Using Craigslist To Measure Pet Overpopulation
Guest blog by Lisa Wahl, Eugene, Oregon
How can animal advocates measure the effectiveness of a spay/neuter program? Is there a way to take the “temperature” of an area’s pet overpopulation problem in a timely way and independent of shelter data, which may be unavailable, inaccurate, or easily distorted by changes in shelter policy?
Specifically, can advocates use Craigslist ads to produce valid and significant results to either 1) measure changes in the pet overpopulation problem in an area over time, or 2) measure differences in the pet overpopulation problem in different geographic areas?
The most likely first effect to be seen from a spay/neuter program is a decrease in “kitten season,” that period of summer and fall when kittens born in the spring are looking for homes.
In 2008 I attempted to count the number of kittens looking for homes on the Eugene, Oregon edition of Craigslist (CL) by manually eliminating duplicate ads. This proved to be very time consuming, but informative, with 39-72 new kittens advertised each week. At the same time, the local county shelter was taking in fewer than 100 cats per month.
Beginning weekly in February 2009 (less frequently later), two counts were taken from CL ads in the Community Pets section for each area being monitored. I recorded results from searches for the words “kittens” (K) and “dog” (D). I added new CL areas to the analysis when this method began to look promising and removed them when results showed limitation in some areas.
The result has become the Craigslist Kitten Index, which is really quite simple:
Craigslist Kitten Index (CLKI) = (K/D) * 10
Why divide kittens by dogs? I hear you ask. I find everything easier to understand with charts. The chart below shows the count of kitten ads appearing in CL. Does that look like kitten season?
To control for differences in the size of the human population, we can look at per capita figures. The chart below shows kitten ads per 1000 people. Does that look like kitten season?
Below is a chart for the CLKI. Now, THAT looks like kitten season! Given the shelter statistics as well as anecdotal evidence, the relative size of each area’s kitten season looks right, too.
We include the count for “dog” searches as a measure of CL’s overall activity level, which is generally increasing, along with increasing use of the internet. Using the dog count as the CLKI’s denominator allows for a quick comparison of different geographic areas rather than any attempt to divide by human population. It evens out the variations of CL activity. Only about half of the ads that come up on the “dog” search are dogs available for adoption (others being lost and found, offers of grooming and pet sitting services, etc.), while 75-94% of ads that come up in the “kittens” search are multiple kittens needing homes.
We can calculate an approximate annual figure by taking an average of the CLKI from mid-May to the first week of October. It’s important to note that CL areas have no definable geographic boundaries, and many factors outside of the defined area can impact the activity level. Nonetheless, the results are very encouraging. Graphing the CLKI shows the expected peak for “kitten season.”
Using shelter data from California available online and CLKI numbers for 15 CL areas, including 17 counties, the CLKI correlates very strongly with the commonly used “shelter deaths per 1000 human population” metric; the correlation is even stronger when looking at cat deaths/1000. It correlates most highly with the shelter death figure for the same year, suggesting that it can be a very timely measurement and available much sooner than shelter data.
However, throughout this process I have also observed some limitations of the CLKI. Not all cities or counties have clearly defined Craigslist areas. Some areas, such as the “Oregon Coast,” are too broad to have an equivalent in census or shelter data. In CL areas where activity is low, the CLKI is more easily affected by outside variables, such as a rescue group that begins to use CL for animal adoptions.
Conversely, in CL areas where the activity is very high and the “dog” search exceeds 1000 ads, the search returns the first 1000 only, rendering the CLKI essentially useless. Recently, CL also changed its search parameters so that if “too few” matches are made, it adds some from adjacent area, usually to total 100, and a manual count has to be done for the “kitten” value.
Still, the CLKI seems to be a valid measurement of an area’s pet overpopulation problem that is accessible to advocates even when shelter data is unavailable. Please share your own experiences with Craigslist or other methods of quickly measuring pet overpopulation by leaving a comment!