Can We Control the Border?
A Look at Recent Efforts in San Diego, El Paso and Nogales

BACKGROUND --
THE "OUT-OF-CONTROL"BORDER

Illegal immigration through surreptitious entry into the country (Entered Without Inspection, or EWI in INS parlance) has not always been as great a problem as in recent years. In 1965, the number of Border Patrol apprehensions of illegal border crossers — mostly those who entered over the U.S.-Mexican border — was 110,000. By 1993, the level of apprehensions had grown to 1.35 million, over a twelve-fold increase (see Figure 1).

The appearance that the border was sieve-like contributed to the adoption — after several years of legislative effort — of the Immigration Reform and Control Act of 1986 (IRCA). This act established sanctions against employers who knowingly hire illegal aliens. A sharp drop in illegal entries (as indicated by a drop in apprehensions) resulted from concern among would-be illegal immigrants that they would be unable to find employment in the United States. This decline in attempted illegal entry is a clear indication of the potential long-term deterrent value of the employer sanctions system, if it operates effectively.

But, by 1989, the apprehension rate resumed its climb. This renewed increase of attempted illegal entry despite enactment of employer sanctions points to their ineffectiveness, as they have been implemented, as a deterrent. The reasons that the IRCA provisions have not worked are various, including a convoluted system of documents that may be used to establish entitlement to legal employment, the ready availability of fraudulent documents, deliberate disregard of the law by some employers who see little threat from limited enforcement efforts, and new arrangements whereby unscrupulous contractors furnish illegal workers to employers.

What Conclusions May Be Drawn from the Apprehension Data? — Because of the unavailability of any alternative timely data on the volume of illegal entry, apprehension data generally are used as an indicator of the relative magnitude of the problem. But INS and Border Patrol authorities readily acknowledge that the number of apprehensions does not equal the number of attempted entries, because the border control effort is less than totally effective.

Some Immigration and Naturalization Service (INS) officials have suggested in the past that two or three aliens slipped past the Border Patrol for every person caught. The Border Patrol Sector Chief in El Paso suggested an even lower rate of apprehension in that sector (one of eight illegal crossers) before the initiation of Operation Hold the Line. The percentage of apprehensions will vary over time and at different locations along the border, depending on Border Patrol interdiction capabilities and the numbers and resources of would-be illegal entrants.

To underscore the difficulty in interpreting apprehension data, we offer an example. To begin, the ratio of apprehensions to EWI border crossings might be 1:2 with 1,000 aliens apprehended and 1,000 avoiding apprehension. At a later time, because of increased Border Patrol resources, the ratio might become 9:1 with 1,800 aliens apprehended and only 200 avoiding apprehension. The latter situation might cause attempted illegal entry to drop in half, so that apprehensions dropped back to 900 and only 100 succeed in avoiding detention. The apprehension data then would reflect only a modest decline of 10 percent from the starting point of 1,000, but the level of illegal entry would be down by 90 percent. Alternatively, Border Patrol efforts could be undermined by diversion of resources to a more pressing responsibility, would-be illegal alien entry attempts might then double, and the ratio of apprehensions to attempted entries could drop to 1:5 with only 800 apprehensions for each 4,000 attempted illegal entries. The apprehension data would reflect a drop by one-fifth, but the illegal entry problem would be 220 percent worse.

In that case, what meaning can be derived from the apprehension data? At best, they are a rough indicator over time of the effectiveness of the Border Patrol. But, it must be remembered that a change in the data may reflect a change in the effectiveness of the Border patrol or a change in the influx of the illegal crossers or, more likely, a combination of factors.

There is one other statistic collected by the Border Patrol that serves as a check on the meaning of the apprehension data — hours of agent time spent on border (line) and non-border (non-line or interior) operations. By looking at the agent-hours (which we convert to agent-days by dividing by eight), it is possible to evaluate whether increased or decreased personnel resources explain some of a change in apprehension level. This becomes important in situations such as the closing of stations in the San Diego sector during Operation Gatekeeper.

If the fact of the closing were not known, and sector-wide or total apprehension data were studied, it would appear that a decrease in apprehensions might be equated with successful deterrence. When the agent-days of operations are looked at, it is clear that such a decrease relates to a reduced level of enforcement activity.

An informal check on apprehension data is in use in San Diego. Border Patrol agents report at the end of their shift on the number of illegal aliens they have detected. This data is refined by comparing it with the observation of others in the area to avoid double-counting and then compared with actual apprehensions. This data may be subject to manipulation, but it offers Border Patrol supervisors an unofficial indicator of the level of success in attempting to apprehend all illegal border crossers.

Despite the limitations in using apprehension data, some inferences may be drawn from them. Most of the EWI problem occurs at the U.S.-Mexican border, and it results primarily from the attempted illegal entry of Mexicans (see Figure 2). Over half of the challenge posed by illegal entry, to the extent that it may be judged by the level of apprehensions, has occurred in two sectors — San Diego and El Paso.

Hispanic advocacy groups have asserted that the disproportionate share of Mexicans in the apprehension data is because the Border Patrol has concentrated its efforts along the U.S.-Mexican border, and that reflects an anti-Hispanic prejudice. That this assertion is groundless is apparent in the results of the IRCA amnesty for illegal aliens. Three-fourths of the over three million applicants — those who had taken advantage of the out-of-control border to enter the United States to find jobs — were Mexican. More than an additional nine percent were from Central American countries. Most of these illegal aliens were in the EWI category. Thus about seven of every eight of the three million illegal alien amnesty applicants had taken advantage of inadequate security at the U.S.-Mexican border.

Recidivism — Another factor that must be understood about apprehension data relates to recidivism. What happens when the Border Patrol apprehends an EWI? If the alien is Mexican (over 95% of the cases), he or she is usually given the opportunity to waive a deportation hearing and accept "voluntary departure." At the border, this means release for return to Mexico or Canada. But other apprehended illegal aliens are also offered the chance to leave on their own — very few are actually de- ported. In some cases, EWI Mexicans are also deported, such as upon release from prison or even along the border when they have forcefully resisted apprehension or in cases of flagrant recidivism. However, since 1990, among apprehended EWI aliens, the largest share deported in any year was only slightly over three percent. The other 97 percent were quickly processed, with minimal data recorded regarding name, date and place of birth, and then "required to depart." These voluntary departure cases represent the vast bulk of the workload of the U.S.-Mexican border enforcement operations.

The Border Patrol has long known that many of the Mexicans whom it apprehends and requires to depart continue to repeat illegal entry attempts in the hope that they will succeed the next time. But, until recently, the Border Patrol had little ability to know how many of those apprehended were repeaters, unless the alien admitted that fact. This means that a million apprehensions in a year is not the same as a million different persons being deterred from illegal entry. Rather, the same person might be apprehended two or three or more times in a single day or week.

The recidivism rate indicates one of the difficulties in relying on apprehension data as an indication of the level of attempted illegal entry. If the Border Patrol is less efficient, it will be easier for aliens to slip into the country, and apprehension data will be lower. If, however, the Border Patrol is more efficient, a determined would-be illegal entrant might be apprehended a half dozen or more times before evading apprehension or giving up the effort. On the other hand, if the Border Patrol ordinarily caught only half of all EWI entrants, and then improved its operation so that it apprehended nearly 100 percent of the attempted illegal entrants, the result as seen in apprehension data would likely be a significant surge in apprehensions at first, and then, as word spread among would-be illegal crossers that the odds had become strongly against successful entry, the number of attempted illegal crossings would then drop off sharply, as would apprehensions.

Controlling for External Influences — One challenge in analyzing illegal alien apprehension data is to adjust for the considerable changes that take place over time. For example, a close look at the recent apprehension data reveals the upward trend since 1989, following the drop due to the adoption of the employer sanctions system in the 1986 IRCA.

Between October 1991 and September 1993, when Operation Hold the Line was launched, total monthly average apprehensions increased more than five percent each year. Then in FY-94, after the El Paso operation was in place, apprehensions dropped (see Table 1). A comparison of just the FY-94 and FY-93 numbers would indicate a drop from an average of about 100,000 per month to under 82,000 (or about 19%). However, the earlier upward trend suggests that the level in FY 94 likely could have been about 120,000, and, therefore, the overall significance of the change caused by better border control was a reduction of about 40,000 (or nearly 40%) below the prevailing trend.

Another clear pattern that may be seen in the data is a seasonal fluctuation. At the end of each year, the apprehensions decline, with December the lowest level of the year (see Figure 3). They increase early in the new year — as the new planting season approaches. There tends also to be an increase near the start of the harvest season. The high apprehension months reflect the attraction of employment opportunities in agricultural labor, and the low apprehension months suggest that the pull of "home for the holidays" is stronger than the lure of U.S. economic opportunities.

Finally, there are also patterns that may be found in the apprehension data that correlate with the days of the week. This is noticeable in El Paso, especially before Operation Hold the Line, when large numbers of Ciudad Juárez residents were illegally entering El Paso for jobs. The weekend apprehension rates would decline, and on Monday they would rise. Careful analysis of border operations requires controlling for this type of variable.

What Are Other Measures of Successful Border Control? — A December 1994 report on border control operations by the General Accounting Office (GAO)* cited Border Patrol officials as identifying four measures for gauging their success: (1) a reduction in apprehensions; (2) an increase in fraudulent attempts to enter the country through ports of entry; (3) an increased level of apprehensions elsewhere on the border; and (4) fewer illegal immigrants detected in the interior of the United States.

Of course, the fourth indicator is the most important one. However, unless some new way of measuring the illegal alien population in the United States is devised, other than the census, we would have to wait until after the results of the 2000 census are available to see if current efforts are succeeding. Even if we were prepared to wait for so long, could we trust the resulting data? Census officials acknowledge a possible undercount of as much as one-third of the illegal alien population.

The current estimate of the Center for Immigration Studies and the INS of the net amount of illegal immigration into the United States each year is about 300,000, about half of whom are border crossers and the other half are legal non-immigrant entrants who violate their status. That estimate is derived from data collected from among the over three million applicants for the IRCA amnesty and from comparing the decennial censuses — that count all residents, including illegal aliens — with the number of immigrants who entered legally during the same period. Also estimated are the number of illegal immigrants who die or return home on their own. These methods are imprecise because the survey data depends on the information supplied by the illegal immigrants, and the census data are flawed by the problem of not counting persons who are anxious to escape official notice because of their illegal status. But most analysts think that the net level of illegal newcomers is in multiples of hundred thousands rather than the millions suggested by estimates based on the apprehension data.

The estimated level of over 300,000 net new illegal immigrants each year indicates that these uninvited newcomers constitute more than one quarter of all annual immigration. The administration and the Congress now agree with the American public that a major effort is required to better management the border and ports of entry and make effective secondary controls at the work site.

Data on aliens deported and "required to depart" also are of limited value in gauging the effectiveness of the border control effort (see Figure 4). If the INS and the Border Patrol were able to effectively identify and remove most illegal aliens who manage to avoid detection at the border or who overstay their terms of legal entry, then the data generated by interior control operations would serve as an indicator of how well the first line of defense is doing. However, compared to the estimates noted above of over 300,000 new illegal immigrants each year, the INS is finding and removing fewer than 50,000.

Like the overall decline in apprehensions in 1986 after the passage of IRCA, the number of deportation cases declined. But, unlike the resumed rise in apprehensions, a lower level of overall deportations has continued, along with increased removal of persons convicted of narcotics and other crimes. Thus, the removal of EWI illegal residents has not been keeping pace, and the data reflect a level of effort that has little relationship to the enforcement challenge. Therefore, little can be inferred from them about border control effectiveness.

INS acknowledges that it has inadequate resources to discharge its law enforcement responsibilities. For example, a March 21, 1995 San Jose Mercury News article quoted William Bonnette, the Livermore, California Border Patrol sector chief explaining that, "There's no way I can handle 52 counties with 52 agents." The report noted that the INS district investigations branch in San Francisco had 150 employees in 1988 and 95 today (65 of whom are agents) to cover 49 counties. It cited a General Accounting Office study that reported that the INS admitted to being unable to pursue 36,000 leads on possible illegal aliens. Even with an increased emphasis on identification and removal of criminal aliens, INS authorities, including Commissioner Meissner, acknowledge that they are not yet able to screen, identify and process for removal all deportable criminal aliens. Chief Bonnette noted that doing the job properly would mean interviewing all foreign-born criminal suspects after booking, not just after conviction.

We may look for some of the answers to questions about what conclusions may be drawn from the apprehension data by turning our focus to the U.S.-Mexican border. In El Paso, apprehensions fell dramatically in 1993. Did that mean the Border Patrol there was less effective in apprehending illegal crossers? Or did it indicate success, because the Border Patrol is deterring would-be illegal immigrants from even trying to enter? The same questions are now relevant for San Diego. Apprehensions were declining even before Operation Gatekeeper began. Does that indicate greater deterrence, as claimed by the Border Patrol? Or, as others suggest, does it indicate that the economic pull factor was weaker because of the depressed California economy?