It is impossible to ignore the recent growth of the U.S. undocumented migrant population, which has almost tripled over the space of the past fifteen years.1 According to recent estimates, the annual flow of undocumented migrants into the U.S. is on par with the number of migrants who are granted legal permanent residence.2
Given these figures, it is not surprising that concerns about controlling unauthorized migration have been at the forefront of recent debates over immigration reform. The inability of the U.S. Congress to agree on an immigration reform bill for the past two years—despite mounting public pressure—provides some indication of the complexity of this issue.3 One reason for the inaction is the political impasse created by the conflicting priorities of immigrant rights advocates, immigration control advocates, and still other factions that see immigration as a key ingredient for long term, economic growth.4
It is important to note, however, that the causes of undocumented migration today are not exactly the same as the factors that were driving these migration flows in the earlier part of the 20th century. Even so, the debate around undocumented migration still seems to be lodged in a paradigm that is organized around border control. Although this emphasis is not entirely misplaced, it also tends to produce distorted explanations of undocumented migration and ineffective strategies for controlling undocumented migration.
Advocates for stronger border control are often quick to point out that the U.S.-Mexico border is one the longest undefended borders on the face of the earth.5 It is also true that Mexican nationals have historically composed the majority portion of the U.S. undocumented population. Even so, trends in U.S. undocumented migration aren’t simply a product of the geographic contiguity of Mexico and the U.S. They have also been shaped by strategies for crafting and regulating migrant labor flows and changes in legal-administrative practice which are used to define the very concept of the “illegal alien.”
Immigration Quotas and the Invention of the “Illegal Alien”
The “illegal alien” category was non-existent for the better part of U.S. history. It was introduced to U.S. immigration law during the final years of the first major wave of European migration (dating from the late 1870s to the early 1920s). Prior to this time, immigration laws and other institutional practices were used to restrict and screen migrants on the basis of national-origin and related eugenicist concerns about the health, productivity, and mental acuity of immigrant “racial” stock.6 These criteria were applied, unevenly, across a diverse range of migrant groups, but the brunt of these exclusionary measures fell upon Chinese migrants.
The U.S. first adopted annual immigration quotas for specific national-origin groups (and hemispheric regions) with the 1924 Immigration Act. These immigration quotas required the U.S. immigration system to pay much closer attention to the number of immigrants admitted from each national origin group in a given year. The quotas also provided the impetus for administrative and enforcement practices that could be used to distinguish between legal entrants (who fell within the quota) and unlawful entrants (who exceeded the quota). Prior to the 1924 Immigration Act, the entire matter of unlawful entry was a non-issue for most European migrants due to the U.S. government’s open borders policy for these migrant groups. But even after the imposition of the quotas, legislators introduced criteria for normalizing illegal aliens that tended to favor Europeans.7 In contrast, Mexican migrants were more often framed as the legitimate targets of border control activities and in many Southwestern states, even Mexican Americans were subjected to legal exclusions that were similar to the Jim Crow laws that were used to segregate whites and blacks in other parts of the U.S. Furthermore, many of the national origin quotas were so low that they made unlawful entry the only possible means of entry to U.S. for some national origin groups.8
Operation Wetback and the Bracero Program: Immigration Enforcement as a Strategy for Labor Market Regulation
These are just a few examples of the selective way that immigration quotas were applied to different national origin groups. Racial stereotypes clearly played a role in shaping this selectivity, but it also bears noting that U.S. immigration policies have always been influenced by political and economic factors. For example, the policing of “illegality” in Mexican immigrant populations has been shaped by seasonal migration trends and labor market dynamics that are specific to the southwestern U.S. As some scholars have noted, border control activities along the U.S.-Mexico border have tended to increase and decrease in tandem with the labor needs of employers.9 During periods of high demand for migrant labor border control has tended to be lax, and these activities usually intensify when there appears to be an oversupply of migrant labor.
One of the most dramatic examples of selective enforcement at the U.S.-Mexico border was Operation Wetback, which was used to reduce the size of the Mexican farmworker population in the southwest during the later years of the Bracero program. Under Operation Wetback, over 800,000 Mexican nationals (including a number of naturalized Mexican Americans) were deported to Mexico.10 But it bears emphasizing that Operation Wetback was implemented as an emergency measure which only lasted 2 years (1953-1955), and did not necessarily reflect the typical enforcement practices of the INS of that era. In the years, immediately after Operation Wetback, INS apprehensions of Mexican nationals dropped precipitously.11 And several years earlier, a Senate Judiciary Committee report on immigration included comments from Senator Pat McCarren, which explained why a certain level of illegal migration at the U.S.-Mexico border was tolerable due to the labor needs of U.S. employers.12 This frank admission also reflects the priorities that led the U.S. government to sign the Bracero agreements with the Mexico government in the 1940s.
The Bracero program is generally regarded as the first, formally recognized, guest worker program adopted by the U.S. government. The guest worker category was attractive to many lawmakers because it gave migrants a renewable, temporary legal status (usually lasting no more than three years) that could be used to increase migrant labor flows but without having to increase immigration quotas. The underlying premise for the guest worker category was that most Mexican farmworkers were seasonal migrants who were looking to make quick money, not to become permanent settlers. But the historical record has shown, these assumptions were incorrect. Many Bracero farmworkers started families and established homesteads on the U.S. side of the border. And instead of controlling undocumented migration, the Bracero program actually stimulated an increase in undocumented migration that developed alongside the flow of legally admitted, guest workers.13
Immigration Today: The Expansion of Temporary Workers and Other “Nonimmigrants”
The history of the Bracero program is very relevant to the current immigration debate, since one of the proposals being discussed is a new expansion of the U.S. guest worker program.14 This has lead to a renewed series of debates over the efficacy of guest worker programs as a labor market supply and as a strategy for controlling undocumented migration. Other scholars and analysts have discussed this matter in much greater detail. Even so, one factor that is worth emphasizing—for the purpose of this discussion—is the changing size and scope of the U.S. guest worker population, since this also provides insights into some of the most distinctive factors driving undocumented migration today.
The legacy of the Bracero program and Operation Wetback have produced an immigration policy discourse that associates guest workers with Mexican farm labor in the southwest. As I noted earlier, it has also produced a corresponding emphasis on U.S.-Mexico border control as the main focus of strategies to control unauthorized migration. This emphasis may be warranted, given that Mexican nationals are still the single largest segment of the undocumented migrant population (at 56 percent), according to recent estimates.15 But this statistic also distracts attention from the slow but steady internationalization of the undocumented population, which includes a sizeable number of Asian migrants (over 1.3 million), Central Americans (over 1.3 million) and Caribbean migrants (over 400,000) among others.16 In contrast to the Bracero era, these unauthorized migrant flows are not concentrated in the southwest or in the agricultural sector. Today, unauthorized migrants can be found in almost every employment sector, including health services, child care, manufacturing, food processing, retail, entertainment, hospitality services, and in some cases, working businesses that contract with the federal government.17
Like the Bracero program, there is also a connection between the growth of the temporary worker population and the growth of the unauthorized migrant population. It has been estimated, for example, that between 25-40 percent of the current undocumented population is composed of people who entered the U.S. legally, with temporary visas, but did not leave when their legal status expired.18 These trends also have to be viewed in light of the massive increase in the flow of noncitizens entering and exiting the U.S. each year.
It is often observed that although U.S. migration flows are currently at historic highs, they are actually no larger than the migration flows of the last major peak of European migration (1870s-1920s). This is true when comparing the absolute number of annual admittances from the 1870s-1920s to the annual number of people granted legal permanent residence from the 1990s to the present.19
But what these figures leave out is the massive increase in temporary visitors (or “nonimmigrants”) who are currently entering and exiting the U.S. every year. Although immigration flows fluctuated widely over the course of the 20th century, the annual flow of nonimmigrants grew steadily from decade to decade. In the early 20th century the flow of nonimmigrants was typically much smaller than the flow of permanent settlers. In the 1920s, the annual flow of nonimmigrants ranged between 150,000-200,000 per year.20 By the late 1950s, this flow had increased to over 1 million persons.21 By the late 1990s, this number increased to over 30 million, where it remains today.22 However, this number only accounts for persons admitted with I-94 forms (who are usually entering the U.S. from other continents). If temporary visitors admitted without I-94 forms are included (who are predominantly from Canada and Mexico) this number increases to over 175 million persons as of 2006.23
Tabulations on the flow of the nonimmigrant population can be complicated because the U.S. government has changed its definition of “nonimmigrant” and changed the way it has gone about recording these arrivals many times over the past 150 years.24 It is also likely that the tabulation of nonimmigrants has increased in recent years because the government has improved its ability to document temporary arrivals. But it is also apparent that there has been a very real and dramatic growth in the absolute number of nonimmigrants entering the U.S. on annual basis. In the early 1900s, the flow of nonimmigrants was often no more than 10-20 percent the size of the flow of legal immigrants into the U.S.25 By the end of the 20th century the flow of legal immigrants was dwarfed by the nonimmigrant flow—being barely more than 3 percent if counting nonimmigrants admitted with I-94 forms, and a small fraction of 1 percent if counting the total nonimmigrant flow.26
Temporary workers have always been a small fraction of the total flow of nonimmigrants. In the early 1920s, there was no specific category for temporary workers27, but this annual flow exceeded 1 million by the late 1990s; being on par with the size of the annual immigration flow (persons granted legal permanent residence).28 It also bears noting that temporary workers are just a fraction of the broader category of longer-term “visitors” who are admitted to the U.S. on visas that span several years. If one only counts students and workers admitted with temporary visas in 2006 (including spouses and children), this category of persons was over twice the size of the number of noncitizens who were granted immigrant status in the same year.29
The popular conception of immigration still has not caught up with the realities of these present day dynamics. Even if we understand that people do not literally file through Ellis island quite the same way as they did in the early 1900s, there still is a tacit assumption that immigrants are new arrivals. On the other hand, undocumented migrants are assumed to be people who “jumped ahead” of the line. However, the vast majority of people counted among the annual flow of new immigrants are not new arrivals. In 2006, for example, two thirds of the annual immigration flow (over 800,000 persons) were people already residing in the U.S. who had their legal status adjusted to that of a legal permanent resident. In contrast, only 447,000 of the persons who were classified as “immigrants” in 2006 (i.e. legal permanent residents with the option to apply for citizenship) were actually new arrivals.30
On the other hand, many of the people who join the ranks of the undocumented are also individuals who had been admitted to the U.S. with a temporary visa or some other provisional legal status, but had their application for permanent residence declined or delayed beyond the time frame of their temporary visas. These people have not “jumped ahead” of the line, they have followed a process used by hundreds of thousands of others who were unable to enter the U.S with permanent legal status. And their unfortunate situation is the product of an immigration system that is accommodating an annual flow of noncitizens that numbers in the tens of millions, but grants a very small fraction of these people permanent legal status in any given year. This is one reason why the annual statistics on immigration are better understood as a legal-administrative construct rather than a literal account of new arrivals; representing the small sliver of the broader noncitizen population who have been redefined as “immigrants” by the immigration system.
It is important to emphasize, however, that a large majority of the nonimmigrants do not stay in the U.S. any longer than the terms of their admission and have no intention of becoming permanent residents. This becomes most apparent when comparing the total annual flow of nonimmigrants (which has ranged between 247-175 million; 2000-2006)31 to estimates of the annual increase of the undocumented migrant population (which average 850,000 per annum; 2000-2005).32 The vast majority of nonimmigrants leave within days or weeks after their arrival—including the hundreds of thousands of visitors and day laborers that move back and forth across the U.S.-Mexico and U.S.-Canada border as well as the millions of tourists who spend their currency in the U.S. each year. The dramatic increase in the size of this population is a direct result of economic globalization, which has expanded transnational markets and liberalized immigration flows (but in ways that tend to favor migrants with resources). It would not be possible to radically reduce this annual flow of tourists, business visitors and seasonal workers without hurting and permanently transforming a large cross section of the U.S. economy. In many respects, this situation takes us back to the labor market pragmatism of Senator McCarran. But instead of acquiescing to the seasonal labor market needs of the southwest agricultural industry, we are faced with the consequences of the growing interpenetration of the U.S. domestic economy and the global economy.
A Multi-Pronged Strategy for Reducing the Undocumented Population
Contrary to the demands of some immigration control advocates, it is not possible to completely insulate the U.S. economy from these global forces or to treat unauthorized migration as an entirely separate matter (i.e. a problem that only concerns the movements of low income migrants across the U.S.-Mexico border). The fact is that the U.S. is accommodating an annual flow of noncitizens that, by the most conservative of estimates, is equivalent to the entire population of Canada. The annual growth of the undocumented population cannot be understood apart from the massive increase in these flows of international students, business visitors, tourists, and temporary workers. Any workable solution to the problem of unauthorized migration will have to be able to balance multiple priorities and will have to wean itself from the idea that reducing the size of the undocumented population necessarily means reducing migration flows. Border control should still play a role in this multi-pronged strategy, but it is not the single most important point of emphasis. Instead, the long term effectiveness of this approach is contingent on improvements made in other areas. I have listed four of these priority areas below.
Tracking and decriminalizing legal status violations
The sheer size of the annual flow of noncitizen requires more effective ways of categorizing these persons and recording their points of entry and exit. In this case, the main goal should be to improve the capacity of the immigration system to document the arrival of all individuals into the U.S. Over the past several years, analysts have developed numerous proposals for creating a database (sometimes referred to as an “electronic border”) which could do a better job of tracking these flows in a noninvasive way, while also making this information accessible to frontline workers.33 This system could make it easier for immigration agents to develop proactive strategies that prevent people from overstaying their visas, as well as contacting individuals just before their legal status is about to expire.
However, the recent get-tough trend in immigration enforcement has created a climate where people with legal status violations (including visa over-stayers) are placed in immigration detention centers, which are often county jails and state prisons that are receiving federal subsidies for providing these services.34 There has also been a tendency to treat these enforcement practices as part of a broader counterterrorist agenda, despite the fact that all of the immigration dragnets conducted since 9/11 have not turned up a single individual who was been deported for conspiring in terrorist activity.35 These trends are not only needlessly expensive (requiring the building of new holding facilities that are contributing to the expansion of the prison-industrial complex) they also send signals that scare noncitizens from interacting with government workers, further contributing to their tendency to go “underground.”
This is why it is important that all matters concerning the adjustment or revocation of legal status (including penalties for violating the terms of one’s legal status) continue to be treated as civil matters, as currently defined under U.S. law. This emphasis should also be reflected in the way these cases are processed and the tone of the legal deliberations on these matters.
Revisiting current immigration quotas
Although most nonimmigrants have no intention of remaining in the U.S. as permanent residents, it is apparent that a small minority of these persons are either entering unlawfully or overstaying their visas. The fact that tens of thousands of these individuals are able to over-stay their visas and find gainful employment (or enter unlawfully and find gainful employment) indicates that there is a gap between labor market demands and current immigration quotas. This problem could be ameliorated by increasing immigration quotas between 10-20 percent36 or making it easier for gainfully employed noncitizens to renew their temporary visas (and for longer periods of time) before they expire. It is important to note, however, that this increase in immigration quotas would not require any increase in new arrivals. It would only require the government to re-categorize a larger share of the temporary visa population as members of the legal permanent resident population.
Eliminating exploitative hiring practices
It is also important to consider that some employers hire undocumented workers precisely because they are undocumented. For example, one reason why the Bracero program produced an increase in undocumented migration was because employers quickly learned that they could pay undocumented workers even lower wages than the Bracero guest workers.37Fear of deportation tends to instill a level of additional compliance that inhibits undocumented workers from demanding higher pay, safer working conditions, less arduous production schedules and so forth. This is why it is important to reduce incentives to exploit undocumented workers. This can be accomplished, in part, by making the administrative process for hiring migrant workers (including guest workers and permanent residents) easier and less expensive for employers. But it will also require more intensive investigations of employment sectors that are known to hire undocumented workers. The goal of these investigations should not just be to “root out” undocumented workers, but to analyze and mitigate the factors that encourage exploitative hiring practices. Besides immigration enforcement, this will require holding employment sectors accountable to existing labor standards and improving these standards were necessary. As a result, the focus of prosecutorial action should not be on the illegality of the worker, but on the illegality of wage levels and employment conditions (regardless of the legal status of the worker).These actions would ensure that migrant workers are being recruited to fill labor shortages or to fill a specific skill-niche that is not present in the available workforce, but not to undercut current wage levels and labor standards.
The U.S. may also be able to learn some lessons from the European Union about how to craft regional immigration policies. Instead of trying to insulate itself from Mexico (by focusing immigration spending on border control) U.S. policy makers and immigration officials could improve their joint planning with Mexican authorities on mitigating the causes of undocumented migration from Mexico. This would require more U.S. involvement and investment in the long term economic growth of the Mexican economy as well as developing closer links between U.S. and Mexican immigration enforcement.
Along with these longer term objectives, U.S. lawmakers should also consider developing legal status options that are better adapted to the complex realities of migration at the U.S.-Mexico (and U.S.-Canada) border. This may require the development of a tracking system and database (as noted earlier) that can keep more comprehensive records of daily cross-border travel. This system could expand renewable visa options for Mexican nationals which are tailored to fit the most typical patterns of daily and seasonal migration (to reduce the temptation to cross illegally). This expansion of legal status options would dramatically reduce the number of people entering the U.S. without documentation at point of entry. This system would also deploy targeted enforcement practices that are much less invasive than the ones currently being used to detect undocumented migrants after they have already entered the nation, which tend to encourage forms of racial-ethnic profiling that violate the civil liberties of large cross sections of the immigrant population.
1 Increasing from approximately 4 million in the early 1990s to approximately 12 million by 2005. J. Passel. 2006. Size and Characteristics of the Unauthorized Migrant Population in the U.S. Estimates Based on the March 2005 Current Population Survey. Pew Hispanic Center. P.3.
2 It has been estimated that the flow of unauthorized migrants into the U.S. averaged 850,000 per year, 2000-2005, which compares to an annual rate of legal immigration (not counting persons admitted with temporary visas) that ranged between 850,000 and 1.1 million for the same period. Department of Homeland Security. 2007. Immigration Yearbook, 2006. Pp.5-6; Passel. P.2.
3 The most recent reform bill that died was Senate Bill 1639 which was rejected by the Congress in summer 2007. The bill received the support of President Bush and Senator Edward Kennedy. Tapper, J. 2007. Immigration Reform Bill Dies in Senate. ABC News Online. June 28, Accessed February 15, 2008, URL http://abcnews.go.com/Politics/story?id=3326113&page=1&CMP=OTC-RSSFeeds0312
4 For a brief review see P. Kretsedemas and D. Brotherton. 2008. Immigration Reform at a Crossroads. Pp 365-374 in D. Brotherton and P. Kretsedemas, eds. Keeping Out the Other: A Critical Introduction to Immigration Enforcement Today. New York: Columbia University Press.
5 The U.S.-Mexico border is actually the 10th on the list of “longest undefended border” and is less than half the size of the U.S.-Canada border which is #1 on the same list. World Factbook 2002. Land Boundaries. Accessed April 8, 2008. URL http://www.umsl.edu/services/govdocs/wofact2002/fields/2096.html
6 For detailed review of these polices and practices see, A. Zolberg. 2007. A Nation by Design: Immigration Policy in the Fashioning of America. Cambridge, MA: Harvard University Press.
7 M. Ngai, Impossible Subjects: Illegal Aliens and the Making of Modern America. Princeton, NJ: Princeton University Press. Pp.64-90.
8 This was especially true for China, which had allowed tens of thousands of Chinese men (and a much smaller amount of women) to migrate to the U.S. between the 19th century and the early 1900s, but was only granted an annual quota of 100 persons. Meanwhile, most European nations were given quotas of at least 1000, and some in the tens of thousands. Ngai. P.28.
9 K. Calavita. 1992. Inside the State: The Bracero Program, Immigration and the INS (After the Law). New York: Routledge; Ngai.
10 Ngai. Pp.155-156.
11 In the years following Operation Wetback, annual apprehensions of Mexican nations dropped by over 90 percent, compared to apprehension levels during the peak of Operation Wetback. Ngai. P.28.
12 Ngai. P.255.
13 Ngai. Pp. 127-166.
14 For a review of some of these proposals (not including the most recent, failed attempt at immigration reform – see endnote 3) see the review provided in Eliot Turner and Marc R. Rosenblum. Solving the Unauthorized Migrant Problem: Proposed Legislation in the US. Migration Policy Institute. September 1 2005. Accessed April 8, 2008. URL http://www.migrationinformation.org. Also see the Bush White House guest worker proposal, The White House, Office of the Press Secretary. President Bush Proposes New Temporary Worker Program. January 7, 2004. Accessed April 8, 2008. 15http://www.whitehouse.gov/news/releases/2004/01/20040107-3.html.
16 Passel. P.6.
17 Passel. P.10. There also are a number of cases where military sub-contractors have employed undocumented workers. Associated Press. 2007. Almost 40 Illegal Immigrants Rounded Up at U.S. Military Bases, January 19; R. Henry. 2007. Federal Raid Leaves Mass. Town a Mess. The Associated Press. March 8.
18 Passel. P.16.
19 In both eras annual admittances peaked at over 1.2 million. The largest peak in the first wave occurring in 1907 (at approximately 1.3 million persons) and the largest peak in the most recent wave being 1991 (at over 1.8 million persons, but this also includes several hundred thousand unauthorized migrants who had their status adjusted as a result of the amnesty provisions of the 1986 Immigration Control and Reform Act). It bears noting, however, that the typical annual admittances of the current era are still much higher (by a difference of 100-400,000 persons per annum) than the annual admittances of the first wave. Immigration Yearbook, 2006. Pp.5-6.
20 R. Barde, S. Carter and R. Sutch, 2006. Nonimmigrants admitted, by class of admission: 1925–1996. Table Ad1014-1022 in S. Carter, S. Gartner, M. Haines, A. Olmstead, R. Sutch and G. Wright, eds., Historical Statistics of the United States, Earliest Times to the Present: Millennial Edition (New York: Cambridge University Press).http://dx.doi.or.
22Immigration Yearbook, 2006. Pp.64-67.
23Immigration Yearbook, 2006. P.63.
24 S. Carter and R. Sutch, 2006. U.S. immigrants and emigrants: 1820–1998. Table Ad1-2 in S. Carter, S. Gartner, M. Haines, A. Olmstead, R. Sutch and G. Wright, eds., Historical Statistics of the United States, Earliest Times to the Present: Millennial Edition (New York: Cambridge University Press). http://dx.doi.org.
25 This estimate is based on an annual flow of nonimmigrants that ranged between 150-200,000 during the first two decades of the 20th century, and an annual immigration flow that ranged anywhere between 300,000 to 1.2 million persons (but was more frequently on the higher end of this range). Carter and Sutch; Barde et.al.
26 This estimate is based on an annual immigration flow of 1 million (typical annual flows ranged between 600.000 and 1.2 million in the late 1990s; 1995-1999) and a nonimmigrant flow exceeding 30 million for persons admitted with I-94 forms, and a total flow of 150-230 million during the same period. Immigration Yearbook 2006; Pp.5-6; 63-67.
27 Barde et. al.
28Immigration Yearbook, 2006. Pp.5-6; 64-65.
29 This being 2.8 million workers and students (including spouses and children) granted temporary visas compared to 1.3 million persons granted legal permanent residence. Immigration Yearbook, 2006. Pp.5-6; 66-67.
30 Immigration Yearbook, 2006. Pp.18-19.
31 Immigration Yearbook, 2006. P.63.
32 Passel. P.2.
33 R. Sagarin, Raphael. 2003. Adapt or Die. Foreign Policy. No. 138: 68-69; J. Steinberg. 2003. Statement to the National Commission on Terrorist Attacks Upon The United States. Fourth public hearing of the National Commission on Terrorist Attacks Upon the United States. October 14. Accessed April 8, 2008. http://www.9-11commission.go.
34 Brotherton and Kretsedemas.
35 A. Ifitikhar. 2008. Presumption of Guilt: September 11 and the American Muslim Community. Pp. 108-137 in D. Brotherton and P. Kretsedemas, eds. Keeping Out the Other: A Critical Introduction to Immigration Enforcement Today. New York: Columbia University Press.
36 If the annual increase in undocumented migrants is approximately 850,000 per annum, and 25-40 percent of unauthorized migrants are visa over-stayers (Passel), this amounts to 220-350,000 persons with temporary visas who are joining the ranks of the undocumented each year. A 10-20 percent increase in immigration quotas (increasing the annual number of adjustments by 200-300,000 persons) could significantly reduce or eliminate the growth of this segment of the unauthorized migrant population. If it was also possible to reduce the inflow of unlawful entrants (which is already occurring with the recent increase in border control activities) and provide options for unauthorized migrants currently living in the U.S. to adjust their status (mainly through temporary visas as proposed by recent immigrant reform bills – see end notes 3 and 14) it would be possible to phase out most or all of the remaining unauthorized migrant population.
37 Ngai. Pp. 127-166.