DISCLAIMER

This paper reports the results of research and analysis undertaken by Census Bureau Staff. It has undergone a more limited review than official Census Bureau publications. This report is released to inform interested parties of research and to encourage discussion.

On March 1, 2001, the U.S. Census Bureau issued the recommendation of the Executive Steering Committee for A.C.E. Policy (ESCAP) that the Census 2000 Redistricting Data not be adjusted based on the Accuracy and Coverage Evaluation (A.C.E.). By mid-October 2001, the Census Bureau had to recommend whether Census 2000 data should be adjusted for future uses, such as the census long form data products, post-censal population estimates, and demographic survey controls. In order to inform that decision, the ESCAP requested that further research be conducted.

Between March and September 2001, the Demographic Analysis-Population Estimates (DAPE) research project addressed the discrepancy between the demographic analysis data and the A.C.E. adjusted estimates of the population. Specifically, the research examined the historical levels of the components of population change to address the possibility that the 1990 Demographic Analysis understated the national population and assessed whether demographic analysis had not captured the full population growth between 1990 and 2000. Assumptions regarding the components of international migration (specifically, emigration, temporary migration, legal migration, and unauthorized migration) contain the largest uncertainty in the demographic analysis estimates. Therefore, evaluating the components of international migration was a critical activity in the DAPE project.

This report focuses on the evaluation of the U.S. Census Bureau's estimate of legal immigration to the United States between 1980-1990 and 1990-2000. Specifically, the review process validated the estimates of the numbers of people who obtained legal permanent residency in the United States, either as a new arrival or by adjusting their residency status. To produce the estimate of legal immigrants, the Census Bureau reviewed data files from the Immigration and Naturalization Service (INS).

Our evaluation resulted in legal immigration estimates of 5.373 million for 1980 to 1990, and 7.543 million for 1990 to 2000. For both time periods, legal immigrants included more women than men, were likely to be non-Hispanic, and the largest numbers were between the ages of 30 and 49.

Legal immigration is the process by which a non-citizen of the United States is granted legal permanent residence in the United States by the federal government. A non-citizen with legal permanent residence status may remain in the country, be employed, and travel outside the United States without restriction. A legal permanent resident may also seek naturalization. Attaining legal permanent residence in the United States is not the same as becoming a United States citizen.

After dropping to relatively low levels during the period 1931 to 1970 (see Table 1), legal immigration to the United States has increased significantly in recent decades, primarily as a result of a change in immigration law in 1965. According to the U.S. Immigration and Naturalization Service (INS), 7.6 million people attained legal permanent residence in the United States in the fiscal year 1991 to 1998 period--an amount equaling the peak levels of immigration early in the Twentieth Century.

The high levels of legal immigration have once again made migration a significant source of population growth for the United States in the 1990s, accounting for more than one quarter of the country's total population increase. Consequently, accurate Census Bureau estimates of the total U.S. population are highly dependent upon accurate statistics on legal immigration.

The project objective for DAPE Team 9 was to answer the question, "In the current population estimates process, are the assumptions concerning the size and characteristics of legal immigration realistic?" The following sections provide a detailed discussion of the data sources, methodology, assumptions, and limitations of the current approach to estimating the flow of legal immigration. The final section will examine the research hypothesis and provide suggestions for enhancing the quality of the legal immigration component.

Major components of legal immigration

There are numerous ways to qualify for legal permanent residence in the United States, and the 7.6 million immigrants who received legal permanent residence in the fiscal year 1991 to 1998 period represent a wide variety of immigrant situations. INS immigration data are therefore partitioned into a number of different categories, based on the INS class of admission. These admissions vary in terms of application procedure and data source. Differences in data sources and application procedures are important because they impact how well we are able to estimate both the total number of legal immigrants and their demographic characteristics. Consequently, throughout this document we will present our findings separately for four types of legal immigrants: new arrivals, adjustees, refugees, and asylees. These four groups are discussed briefly in the next section.

Two main administrative routes are open to aliens who wish to become legal permanent residents of the United States, depending on their residence at time of application. New arrivals are people living outside the United States at the time of application for legal permanent residence. They must apply through the U.S. consular office in their country and are unable to enter the United States as legal permanent residents until receiving their visa. From fiscal years 1991 to 1998, 3.5 million people received legal permanent residency as new arrivals.^/1

Adjustees are people already living in the United States at the time of application for legal permanent residence. These people are adjusting from a nonimmigrant status--such as foreign student, temporary worker, refugee, or undocumented migrant--to legal immigrant status. Because adjustees are already living in the United States when they apply for legal permanent residence, they do not apply through the consular office in their previous country. Instead they apply directly to the Immigration and Naturalization Service (INS) for permanent resident visas. Roughly 4.1 million adjustees obtained legal permanent residence in the United States between 1991 and 1998.^/2 This adjustee number includes 0.9 million refugees and asylees who adjusted status.

Among those adjusting status, there are two main groups of immigrants. Refugees are people who cannot return to their country of nationality because of persecution or a well-founded fear of persecution. Refugees must apply for refugee status from outside the United States, and must reside in the U.S. for a minimum of one year before applying for legal permanent residency. According to INS figures from the 1998 yearbook, 0.8 million refugees received legal permanent residency between fiscal years 1991 and 1998. The Office of Refugee Resettlement (ORR) within the department of Health and Human Services provides data on refugee arrivals to this country.

Asylees are defined as people who (1)  are unable or unwilling to return to their country of nationality because of persecution or a well-founded fear of persecution, (2)  applied for asylum while living in the United States or upon arriving at a port of entry, and have been granted asylum, or (3) applied for asylum during deportation and were granted asylum by the Executive Office of Immigration Review (EOIR). The applicant's legal status at time of application is not taken into account when INS makes the decision about asylum. The Census Bureau includes asylees in the legal immigration component only when they change status to legal permanent residence. About 106,000 asylees converted to legal permanent residence during fiscal years 1991 through 1998, according to the 1998 INS yearbook.

The asylee process is a complex one involving multiple steps. In this paper, we will use the term 'asylee applicant' to refer to people applying for asylee status. Successful applicants are known as 'asylees.' Asylees applying for legal permanent residence will be referred to as 'pending asylee adjustees,' while those who are granted this status are known as 'asylee adjustees.' For DAPE purposes, it is important to realize that for both the asylee and refugee components there are people in various stages of application for legal permanent residence.

Data sources on legal immigration

The primary source of data on legal immigration is the Justice Department's INS public use immigrant file. Each year the INS provides the Census Bureau with a file containing individual records of all legal immigrants admitted during the preceding fiscal year (October 1 to September 30). Each record, stripped of the individual's name, contains information for the following variables:

• Month of admission
• Year of admission
• Class of admission
• Country of birth
• Sex
• Age
• Zip code of intended residence
• Nonimmigrant year and month of arrival (used for adjustees and asylees only)

It is important to note that the INS microdata file does not contain information concerning the race or Hispanic origin of immigrants; these characteristics must be imputed.

Other sources used include:

• The Office of Refugee Resettlement (ORR) within the Department of Health and Human Services (DHHS). The ORR gathers data on the number of refugees by month of entry and country of citizenship who are eligible to receive benefits from DHHS.
• Provisional data from the INS on the number of pending applications for adjustment to legal permanent residence, by fiscal year.
• Decennial census data on race and Hispanic origin of the foreign-born population, by country of birth.

In section 2 of the statement of findings we detail how each of these data sources is used and in some cases modified in the creation of national population estimates.

This section examines the methodology used by the Census Bureau to estimate the number and characteristics of legal immigrants.^/3 We see from the preceding section that the Census Bureau relies on INS, ORR, and other federal agencies for basic information on the number of legal immigrants. The data sets are created and maintained to serve the needs of their particular agencies, and often do not provide the full demographic detail necessary for creating national population estimates. The Census Bureau modifies these data sets to correct for two main limitations. First, as was previously mentioned, INS and ORR data do not contain information on race and Hispanic origin of immigrants. Because this information is necessary for the Census Bureau to produce national population estimates containing race and Hispanic origin detail, these characteristics must be imputed for immigrants. Secondly, the act of becoming a legal permanent resident (important from an INS perspective) does not always coincide with time of original arrival in the United States (important from a population accounting perspective). This difference between legal and demographic perspectives is an important issue in the context of DAPE, and we will explore it in more detail later on.

The Census Bureau takes the following steps to convert the INS microdata into estimates of legal immigration.

• New arrivals are distinguished from adjustees by using the class of admission variable.
• Refugee records are written to a separate file, because they are covered in a separate file from the Office of Refugee Resettlement (ORR).
• Age and sex are imputed for all adjustees and for new arrivals when the values are missing.
• Race and Hispanic origin are estimated for all immigrants by using race and Hispanic information from the foreign born in the census with the same country of birth.
• The adjustee backlog at INS is accounted for (starting in fiscal year 1995)
• The year of arrival is identified for "one-time" admission laws
• Once Census has created legal immigration estimates from the latest INS annual file, legal immigration is projected forward for an additional 9 months to make it consistent with the current population estimate date of July 1. INS immigration data are available only through the end of the prior fiscal year (September 30).

When all steps are done, we have a matrix of legal immigration by year since the 1990 Census with full demographic detail (single year of age, race, sex, and Hispanic origin).

New Arrivals
We obtain data on new arrivals from the INS file. For new arrivals, it is assumed that the month and year of admission, age, sex, country of birth, and location of intended residence are accurate when the new immigrant enters the United States. Legal permanent residency is granted when they pass through the port of entry.

Adjustees
Adjustee data are also obtained from the INS file. Adjustees, as previously mentioned, are people already living in the United States at the time of their application for legal permanent residency. Consequently, some of the data for adjustees at their time of entry into the United States (such as age and residence) may not match their data when legal permanent residence is obtained.

Estimating the total numbers of expected adjustees in the United States at any given time is difficult because many of these people, while already in the U.S., have not yet applied for legal permanent residency or may never apply. The Census Bureau uses a proxy rule to account for this lack of data. This proxy rule assumes that the number of people adjusting status in a given year is equal to the number of future adjustees entering the country in that same year. The Census Bureau excludes the following categories of adjustees when applying the proxy rule: refugees and some Cuban and Haitian asylees; legal temporary residents, who will not become legal permanent residents; unauthorized migrants, who will not become legal permanent residents; and special "one-time" asylees who arrived prior to 1990. Refugees and one-time asylees are not included in the proxy rule because their numbers may fluctuate greatly from year to year, therefore violating the assumption of year-to-year stability.

The assumption behind the proxy rule is robust, provided that the number of nonimmigrants adjusting to immigrant status is constant from year to year. While this assumption has historically been reasonable, starting in fiscal year 1995 it became increasingly questionable. In that year, a change in immigration law allowed qualified nonimmigrants to apply for legal permanent residency from within the United States. Previously they would have had to leave the country to apply, and their applications would have been processed by the Department of State. Now, however, their applications were being processed by the INS. One consequence of this change in policy was a substantial increase in the mid 1990s in the numbers of applications for adjustment of status to legal permanent residence, as shown by Table 2.

But while the numbers of applications for adjustment was increasing substantially from 1995 through 1997, the number of applications for adjustment actually completed by INS was not. For various reasons, the numbers of applications completed declined for fiscal years 1997, 1998, and 1999. The result was a large increase in the backlog of pending applications for adjustment of status, as shown by Table 3.

These shifts in application processing are problematic in several ways. First, they cloud the immigration data, making it difficult to know whether year to year changes in immigration numbers are due to actual changes in immigration patterns or simply due to changes in processing. Second, the existence of a large backlog has created a shadow, quasi-legal immigrant population that does not fit easily into existing international migration categories. We'll return to this problem later on.

To correct for this backlog, the Census Bureau shifts adjustee cases from later years backwards, so that the number of adjustees in a given fiscal year matches the number of applications in that year (with certain exceptions and allowances). The class of admission variable is used to make sure that none of the shifted adjustees exceeds existing numeric quotas in those immigration categories where quotas exist. After the supply of adjustees from these later years is exhausted, cases from existing files are replicated, under the assumption that future adjustees will have characteristics similar to current successful adjustees. The result of this attempt to account for the adjustee backlog has been an increase in the number of estimated adjustees for all years since 1995, in some cases by a large amount (see Table 4).

Refugees
The Census Bureau produces an estimate of the number of refugees admitted to the U.S. each year by using data from the Office of Refugee Resettlement (ORR) of the Department of Health and Human Services (DHHS). After the ORR data on refugees are compiled, we use INS microdata to impute missing age and sex data for this population. Imputing race and Hispanic origin characteristics for refugees is done similarly, but uses the 1990 census data on foreign-born migrants who entered the United States in the preceding five years.

The refugee component is unique in that refugees often arrive in the U.S. in sudden "waves" as a result of sudden political upheaval in foreign countries. Consequently, there is no assumption of a year-to-year stability in the flow of refugees from a particular country. Because of the instability in refugee flows, the time series of when refugees adjust to legal permanent residence would not be an accurate reflection of their original entry to the United States. Refugees adjusting to legal permanent residence are thus included as refugees at their time of arrival, based on time series of arrival data from ORR, and are not included in the proxy rule.

For DAPE, our goal was to estimate the legal immigration component for the period 1990 to 2000. We limited our refugee component to include only those refugees who arrived in the United States after April 1, 1990 and received legal permanent residence before April 1, 2000. While 1.04 million people entered the United States as refugees during the 1990s, only 0.8 million received legal permanent residence before April 1, 2000. Some of the remaining 0.2 million refugees who have not yet received legal permanent residence will undoubtedly adjust their status in the future.

Asylees
The Census Bureau does not currently include asylees in the population estimates universe until they apply for permanent residence status. Consequently, asylees' legal immigration information does not coincide with their initial entry into the United States. Data for asylees who have adjusted their status are available in the INS microdata file, and are processed in the same manner as the data for the rest of the adjustee population. They are tallied and accounted for separately in the total adjustee population.

Asylees changing status to legal permanent residence are constrained by statutory limits. Through 1990, the limit was 5,000 asylees adjusting status per year. In 1990 the limit was changed to 10,000 (effective in 1991), with 22,700 asylees adjusting status in 1991 under a grandfather clause in the 1990 legislation. For fiscal years 1992 through 1998, asylee adjustees have averaged slightly under 9,500 per year. Asylees adjusting status are also subject to country limitations (as are the balance of the adjustees). This requires that they wait until slots are available for their country of nationality.

There has been a large increase in the number of applications for asylum in recent decades. In the 1970s applications averaged 3,100 per year. In the 1980s, applications increased to an average of 55,000 per year, and in the 1990s further increased to an average of 123,000 per year. A relatively small and declining percentage of the applicants are granted asylum: from 22.5 percent of those applying in the 1970s to 7.8 percent in the 1990s. Some applicants have become legal permanent residents through legislation benefiting specific countries, such as El Salvador, Guatemala, Haiti, Nicaragua, and countries of the former Soviet Union. Applying for legal permanent residency through this legislation removes the applicants from the regular asylum process.^/4

Once applicants are granted asylum, they must wait one year to apply for legal permanent residence, and are subject to the same overall limit, and specific country limits, set for all adjustees. About 71,000 people were granted asylum by INS in fiscal years 1993 through 1997. During that same time, about 44,000 asylees became legal permanent residents. People can also appeal to Executive Office of Immigration Review (EOIR) for asylum. Between 1993 and 1997, an additional 21,000 people received asylee status through EIOR. These figures are not included in any INS statistics on asylees. Thus, while a total of 92,000 people were granted asylum between 1993 and 1997, only 44,000 were granted legal permanent residence, resulting in a backlog of more than 48,000 registered asylees awaiting adjustment to legal permanent residence (Table 5).

One of the limitations of legal immigration data is that they are collected primarily for legal purposes and not explicitly to measure population change. Consequently, most immigration statistics reflect not only demographic processes but legal and political processes, too. As the General Accounting Office (GAO) noted in a report on the quality of immigration statistics, "INS statistical staff told us that annual trends in the number of new LPRs do not convey a meaningful indication of any demographic concept." (GAO/GGD-98-164, p.28.)

The challenge for demographers, then, is to find meaning in immigration statistics that are to some extent distorted by legal processes, policy and political changes, and logjams in the processing of applications. This is not an easy task. Producing accurate estimates of each of the components of legal immigration each year is hampered by limitations of some sort.

Perhaps the bigger limitation of legal immigration data is simply that it is increasingly difficult to easily define what we mean by a "legal immigrant." The Immigration and Naturalization Act (INA) defines a legal immigrant as a non-citizen of the United States who is granted legal permanent residence in the United States. This is a straightforward, easily understood definition. However, its weakness is that it makes no distinction between people who have applied for--but not yet received--legal permanent residence and unauthorized people. The large backlog of adjustee applications is but one group of quasi-legal immigrants. While adjustee applicants may not yet have legal permanent residence (and some percentage of them will no doubt be denied legal permanent residence when their applications are eventually processed), many of them will eventually become legal permanent residents. In the meantime, most of them have the ability to legally seek employment in the United States. By not including these categories of people with the legal immigration component, we are including them in the residual foreign born. Thus, by dividing immigration into the categories of legal immigrant (those with LPR) and unauthorized (all others who entered the country but do not have LPR), we lose a category of authorized people who do not have LPR.

Adjustees
The adjustee component of legal immigration has perhaps the most limitations of any of the four components. It has neither the refugee component's accurate system of administrative records nor the new arrival component's conceptual simplicity. Adjustees are people already resident in the United States at the time of their application for legal permanent residence; therefore data on adjustees at their time of entry into the United States do not necessarily match data when legal permanent residency is obtained. At what point, then, does someone make the transition from unauthorized to authorized? Adjustee data are also complicated by definitional difficulties. At what point should an adjustee first be included in our estimates of the resident United States population: when they apply for legal permanent residence, or when they actually receive legal permanent residence?

The proxy rule we use to estimate the number of future adjustees entering the U.S. in the current year also has its limitations. While the proxy rule may generally hold true when levels of new arrivals and adjustees are constant, it is less applicable when annual levels of immigration are not constant. When the number of new arrivals is increasing substantially each year, the proxy rule will underestimate the number of future adjustees. When the number of new arrivals is decreasing each year, the proxy rule will overestimate the number of future adjustees.

The increasing backlog of adjustee applicants has greatly complicated the ability to make accurate estimates of the size and composition of the adjustee population. Currently we know very little about the backlog population's demographic composition. After the supply of adjustees from these later years is exhausted, cases from existing files are replicated, under the assumption that future adjustees will have characteristics similar to current successful adjustees.

New Arrivals
Unlike the adjustee component, the new arrival component is relatively straightforward and uncomplicated. To examine the completeness of the flow, we compared INS new arrival data with numbers of visas issued by the U.S. State Department's Bureau of Consular Affairs (BCA) for fiscal years 1993 through 1998 (see Table 6). Both the INS and BCA data exclude data on refugees and asylees. The BCA totals are always higher, ranging from 1,167 higher in FY1996 to 35,996 higher in FY 1997. Possible explanations for the discrepancies include situations where the applicants receiving the visa never actually entered the U.S, applicants changed their minds, their circumstances changed, the political situation in their country changed, the home country government created bureaucratic or legal impediments, or the applicants became ill or died.

Other explanations are that the applicant entered the U.S. in the year following that in which the visa was granted, or that the applicant received the visa but INS lost the paperwork, so that the applicant was not in the INS file as a new arrival. INS indicated that paperwork does disappear on occasion, and that the other possible explanations were credible. In any event, the discrepancies for these years are relatively small and appear to be explainable. This suggests that the new arrival data we receive from INS are reasonably complete and contain relatively few limitations.

Refugees
In general, the quality of refugee data is considered to be high, due to the existence of accurate administrative records from the Office of Refugee Resettlement. The major source of error for the refugee component is simply agreeing when to include refugees as legal immigrants. For DAPE, our estimate of refugees for 1990 to 2000 was restricted to the 800,000 refugees who entered the United States during these years and also adjusted status to become a legal permanent resident. ORR data indicate that slightly more than one million refugees entered the country during the 1990s, indicating that roughly 200,000 non-adjusted refugees (one million minus 800,000) are erroneously excluded from DAPE's legal immigration component and are thus included in the residual foreign born.

Another limitation is the process by which demographic characteristics are assigned to refugees. We assume that the race and Hispanic origin of refugees from a particular country matches the race and Hispanic origin of recent immigrants from that country. This assumption is probably the most reasonable way of establishing race and Hispanic origin detail for refugees. However, it might not hold true in those circumstances where refugees are fleeing a particular country specifically because they are ethnic minorities. This is certainly a possibility, because refugee status is conferred on people who claim fear of persecution based in part on race, sex, religion, nationality, or membership in a particular social group.

Asylees
The quality of existing statistics on asylees is hampered by several general limitations. First, there is currently some degree of overlap and doublecounting between asylees and other components of legal immigration, because asylees can have any legal status when they apply for asylum. For instance, most of the people in temporary protected status are also seeking asylum.

Second, because the processes of applying for asylum, applying for legal permanent residence, and receiving legal permanent residence often take a number of years, there is a large quasi-legal population of pending asylees. As with the quasi-legal refugee component, these are not unauthorized people in the strict sense of the word. Asylee applicants, for instance, are now authorized to work after six months. Once registered, they are able to obtain a social security number and apply for other benefits through ORR. Asylee applicants are not currently accounted for in any population estimates methods, which relegates them to the residual foreign born.

We see from Table 7 that as of FY 1998, the overwhelming majority of asylee applicants come from twelve countries, including nine countries in Central and South America, suggesting that a large percentage of asylee applicants are Hispanic. Table 7 indicates also that only about half (32,822 of 69,407) of cases granted asylum between 1990 and 1998 were for these 12 countries, even though they constitute an overwhelming majority of cases filed during the period and cases pending at the end of the period.

Finally, a recent GAO report suggested that asylees are undercounted by a significant degree. INS maintains statistics on asylees, while data on asylee appeals are maintained by EOIR. Some people win asylum on appeal through EOIR, while in other cases "trailing relatives" (some already in the US, others new arrivals) are not counted. In 1998 GAO estimated that these additional categories would have increased the asylee count for 1996 from 18,000 to at least 29,000. By the end of the decade, this asylee undercount had become sizable.

Summary of General Limitations of Current Legal Immigration Data

1. Immigration data sets do not provide full demographic detail for legal immigrants, and thus these characteristics must be imputed. The imputation process itself rests on various assumptions about size and composition of legal immigration, and changes in immigration trends may weaken the reasonability of these assumptions.
2. Changes in immigration laws can impact the volume and processing of immigration applications, which impact the ability to process applications. Processing difficulties can then impact the interpretation of data trends.
3. Asylees are accounted for only when they become legal permanent residents. To the extent that we utilize a residual method to calculate the number of residual foreign born, asylee applicants are being erroneously included as residual foreign born.
4. Legislative changes in immigration law can significantly impact year to year trends in the size and composition of various international migration components. This also makes it more difficult to assume the stability of patterns over time.
5. The lack of an adequate tracking system prevents us from doing analysis on the impact of the delays on the applicant populations. We cannot account for demographic events (births, deaths) for these populations, and records do not contain unique identifications for individuals. There is also no way of telling when people have filed multiple applications under different INS programs for legal status in the U.S.
6. Changes in numbers of adjustees are in part an artifact of problems in processing (a result of the backlog).

DAPE Team 9 estimates that for the period 1980 to 1990 legal immigration to the United States was 5.3 million (see Appendix table A.1). We see from table A.1 that slightly more females than males immigrated in the 1980s and roughly 0.7 million of the 5.3 million immigrants-roughly 13 percent-were Black. People of all ages immigrated to the United States during this time. Roughly 0.9 million were aged 0 to 17, and another 0.9 million were age 50 or above.

Table A.3 expands the race categories and includes a Hispanic component. We see that Hispanic immigration totaled 1.6 million for the 1980s, with slightly more males than females. Immigration among Asians and Pacific Islanders was sizable in the 1980s, accounting for nearly 2.2 million immigrants. Immigration for the Native American and Alaska Native category was quite small, at just under 10,000.

The previous estimate of legal immigration for the period had been 5.8 million. Looking at Table 8, we see that the revised number boosted the number of Black immigrants from 0.6 million to 0.7 million. The number of Nonblack immigrants decreased from 5.2 million to 4.7 million. These estimates exclude 2.0 million immigrants who legalized under the IRCA provisions a program discussed in full detail in the Team 5 Working Paper.

For the period 1990 to 2000, DAPE Team 9 estimates the legal immigration component to be 7.5 million. The female component is slightly larger than the male component. Black immigration was estimated to be just under 0.9 million, with just over 0.7 million of these Black immigrants estimated to be non Hispanic. Asian Pacific Islanders again were a sizable immigration component, accounting for 2.4 million immigrants.

We see from Table 8 that the 1990 to 2000 DAPE estimate is about the same as the previous estimate for ages 30 and above, but is lower for ages 0-17 and 18-29. For ages 0 to 17, for instance, the DAPE estimate was nearly 0.4 million lower than the original estimate.

The Census Bureau estimate of the number of new arrivals entering the country as legal permanent residents appears to be reasonable. State Department data closely match INS data for this component (see Table 6). We do not see any need for improvement of this component.

The Census Bureau estimate of the characteristics of new arrivals also appears to be reasonable. Race and Hispanic origin characteristics are imputed for new arrivals based on their country of origin, and this also appears to be a reasonable method. We see no need for improvement of this component.

The assumed flow of adjustees may be understated due to the existence of a large and growing backlog of pending applications for adjustment of status. This backlog has increased greatly in recent years, and at the end of FY 2000 was nearly nine times larger than at the end of FY 1994. The INS estimates that legal immigration during the period 1995 to 1998 would have been 450,000 to 550,000 higher if the increase in pending applications had not occurred. Since 1998 the backlog has continued to increase: from 800,000 at the end of FY 1998 to slightly more than 1.0 million at the end of FY 2000 (see Table 3). As a result, we suspect that legal immigration for the period 1995 to 2000 could have been underestimated by between 850,000 and 950,000 people. While Census Bureau estimates corrected for some of the adjustment backlog (approximately 700,000), our DAPE work leads us to believe that the Census Bureau underestimated adjustees by 250,000. Based on observed trends in asylee applicant data, we suspect that a high percentage of these missed people in the adjustee backlog are Hispanic.

The assumed characteristics of adjustees may be underestimating the Hispanic component of the adjustee population. We do not know for certain, but evidence suggests that a high percentage of the backlog adjustees missed by the Census Bureau are Hispanic, because one of the main groups impacted by the 1995 change in immigration law was unauthorized migrants. This would suggest a sizable Hispanic population in the adjustee backlog, but we currently know very little about the composition of the backlog population. One possibility would be for the Census Bureau to formally request an INS tabulation of the demographic characteristics of the adjustee backlog population each year. This would give us some knowledge of the people who will be adjusting to legal permanent resident status at some point in the future. Our DAPE work suggests that the assumed characteristics of adjustees are probably underestimating the percent of adjustees who are Hispanic.

Our DAPE estimate of 800,000 refugees only includes those who arrived after 1990 and adjusted their status to legal permanent resident. This figure does not include the roughly 200,000 refugees who arrived after 1990 but have not adjusted to legal permanent resident. In general, the Census Bureau estimates of the refugee component have traditionally been very good, because they utilize the ORR file containing all refugees, not just those who adjusted status.

The imputations of race and Hispanic origin characteristics for refugees are necessary, and the assumptions guiding the imputation process appear to be reasonable. While there may be instances in which the refugees from a country are of a race or ethnic group very atypical of the country's overall race and ethnic composition, our DAPE work leads us to believe that the estimated characteristics of the refugee flow are generally reasonable.

Traditionally only asylee adjustees are included in legal immigration figures; asylee applicants and registered asylees are not currently part of the accounting process. Our DAPE work suggests that the Census Bureau consider including asylee applicants as part of the legal immigration component; otherwise in a residual method they are included in the residual foreign born, despite the fact that asylee applicants are already living in the U.S. and after 180 days are generally allowed to seek employment. The process of receiving asylum and then, eventually, receiving legal permanent residence is often a long one. Because of the current limits placed on the number of people granted asylum (10,000 per year), applicants may be in the U.S. for years before they can be given asylee status. Once they have asylee status, they must wait a minimum of one year before applying for legal permanent residence. While INS is approving roughly 10,000 cases each year and is adjudicating a total of 40,000 to 60,000 cases per year, it had a backlog of roughly 328,000 asylee applications as of September 2000. Each case translates to approximately 1.4 people.^/5 This suggests that if we included (1) asylee applicants and registered asylees in the same manner in which we include adjustee applicants, and (2) asylees granted by EOIR as well as INS, we would add approximately 500,000 people to the legal immigration estimate, which would reduce estimates of the residual foreign born by the same amount.

We know from tables published in the annual INS Yearbooks that the asylee applicant population is demographically different from the other legal immigration components. If we were to include asylee applicants as part of the legal immigration component, their characteristics would alter the characteristics of the overall legal immigration component. Our DAPE work leads us to believe that asylee applicants are more likely to be Hispanic and to be younger than the overall legal immigrant population.

If our definition of legal immigration refers only to those people who have obtained legal permanent residence in the United States, our estimates of the components of legal immigration are reasonable. Thus, the total estimate for legal immigration is also reasonable.

If, however, we want our accounting process to include non-citizens who have expressed the desire for legal permanent residence but are in the application process, then our overall estimate of legal immigration is too low. Our DAPE research suggests that our legal immigration figure of 7.5 million for 1990 to 2000 omits the following estimates:

• a 250,000 underestimate of the adjustee component
• a 200,000 underestimate for the refugee component
• a 500,000 underestimate for the asylee component

The overall potential underestimate, then, is 950,000. If we were to add this to our figure of 7.5 million, our estimate of total legal immigration between 1990 and 2000 would increase to 8.45 million. It should be noted, however, that the 950,000 estimate does not affect the total DA estimate, because if we understate legal immigration we are also overstating the residual foreign born by the same amount.

Some researchers consider people who have begun the application process to be legal immigrants; others consider such people to be illegal immigrants until their paperwork has been processed. The Census Bureau needs to decide how to treat these people, because this pool of people who are neither immigrants nor temporary migrants has increased greatly in recent years. It should be noted that virtually all of these categories of quasi-legal immigrants who currently are not included in our estimates are allowed to work while in the United States.

^/1 Actual number is 3,521,051 Reference is 1998 yearbook, Table 4. Immigrants Admitted by Type and Selected Class of Admission; Fiscal Years 1991-98.
^/2 Actual number is 4,084,017. FYI: This includes 0.9 million special agricultural workers and 0.2 million people residents since 1982, all legalized under IRCA. Reference: 98 yearbook, Table 4.
^/3 See Mulder et al. (2001) for an in-depth description of the methods briefly discussed in the text.
^/4 In 1997 Congress passed and the President signed the Nicaraguan Adjustment and Central American Relief Act (NACARA). Nicaraguans and Cubans, residing in the U.S. since Dec. 1, 1995, could apply for legal permanent residence, and had until March 31, 2000 to do so. They had to meet the requirements of the Immigration and Naturalization Act, except for those exempted by NACARA, which included being a public charge, lack of labor certification, present without admission or parole, lack of valid immigrant visa, and unlawful presence. This resulted in some of the pending asylee applications being closed.
^/5 U.S. Immigration and Naturalization Service. 2000. Monthly Statistical Report, September 2000 FY Year End Report. Asylum Section. Also INS Statistical Yearbook, 1998, Table 29.

Demographic Analysis Population Estimates (DAPE) Research Project Reports Related to Evaluating Components of International Migration (in order of Working Paper Series Number)

Deardorff, K. and L. Blumerman. 2001. Evaluating Components of International Migration: Estimates of the Foreign-Born Population by Migrant Status: 2000. (Population Division Working Paper #58) (December 2001) U.S. Census Bureau.

Perry, M., B. Van der Vate, L. Auman, and K. Morris. 2001. Evaluating Components of International Migration: Legal Migrants. (Population Division Working Paper #59) (December 2001) U.S. Census Bureau.

Cassidy, R. and L. Pearson. 2001. Evaluating Components of International Migration: Temporary (Legal) Migrants. (Population Division Working Paper #60) (December 2001) U.S. Census Bureau.

Costanzo, J., C. Davis, C. Irazi, D. Goodkind, R. Ramirez. 2001. Evaluating Components of International Migration: The Residual Foreign Born. (Population Division Working Paper #61) (December 2001) U.S. Census Bureau.

Mulder, T., B. Guzm n, and A. Brittingham. 2001. Evaluating Components of International Migration: Foreign-Born Emigrants. (Population Division Working Paper #62) (December 2001) U.S. Census Bureau.

Gibbs, J., G. Harper, M. Rubin, H. Shin. 2001. Evaluating Components of International Migration: Native Emigrants. (Population Division Working Paper #63) (December 2001) U.S. Census Bureau.

Christenson, M. 2001. Evaluating Components of International Migration: Migration Between Puerto Rico and the United States. (Population Division Working Paper #64) (December 2001) U.S. Census Bureau.

Cresce, A., R. Ramirez, and G. Spencer. 2001. Evaluating Components of International Migration: Quality of Foreign-Born and Hispanic Population Data. (Population Division Working Paper #65) (December 2001) U.S. Census Bureau.

Malone, N. 2001. Evaluating Components of International Migration: Consistency of 2000 Nativity Data. (Population Division Working Paper #66) (December 2001) U.S. Census Bureau.

Other materials:

Beck, Susan. 1992 "Cast Away," The American Lawyer (October 1992), p. 54-59.

Beyer, Gregg A. 1992. "Affirmative Asylum in the United States," Georgetown Immigration Law Journal (Vol. 6, No. 2, June 1992), p. 253-284.

Beyer, Gregg A. 1992. "Establishing the United States Asylum Officer Corps: A First Report," International Journal of Refugee Law (Vol. 4, No. 4, July 1992), 39 p.

Beyer, Gregg A. 1994. "Reforming Affirmative Asylum Processing in the United States: Challenges and Opportunity," The American University Journal of International Law and Policy (Vol. 9, No. 4, June 1994), p. 43-78.

Conover, Ted. 1993. "The United States of Asylum," New York Times Magazine (September 20, 1993), p. 56.

Cooper, Bo. 1997. "Procedures for Expedited Removal and Asylum Screening under the Illegal Immigration Reform and Immigrant Responsibility Act of 1996," Connecticut Law Review (Vol. 29, No. 4, Summer 1997), p. 1501-1524.

Martin, David A. 1995. Making Asylum Policy: The 1994 Reforms," Washington Law Review (Vol. 70, No. 3, July 1995), p. 725-755.

Mulder, Tammany et al (2001) U.S. Census Bureau Measurement of Net International Migration to the United States: 1990 to 2000. Estimates and Projections Working Paper No 51. Issued April 2001.

U.S. General Accounting Office. GAO has done reports on a variety of immigration issues. Their website is: http://www.gao.gov/

GAO-01-488 Immigration Benefits: Several Factors Impede Timeliness of Application Processing GAO-98-64 Quality of Immigration Statistics

U.S. Immigration and Naturalization Service. Website: www.uscis.gov

There is a great deal of material regarding asylum under /Immigration Services and Benefits/ Asylum and also under /Statistical Reports. Additional material regarding asylum can also be found by doing searches on words relating to asylum.

U.S. Immigration and Naturalization Service, Information Resource Center. INS web address is: http://www.uscis.gov/graphics/services/asylum/ric/

The Center has information about conditions in the countries of origin for asylees and refugees. See also The Human Rights Documentation Exchange for information about the Center's publications not on its web site. Their website is: www.hrde.org

Center for Immigration Studies. Their web site is: http://www.cis.org/

Gordon, Linda W. 1993. Draft Material for the IMMACT90 Report for FY 1992: Information on the Alien Population in the United States. U. S. Immigration and Naturalization Service, Washington, D. C.

Gordon, Linda W. 1993. Immigration Issues, No. 2. Political Asylum Statistics, 1991-92: A Time of Transition. U.S. Immigration and Naturalization Service.

Kramer, Roger G. 1999. Developments in International Migration to the United States: 1999. Immigration Policy and Research Working Paper 34. U.S. Department of Labor, Bureau of International Labor Affairs, 1999.

U. S. Commission on Immigration Reform (now dissolved). Their web site is: http://www.utexas.edu/lbj/uscir/

"The bipartisan U.S. Commission on Immigration Reform was authorized by Section 141 of the Immigration Act of 1990. The Commission expired on December 31, 1997. The mandate of the Commission was to review and evaluate the implementation and impact of U.S. immigration policy and to transmit to the Congress reports of its findings and recommendations." (From Mandate on website)

U.S. Department of State, Bureau of Population, Refugees, and Migration. They have "primary responsibility for formulating policies on population, refugees, and migration, and for administering U.S. refugee assistance and admissions programs."
Web site subsequent to Jan. 20, 2001: http://www.state.gov/g/prm/
Web site prior to Jan. 20, 2001: http://www.state.gov/www/global/prm/

U.S. General Accounting Office. GAO has done reports on a variety of immigration issues. Their website is: http://www.gao.gov/

U.S. General Accounting Office. 1998. Quality of Immigration Statistics (GAO-98-64). Washington, D.C., 1998.

U.S. General Accounting Office. 2001. Immigration Benefits; Several Factors Impede Timeliness of Application Processing; Report to Congressional Requesters (GAO-01-488). Washington, D.C., 2001.

U.S. Immigration and Naturalization Service. Last modified. A Symposium and Celebration: The Fifth Anniversary of the 1995 Asylum Reforms. From INS website at: http://149.101.23.2:80/graphics/services/asylum/symposia.htm Website contains links to the following:

Beyer, Gregg A. Last modified 08/14/2001/ Striking a Balance: The 1995 Asylum Reforms; "A Walk Down Another Street." (Full text version). From the INS website at: http://149.101.23.2:80/graphics/services/asylum/Beyer.pdf [PDF]

Langlois, Joseph. Last Modifed 08/14/2001. The Success of Asylum Reform and the Evolving and Expanding Role of INS Asylum Officers. From INS website at: http://149.101.23.2:80/graphics/services/asylum/langlois.htm

Martin, David A. Not dated. Asylum Reform: A Global Perspective. From the INS website at: http://www.ins.gov/graphics/services/asylum/Martin.pdf [PDF]

Five Years After Asylum Reform: The INS Regains Control; Practitioners Say Reform Still Needed. 2000. Refugee Reports, U.S. Committee for Refugees, February 2000. From U.S. Committee for Refugees website at: http://www.refugees.org/world/articles/reform_rr00_2.htm

U.S. Immigration and Naturalization Service. 2000. Asylum Reform: 5 Years Later. News release issued February 1, 2000. From the INS website at: http://www.ins.gov/graphics/publicaffairs/newsrels/asylum.htm

U.S. Immigration and Naturalization Service. Not dated. Asylum Reform: 5 Years Later. From the INS website at: http://149.101.23.2:80/graphics/services/asylum\asylum_brochure.pdf [PDF]

U.S. Immigration and Naturalization Service. Last modified 08/13/2001. This Month in Immigration History: January 1995. From the INS website at: http://149.101.23.2:80/graphics/aboutins/history/jan95.htm

U.S. Immigration and Naturalization Service. 1999. History of the United States INS Asylum Officer Corps and Sources of Authority for Asylum Adjudication. From the INS website at: http://www.ins.gov/graphics/services/asylum/history.pdf [PDF]

Weissbrodt, David. 1998. Immigration Law and Procedure in a Nutshell. West Group, St. Paul, MN.

U.S. Department of Justice, Immigration and Naturalization Service. Not dated. 1980 Statistical Yearbook of the Immigration and Naturalization Service. Table 11.

U.S. Department of Justice, Immigration and Naturalization Service. Not dated. 1981 Statistical Yearbook of the Immigration and Naturalization Service. Table 11.

U.S. Department of Justice, Immigration and Naturalization Service. Not dated. 1982 Statistical Yearbook of the Immigration and Naturalization Service. Table REF 4.2.

U.S. Department of Justice, Immigration and Naturalization Service. Not dated. 1983 Statistical Yearbook of the Immigration and Naturalization Service. Table REF 3.3.

U.S. Department of Justice, Immigration and Naturalization Service. Not dated. 1984 Statistical Yearbook of the Immigration and Naturalization Service. Table REF 3.3.

U.S. Department of Justice, Immigration and Naturalization Service. 1986. 1985 Statistical Yearbook of the Immigration and Naturalization Service. U.S. Government Printing Office: Washington, D.C. Table REF 3.3.

U.S. Immigration and Naturalization Service. 1988. Statistical Yearbook of the Immigration and Naturalization Service, 1987. U.S. Government Printing Office: Washington, D.C. Table 32.

U.S. Immigration and Naturalization Service. 1989. Statistical Yearbook of the Immigration and Naturalization Service, 1988. U.S. Government Printing Office: Washington, D.C. Table 32.

U.S. Immigration and Naturalization Service. 1990. Statistical Yearbook of the Immigration and Naturalization Service, 1989. U.S. Government Printing Office: Washington, D.C. Table 33.

U.S. Immigration and Naturalization Service. 1991. Statistical Yearbook of the Immigration and Naturalization Service, 1990 (M-367). U.S. Government Printing Office: Washington, D.C. Table 31.

U.S. Immigration and Naturalization Service. 1992. Statistical Yearbook of the Immigration and Naturalization Service, 1991 (M-367). U.S. Government Printing Office: Washington, D.C. Pp. 77-78.

U.S. Immigration and Naturalization Service. 1993. Statistical Yearbook of the Immigration and Naturalization Service, 1992 (M-367). U.S. Government Printing Office: Washington, D.C. Pp. 76-78, Table 30.

U.S. Immigration and Naturalization Service. 1994. Statistical Yearbook of the Immigration and Naturalization Service, 1993 (M-367). U.S. Government Printing Office: Washington, D.C. Table 30.

U.S. Immigration and Naturalization Service. 1996. Statistical Yearbook of the Immigration and Naturalization Service, 1994 (M-367). U.S. Government Printing Office: Washington, D.C. Table 30.

U.S. Immigration and Naturalization Service. 1997. Statistical Yearbook of the Immigration and Naturalization Service, 1995 (M-367). U.S. Government Printing Office: Washington, D.C. Table 29.

U.S. Immigration and Naturalization Service. 1998. Statistical Yearbook of the Immigration and Naturalization Service, 1996 (M-367). U.S. Government Printing Office: Washington, D.C. Table 29. Also available on INS website at: http://www.ins.gov/graphics/aboutins/statistics/statyrbook96/introdu.pdf [PDF]

U.S. Immigration and Naturalization Service. 1999. Statistical Yearbook of the Immigration and Naturalization Service, 1997 (M-367). U.S. Government Printing Office: Washington, D.C. Table 29. Also available on INS website at: http://www.ins.gov/graphics/aboutins/statistics/1997YB.pdf [pdf]

U.S. Immigration and Naturalization Service. 2000. Statistical Yearbook of the Immigration and Naturalization Service, 1998 (M-367). U.S. Government Printing Office: Washington, D.C. Table 29. Also available on INS website at: http://www.ins.gov/graphics/aboutins/statistics/1998yb.pdf [PDF]

U.S. Immigration and Naturalization Service. 2000. Monthly Statistical Report, September 2000 FY Year End Report. Issued October 31, 2000. From INS website at: http://www.ins.gov/graphics/aboutins/statistics/msrsep00/index.htm

U.S. Department of Justice, Executive Office for Immigration Review. 2001. Statistical Yearbook; 2000. From EOIR website at: http://www.usdoj.gov/eoir/statpub/SYB2000Final.pdf [PDF]

U.S. Department of Justice, Executive Office for Immigration Review. 2001. Chart 4. Asylum Statistics, Fiscal Year 1997. From EOIR website at: http://usdoj.gov/eoir/FY97AsyStats.pdf [PDF]

U.S. Department of Justice, Executive Office for Immigration Review. 2001. Chart 4. Asylum Statistics, Fiscal Year 1998. From EOIR website at: http://usdoj.gov/eoir/FY98AsyStats.pdf [PDF]

U.S. Department of Justice, Executive Office for Immigration Review. 2001. Chart 4. Asylum Statistics, Fiscal Year 1999. From EOIR website at: http://usdoj.gov/eoir/FY99AsyStats.pdf [PDF]

U.S. Department of Justice, Executive Office for Immigration Review. 2001. Chart 4. Asylum Statistics, Fiscal Year 2000. From EOIR website at: http://usdoj.gov/eoir/FY00AsyStats.pdf [PDF]

U.S. Department of Justice, Executive Office for Immigration Review. Undated. Asylum Statistics For Period 10/01/89 - 09/30/90. Base City: All Offices Included. Unpublished table. Similar unpublished tables for fiscal year periods through 10/01/95 through 09/30/96.

Office of the United Nations High Commissioner for Refugees. 1992. Handbook on Procedures and Criteria for Determining Refugee Status, (Geneva, 1992). Website: http://www.unhcr.ch/refworld/legal/handbook/handeng/hbtoc.htm

U. S. Department of State, Bureau of Consular Affairs. 2000. Report of the Visa Office, 1998 (DOS Pub. 10799).

U. S Department of State, Bureau of Consular Affairs. 1998. Report of the Visa Office, 1997 (DOS Pub. 10577).