty one households (2%) in the survey acknowledged feeding dogs they did not own, with an average of 2.3 dogs per feeding household. An examination of zip codes indicated that the majority of these dogs were found in the downtown and north and east sides of the city of San Jose. With some exceptions, these areas are in the lower socio-economic range of households in San Jose. Using US census data, this leads to an estimate of approximately 15,650 transiently or permanently stray dogs throughout the county, or 4.7% of the county’s dog population.

Cat survey results

Twenty five percent of households reported owning cats, representing a decrease from 30% in 1993 (p = 0.013). With an average of 1.7 owned cats per household (a figure unchanged since 1993), the county’s owned cat population was estimated at 256,000 cats. Most cats (85%) were characterized as domestic varieties; only 3% were claimed to be registered pedigree (a figure unchanged since 1993), while others were described as unregistered pedigreed or unknown breed. The percentage of cats kept strictly indoors rose from 33% in 1993 to 49% in 2005 (p < 0.001); only 8% were currently described as strictly outdoors, down from 14% in 1993 (p < 0.001).

The most common source of owned cats was from a friend or relative (42% in 2005 versus 33% in 1993), followed by being found as a free-roaming homeless cat (20% in 2005 versus 32% in 1993), a public or private animal shelter (16% in 2005 versus 12% in 1993), a breed rescue group (9% in 2005 versus 2% in 1993), a breeder (4% in both years), an ad in a newspaper or adopted or purchased in a pet store (2% in 2005 versus 6% in 1993), a negligible percentage born at home (<1% in 2005 versus 6% in 1993), and the remainder coming from various minor or unknown sources. The p-value comparing the source distribution of owned cats between 2005 and 1993 was <0.001.

In 2005 most cat owners (92.8%) had their cats surgically sterilized, compared to 86% in 1993 (p < 0.001). Within the 7.2% of cats not spayed or neutered, 48% had owners that deliberately did not want their cats to be sterilized, 13% had owners who wanted to retain the cat for breeding, 13% were kittens, 13% had owners claiming that the costs were prohibitive, and the remainder gave two or more reasons (the most common of which was lack of time to transport the cat for surgery). Thus, only approximately 6% of owned cats were sexually mature and capable of breeding, approximately half of which were female. However, less than one-half of 1% of owners of sterilized female cats allowed their cats to have a litter prior to sterilization.

The rate of reproduction of owned cats in Santa Clara County was 89 cats per 1000 households, in contrast to the higher rate of 112 kittens per 1000 households in the 1996 National Council on Pet Population Study and Policy (Salman et al., 1998). This may be attributable to the high proportion of altered cats in the county (93%) relative to the comparable 1993 figure and the 2005 national average of 86%. In addition, while in 1993 16% of owned cats had a litter prior to altering, in the current study this figure was less than one-half of 1%. While it was beyond the scope of this study to determine the reasons for this change in attitude, it is likely that greater awareness prompted by considerable multimedia public education about the county-sponsored voucher program instituted in 1993 bore at least some responsibility.

When owners were asked about whether their cats were declawed, 8% stated that they were, but 29% of them obtained the cat in that condition. The most common reason given by owners (84%) for declawing was to protect furniture. Owners not electing to declaw their cats protected their furniture through a variety of means, including having scratching posts and mats, using a spray bottle, clipping the claws, applying double-sided tape, and making loud deterrent noises.

Many individuals fed stray cats: 7% of household respondents admitted to feeding an average of 3.2 cats, a decrease from 10% with an average of 3.4 cats in 1993. Relying on U.S. census data, the estimated fed stray cat population is therefore approximately 135,000 cats, or approximately 35% of the total owned and fed free-roaming/unowned cat population in the county (391,000 cats, which represents a drop from 416,000 in 1993). Only 5.5% of these cats were either trapped or taken to be surgically sterilized by their people feeding them. Fifty six percent of the cats were fed daily, while the remaining cats were fed from once every other day to only occasionally. The cats were most commonly fed on the doorstep of a person’s home (62%), followed by an office (15%), a park (12%), and a shopping center (<1%). Fed stray cats were either alone or belonged to colonies ranging in size from 2 to 25 cats. Two-thirds of the fed stray cats were too wild to be picked up and were defined as feral; the remaining third were classified as unowned (although some of these may have had owners unknown to the survey respondent). Forty seven percent of the female stray cats were known to have had at least one litter, which is probably a conservative estimate. Over half of the known litters were allowed to remain free and disperse into their neighborhoods. Of the remaining kittens, half were kept by the feeder, while the others were given away or taken to an animal shelter. Of the females who had litters, 58% were not trapped or taken to a veterinarian after having a litter, remaining free to potentially breed again.

Population changes at Santa Clara County animal shelters

Changes in dog shelter intakes for Santa Clara County (and the Peninsula Humane Society and SPCA shelter in adjacent San Mateo County for comparison) are shown in Fig. 2. Dog intakes declined 13,643 to 8,441 (38.1%) from 1992–2005 in Santa Clara County. An external explanation for the observed trend is supported by the findings in adjacent San Mateo County, where dog intakes declined by a similar 35.7% between 1990 and 2004. These proportions were not significantly different (p = 0.11).

A substantially different picture emerged when examining changes in cat shelter intakes in Santa Clara and San Mateo Counties (Figs. 3 and 4). Intakes in Santa Clara County dropped 22,473 to 16,369 (27.2%) from 1993 to 2004 and 22,473 to 16,807 (25.2%) from 1993 to 2005, compared to a drop of 8,252 to 6,078 (26.3%) in San Mateo County from 1993 to 2004. Although the two 1993 to 2004 proportions were similar (p = 0.16), there was an overall decline in annual intakes in Santa Clara County of 6,104 cats to 2004 (509 cats per year) for the 12 year period, compared to 2,174 cats in the same 12 year period for San Mateo County (181 cats per year). The absolute changes are economically more germane to counties with respect to shelter expenses because expenditures are based on the per diem cost of maintaining individual cats. The results of the ARIMA (1,0,1) projections indicated a higher-than-expected cat intake to shelters in Santa Clara County during the years when the voucher program was in effect, i.e., 1994–2005 (Fig. 5). The ARIMA (1,0,1) projections further showed that the observed numbers of cats brought in by the field service did not substantially differ from the expected numbers during the same time period (Fig. 6).

Information provided by the HSSV shelter indicated that the majority of cats entering the shelter were unweaned kittens and feral cats. From 2000 to 2004, the HSSV euthanized 53,419 cats deemed unadoptable: 14,406 were too young (under four weeks of age), 7,912 were unsociable, and 7,595 were feral.

Under the voucher program, 20,419 cats were surgically sterilized from 1994–2001 and an additional 6,231 cats were sterilized from 2001–2003. While the program was initiated at the end of 1994, public interest did not start until mid-1995, when a local television station and newspaper ran a story about it.

The San Jose program was initially free to the public; however, various program changes over time were instituted. Veterinarians were reimbursed at a set fee of $25 female and $15 male. Pregnancies could add to the veterinarian reimbursement up to $50, and anatomical issues adjusted the price to as high as $150. In 1996 modifications included requiring a $5 co-pay, and a requirement that cat owners obtain a $5 license and rabies inoculation. While these changes increased the veterinarian reimbursement, they also created a negative effect on the program, as voucher requests declined from 5,600 in the first 16 months of the program to only 2,800 for the year following the changes.

The San Jose voucher program ended in 2003, but the county program continued. Utilizing assumptions in Table 1, if no voucher program had been initiated, the same cats enrolled in the voucher program (assuming that 65% were owned, based on the 2005 survey results) would have produced approximately 312,000 kittens between 1994 and 2005, and approximately 8,600 additional cats would have entered (6,200 surrendered and 2,500 brought in by the field service) the shelters in Santa Clara County. This would have incurred an additional cost of approximately $2.15 million, with the HSSV charge to cities for stray cat services under their contract cost of $250 per cat. If the cost per cat for spay/neuter surgery in 2001–2002 ($23.21 average for all surgeries) can be assumed to be constant from 1994–2005, then the expected cost of the HSSV voucher program was approximately $620,000. Thus, the net gain of the program from reducing the number of cat shelter intakes was approximately $1.53 million. Not counted would be the added burden of approximately 44,000 cats to the stray population in the county.

The proportion of feral cats actually altered in the program considerably fluctuated: from 77% in 2006 to 82% in 2007 to 48% in 2008 (the latter data is from San Jose only). Table 2 contains projections of how county cat and shelter populations would be expected to change in the absence of the voucher program under different owned versus feral cat ratios. Under all plausible scenarios, ranging from 20% to 80% of the altered cats being feral, the costs to the shelters would have likely exceeded $2 million over the 12-year life of the program, and at the higher proportion of feral cats the costs would have likely exceeded $6 million.