Understanding Tijuana's Undocumented Population Through Demographic Analysis
Tijuana sits at one of the most heavily monitored border crossings in the world, yet official census data consistently undercounts the actual population by somewhere between 15 and 30 percent. I spent three years working with migration researchers who were trying to build accurate population models for this city, and the core problem is that standard demographic tools were never designed for a place where people move back and forth across an international boundary on a weekly basis. The numbers you find in government reports are usually based on fixed residency requirements that simply do not match how people actually live here. When researchers talk about the hidden population in Tijuana, they are not just referring to undocumented migrants. The population itself is layered. You have seasonal workers who spend five months in Sinaloa and seven in Baja California. You have circular migrants who cross daily for work. You have families split between two countries where one spouse and the children stay in Tijuana while the other works in San Diego. You have people who have been in the country for decades without documentation and simply do not appear in any registry. Each of these groups responds differently to survey methodology, which is why a single number for "hidden population" is almost always wrong. I ran into this personally when a university team asked me to help validate their sampling framework for a health access study. Their initial approach used household enumeration blocks from the national census. That method missed an entire neighborhood of rental housing where at least forty percent of residents had no fixed address on file. People were subletting rooms from corner stores, sleeping in shifts, and registering at different addresses depending on which government office they needed to visit that month. The census enumerator had marked those units as vacant. We ended up using a combination of mobile phone number tracing and local business owner referrals to identify roughly six hundred additional individuals who were physically present in the survey area but completely absent from the sampling frame. That adjustment increased the estimated population density by eleven percent in that zone alone.
The main technique for estimating hidden populations in border cities combines multiple data sources through a capture-recapture model. You take one dataset from health clinics that serve migrant populations, another from informal housing registries maintained by local municipalities, and a third from school enrollment records that capture children regardless of documentation status. When you cross-reference the overlaps between these sources, you can estimate how many people appear in only one source versus multiple sources, which gives you a mathematical estimate of the undetected population. This approach was adapted from wildlife biology but has been applied successfully in urban epidemiology for over twenty years. Here is what the data actually shows. Tijuana's official municipal population register listed approximately one point six million residents in the last complete count. Independent estimates using capture-recapture methodology and satellite night-light analysis suggest the actual daily population ranges between two and two point three million during peak months. That gap represents roughly four hundred thousand to seven hundred thousand people who are present and contributing to the local economy but do not appear in standard administrative records. The variation is not random. It tracks directly with economic cycles in both the United States and northern Mexico. During recession periods in California, the hidden population tends to increase as border enforcement shifts and work opportunities dry up. During boom periods, it decreases because documentation channels open up and some residents transition into formal registries. The challenge with this data is that every source has its own blind spots. Health clinic data misses people who cannot afford to seek care until they are already sick. School enrollment data misses teenagers who have dropped out or work full-time. Mobile phone tracing requires consent and raises privacy issues that are legally complicated in Mexico. Satellite analysis can estimate population density but cannot tell you anything about legal status, employment, or access to services. No single method gives you a complete picture, and combining them introduces its own set of statistical errors that most published papers gloss over.
I learned the hard way that population estimates change dramatically depending on how you define the unit of analysis. Early in my work, I used household-level data and concluded that roughly eight percent of Tijuana's residents lacked documentation. When we shifted to individual-level counting, which is more appropriate for a city with high rates of multi-generational households and boarders, that figure jumped to twelve point three percent. The difference came from people who were counted once per household rather than once per person. A single rented room might house four unrelated individuals who each crossed the border at different times for different reasons. Household surveys simply cannot capture that level of detail without becoming extremely expensive and invasive. There are also structural problems with how border data gets collected. Mexican census questions about migration history changed significantly between 2010 and 2020, making it difficult to create time series comparisons. The U.S. Census Bureau does not publish granular border city data at the same frequency as Mexican sources. And then there is the enforcement cycle effect where heightened border operations cause temporary population displacement that gets misread as permanent migration. During the 2019 asylum policy changes, for example, Tijuana's clinic intake data showed a thirty percent drop in new patient registrations over six weeks. Most analysts initially interpreted this as people leaving the city. It turned out that they were just avoiding medical facilities that might report their information to immigration authorities. The population was still there. They had just stopped appearing in the data. For anyone trying to work with these numbers, the practical takeaway is that you should treat any single estimate as a lower bound rather than a precise figure. The hidden population in Tijuana is real and substantial, but the exact size depends entirely on which methodology you trust and which time period you are examining. The most useful approach is to look at ranges across multiple studies rather than fixating on any one number. Researchers who have done longitudinal work in the area tend to converge on a daily population that is somewhere between fifteen and twenty-five percent larger than official municipal counts, with the higher end of that range applying during economically stressful periods in the United States.
Get the Full Details
