Literature Review
    There is very little research that focuses particularly on the experience of immigrants in community colleges.  One exception was a 1996 Rand Corporation report on the education of immigrants, although even in that report, the discussion of community college was brief and superficial (Vernez and Abrahamse 1996).  To the extent that this study generated insights into community colleges, the authors concluded that race and ethnicity were more important factors than nativity.  That is, after taking account of race and other personal characteristics, whether an individual was born abroad or in the United States did not have any independent influence on college enrollment and completion.
    Researchers have studied immigrants in education in general and much of that research emphasizes that the relationship between nativity and education differs by ethnic group.  Rong and Grant (1992) conclude that "regressions predicting school years completed show variable generation-by-ethnicity effects.  Asian attainment increases sharply between immigrant and child-of-immigrant generations, leveling off thereafter.  Hispanic attainment improves with successive generations of U.S. residence.  Non-Hispanic white attainment peaks in the child-of-immigrant generation and declines for later generations" (1992: 625).  Data collected during the 1996-7  for the  National Post-Secondary Aid Survey also lends some support to the idea that race has a stronger influence than nativity on whether a student enrolls in a two and four-year programs (NCES 1999).  For example, 62 percent of all foreign born students who were enrolled in college that year were in two-year programs.  But 61 percent of the native born were. This difference was not statistically significant.  In contrast, blacks were much more concentrated in the two-year programs than whites (non-Hispanics)—69 percent for blacks compared to 60 percent for whites.  Hispanics were even more dependent on community colleges (70 percent).
    Some analysts have tried to explain the apparent higher levels of educational attainment among immigrants than among native minorities, especially blacks.  In a controversial argument, Ogbu (1991), for example, emphasizes cultural differences  engendered by a group's perceptions of the opportunities open to it in the society.  The progress of “caste-like” minorities, such as African Americans, is inhibited by a "low effort syndrome" that  initially developed as a coping response to the experience of subordination—and specifically, formal and informal exclusion from all but the most menial positions—and which leads to the emergence of an "oppositional culture" (vis-a-vis all institutions controlled by the dominant group) that becomes entrenched, such that it continues to influence behavior even after the group's situation has been transformed.
    Asians have attracted a great deal of attention because of their high rates of educational attainment.  A popular notion is that Asians have a particular cultural value that promotes education, but Sue and Okazaki (1990) argue explicitly against an "Asian values" explanation.  Hirschman and Wong (1986) invoke selective immigration, as well as historical exclusion from certain occupations such as crafts, leading to an emphasis on other avenues of mobility (including education), as explanations for greater Asian attainment.   But none of this research directly addresses the role of community colleges in the educational attainment of immigrants or in the differences between immigrant and native minority education levels.
    In contrast to research on the situation in the country as a whole, analysts at CUNY's administration have conducted some research on the experience of immigrants in the University, including some analysis of the differences among two- and four-year institutions (CUNY 1995).  Their report, published in 1995, pointed out that more than one third of the first time CUNY freshmen in 1990 were born abroad (and almost all of those were not US citizens), when only about 28 percent of the city's population was not born in the US.  Moreover, the report predicted that by the year 2000, the foreign born would account for almost one half of the starting freshmen.
    This paper builds on that research in two broad ways.  First, we focus particularly on the two year colleges.  Although the CUNY report did provide some comparisons between students in two- and four-year programs, we carry out a much more detailed analysis of the determinants of enrollment in the two types of program, and add a consideration of native minorities, something not done by the CUNY report.  Second, the CUNY report does not analyze educational outcomes.  Access to post-secondary education is certainly an important issue, but what students do with that access—that is, how much education they actually accumulate—is also important.  Moreover, in research on higher education in general, analysts have paid much less attention to educational outcomes than they have to access and enrollment.
    A  focus on the relative roles of the two- and four-year programs and colleges is particularly important at this time given the controversy in New York, and indeed in the country as a whole, about the relative roles of  these two types of institutions and programs.  First of all, community colleges enroll a far larger percentage of college students than most people realize.  For example, while most of the controversy at CUNY has focused on the four year colleges, in 1997, almost two thirds of CUNY first-time, first-year students enrolled in associates degree programs.2  As immigration has accelerated in the last decade, community colleges appear to be institutions that are well-suited to provide access to higher education for these newcomers.  Thus we would expect immigrants, especially recent immigrants, to be particularly concentrated in the two-year institutions.
The roles of the different programs is a particularly important issue for CUNY and other systems that have sought to shift remediation from the four-year to the two-year programs.  In the wake of the 1999 publication of the Schmidt Report, the CUNY administration promulgated a policy that remediation would no longer take place at the four-year schools, only at the community colleges.  Advocates of this policy suggest that eliminating remediation from the senior colleges will raise standards at those institutions and help to strengthen their educational benefits for those students who are already prepared to  do college level work (Schmidt, et. al.1999; Macdonald 1994). But  critics of the increased CUNY selectivity feared that it would restrict opportunities for the immigrants and minorities (New York Times 1998).  According to this perspective, restricting access to the senior colleges is a serious impediment mobility because a bachelor's degree is considered  the key to economic opportunity, and enrollment in a community college lowers the probability of earning a BA (Lavin and Hyllegard 1996,  Dougherty 1994).  Therefore, from this point of view, access to the four-year schools is the most important indicator of the extent to which the university provides economic opportunity.
    The controversies about enrollment in the two types of programs go well beyond the specific issues of concern to CUNY.   The large majority of traditional-aged college students, including those in community colleges, state that they would like to earn at least a bachelor's degree, yet students in two-year programs are much less likely to complete a BA than those in four-year programs (Dougherty 1994).  Alternatively, community college advocates suggest that these colleges offer a step into higher education that would be more difficult if only four-year schools were available. For example, Rouse (1999) found that community colleges did in fact lower the educational attainment of students seeking bachelors degrees, but it also provided access to higher education for students who probably would not have enrolled in a four-year school.  Nevertheless, Rouse's study is based on the existing situation in higher education and does not imply that reforms at the four-year schools might not make them more welcoming to more students, or indeed, that reforms at the two-year schools might not make it easier for their students to transfer.  Indeed, at root, the CUNY controversy is about how open the four-year schools should be to students who might be expected either to have difficulty succeeding in higher education or to avoid it altogether.
    But regardless of  the merits of the CUNY policy on remediation, the policy change is taking place in an era, as we have seen, of extreme growth of the enrollment of immigrants at the university.   Therefore, it is important to understand whether immigrants are currently concentrated in the two- or the four-year programs and the extent to which they use the two-year programs successfully to accumulate credits and earn degrees.

Data
    CUNY’s Office of Institutional Research and Analysis (OIRA) maintains data files for every fall semester cohort of first-time freshmen entering the University.  These files contain a record for each student in the given cohort, and they are updated annually.  They include a great deal of information collected during the application process, as well as numerous variables that record aspects of the students’ educational careers in the CUNY system.
    This paper is primarily built around an analysis of the fall 1990 cohort; however, descriptive statistics are also presented from the fall 1997 cohort file, in order to indicate the nature and scope of certain changes that have occurred in the intervening years.  As can be seen, for example, the proportion of foreign-born students in the freshmen cohorts rose dramatically over the course of the 1990s.  A few words are in order about these data.
    The 1997 cohort has 25,173 students.  Information on place of birth comes from the CUNY application form.  (This question has been included on the application since 1992.)  Data on place of birth are missing for 17.0 percent of cohort members; the percentage is somewhat higher for associate degree entrants (18.9 percent) than for bachelor’s entrants (13.6 percent).
    The 1990 cohort has 26, 575 students.  The application form for 1990 did not include a question on place of birth.  However, a survey was sent to all of the individuals who applied to be first-time freshmen.   The survey data for all respondents who entered the University has been merged with their cohort records.  The survey did include a question on place of birth, as well as various other aspects of individuals’ social background.  The analyses in this paper draw primarily upon the combined survey and cohort data.
    Of the members of the 1990 cohort, 8,332 responded to the survey.  In order to adjust for survey non-response, a set of weights has been calculated.  The weights were devised by running a logistic regression equation predicting the probability of response on the basis of factors for which population-wide data were available (the weight is the inverse of the predicted probability).  All of the figures we report for 1990 entrants represent weighted data.  Information on place of birth is missing for 18.3 percent of the weighted sample.  The proportion is slightly higher for associate entrants (19.5 percent) than for bachelor’s entrants (16.3 percent).
 

Immigrants at CUNY
    Many students born abroad get their post-secondary education at CUNY.  Indeed, in 1990, just over 33 percent of the entering students were born abroad (immigrants accounted for about 28 percent of the city's population in 1990).  Asia accounted for over 22 percent of the foreign-born students (10 percent from China alone) and South and Central America and the Dominican Republic accounted for another 21 percent. Another 50 percent were from the Caribbean (including 9 percent from Puerto Rico3 and 9.5 percent from Jamaica).
    By 1997, 48 percent of the entering class was born abroad.   This is perhaps the single most striking finding of this paper.  In seven year, the foreign born share of the entering cohort rose by 15 percentage points or 45 percent.  Adjusting to a shift of that magnitude would be difficult for any institution.
    Table 1 displays demographic information about the cohorts that entered CUNY as first year students in 1997 and 1990, including the breakdown between those entering 2- and 4-year programs. A few highlights of this table are worth emphasizing.  First, the large majority (82 percent) of the foreign born students in CUNY attended secondary school in the United States.  Second, contrary to our expectations, as a whole, the foreign born were not any more concentrated in 2-year programs than they were in CUNY as a whole.   Indeed, it was the native born who were slightly overrepresented in the associates degree programs—they accounted for 52.2 percent of all starting CUNY students but 53.4 percent of all starting 2-year students.  The two foreign-born groups relied somewhat differently on the two- and four-year programs, with immigrants who attended high school abroad slightly overrepresented and the US high school groups slightly underrepresented in the two-year programs.
Thus, overall, immigrants were not particularly concentrated in the two-year programs.  But the foreign-born are not a homogeneous group, as is suggested by Table 2, which displays the countries of origin for the immigrant members of the 1990 entering cohort.  Here we do see that some immigrant groups do in fact rely more on the two-year than the four-year programs.  The Asians are more likely to enroll in the four-year programs while immigrants from the Western Hemisphere are more concentrated in the community colleges.   So for some groups, the two-year programs do represent an important point of access to the CUNY system.
    Racial and ethnic differences among the foreign born are another source of variation.  We pointed out that much of the research literature that does focus on immigrant educational attainment highlights the ethnic and racial differences among nativity groups. Data displayed in Table 3 does show that both native and foreign born whites and Asians are overrepresented in the four-year programs, while blacks and Hispanics from both nativity groups are concentrated in the community colleges and two-year programs.

Assessment Tests
    According to the new CUNY policy, students who do not pass the CUNY assessment tests and therefore require remediation should enroll in the two-year programs. But the policy does include a significant exception that has important implications for the enrollment of immigrants.  Students categorized as English as a Second Language Students (ESL) would be exempt from this requirement.  Thus ESL students who failed assessment tests would still be able to enroll in the four-year schools.  By examining passing rates on assessment tests, we can get an idea about how that policy and the ESL exception, will interact with the growing immigrant enrollment at CUNY.
    Data presented in Table 4 suggests that the two-year programs already enroll disproportionately more students with low test scores.  Only 40 percent of two-year entrants pass the math test and about 45 percent pass the reading test.  At the same time, these data make it obvious why the policy is so controversial.  Although the four-year entrants do better than their two-year counterparts, over one fifth of the four-year entrants fail the math test and almost a third fail the reading test.  These data also show that the two-year programs tend to enroll native students with low math scores and immigrant students with low reading scores.  Thus immigrants in the two-year programs have low passing rates for the reading test and high passing rates for the math tests.  This probably reflects language problems among the immigrants.4  Disaggregating the immigrant groups by race and ethnicity shows that white, Asian, and Hispanic four-year entrants do particularly poorly on the reading test.  In contrast, the Asian four-year entrants do very well on the math test while the Hispanic immigrants have low scores on the math test as well.
    We draw two broad conclusions from the assessment data.  First, as expected, the two-year schools already enroll a disproportionate number of students with low assessment scores. Second, even though many of the foreign-born students in the four-year schools do not pass the assessment tests, their deficiencies are most pronounced on the reading tests.  Thus many of them will be ESL students and therefore exempt from the new policy.  As a result we do not expect to see a significant shift in immigrant students from the four-year to the two-year programs.

Analysis of the Allocation of Students in Two and Four-Year Programs
    In the following section, we shall make a more comprehensive examination of the determinants of enrollment in either a two or four year CUNY program.  Our single variable comparisons so far suggests that ethnicity is a much more important predictor than nativity of whether CUNY entrants will enter a two or a four year program.  It may be that ethnicity is related to another variable that we have not considered and that that variable is what is driving these results. Alternatively, nativity status may be negatively related to some other important variable and therefore nativity's actual influence on enrollment could be obscured.
    In this analysis, we will use the 1990 entering cohort primarily because we have data on a wider set of demographic variables, based on the special survey of that cohort conducted by CUNY, than we do for later cohorts.  In addition to the variables that we have already considered—nativity, ethnicity, and test scores, we will also consider the influence of the variables listed in Table 5.  The additional variables that we use include general demographic characteristics such as age and gender, measures of background socioeconomic variables, and of competing commitments, such as employment and parenthood.  We will discuss each briefly below.
    Community colleges often attract students who have other commitments and responsibilities.  Not only are community colleges almost always cheaper than four-year colleges, but they often schedule courses in such a way as to facilitate adult participation.  Indeed, nationally and in New York, community college students tend to be older.   For example, in 1997, three quarters of all CUNY four-year program first-time entrants were 18 and younger, while only 42 percent of the associate’s degree entrants was that young.  Employment, either full or part-time, and whether the student has a child under 18, also reflects competing commitments. Although certainly many working parents enroll in four-year colleges, community colleges tend to cater much more to the needs of these types of students.  And indeed, our data show that community college students are more likely to be working and to have their own children.
    In addition to the test scores, which we standardized as z scores in order to allow comparison between the tests, we also included other variables that would reflect the type of educational preparation received by the students.  Research has shown that a General Equivalency Degree is not in fact equivalent to a high school degree (Cameron and Heckman 1993).  Therefore, we would expect that students with GEDs would be more likely to enroll in a community college.  We also included dummy variables to control for the educational background of the student's parents (coded to reflect the highest level attained by either parent).  Students whose parents have more education tend to do better in school.  Thus we expect that higher levels of parental education will be associated with a greater likelihood of entry into a four-year program.  And, it is a longstanding finding of educational research that wealth also tends to be associated with higher school performance, net of other factors.      We therefore also include controls for household income.5
    The results of the logistic regression of the determinants of whether a first-time, first-year student starts in a two-year or four-year program are displayed in Table 6.  In this regression, positive coefficients indicate that the variable increases the probability of enrolling in a 4-year program.  The analysis shows that foreign born students who graduated from a US high school are more likely to enroll in a four-year program, while those who graduated from high school abroad are more likely to enroll in a two-year program.  Most of the control variables have the expected influence.  Those who have earned a GED, older students, those with jobs, and those with childcare responsibilities, students are all more likely to enroll in a community college even after controlling for all of the other included variables.  The influence of scores on math and reading assessments have the expected sign.  Women and students who state that they want a BA are more likely to enroll in a four-year program.
Introducing the control variables has interesting effects on the influence of ethnicity.  Blacks are still much more likely than whites (the comparison group) to enroll in a two-year program.  But surprisingly, there is no statistically significant effect for Hispanics or Asians.  This suggests that the Hispanic concentration in community colleges is a result of their low test scores, rather than some particular "Hispanic" effect.    Similarly, the Asian overrepresentation in four-year program can be explained by the characteristics of the Asian students without resorting to an "Asian culture" argument.
    Therefore, to some extent the CUNY community colleges are playing their expected role of  providing access to higher education for the city's immigrant population, especially those with very weak language skills, but the four-year schools also very much play that role, although they do so somewhat less for the foreign born whose took their secondary education abroad.  It probably makes more sense to think of CUNY as a whole as an extremely important immigrant-educating institution.  While we have identified some differences in the enrollment of immigrants in the two- and four-year programs, these differences are probably less important than the overall role of CUNY.
    Another important conclusion is that we have found nativity to be at least as important as race and ethnicity in determining the choice between these two types of schools.  This conclusion is in conflict with the general consensus in the research literature which tends to downplay nativity and emphasize ethnicity.  On the other hand, the analysis suggests that black students do have a statistically significant lower probability of enrolling in a four-year program.    We shall return to this result later in the paper.

Educational Outcomes for CUNY Native and Foreign Born Two-Year Entrants
    In this section we will focus on those CUNY students who start in two-year programs.  Fundamentally we would like to know how well immigrants are able to use the CUNY community colleges to achieve their economic and educational goals.  Since our data do not include any information on eventual economic outcomes, we focus here on educational outcomes.  From this point of view, what is the most appropriate measure of success for a community college student?  Community college advocates and critics have argued about this for many years.  As we shall see, after eight years, only 23 percent of those students who started in a two-year program have earned an associates degree.    Certainly a 23 percent graduation rate for a high school would be considered disastrously low.  On the other hand, many community college students transfer to a four-year institution without earning an AA degree, and such students should not be seen as a college "dropouts"; indeed, most community college personnel would see a successful transfer, even without a degree, as a successful outcome for the college.  Moreover, community college advocates argue that many of their students are looking for specific skills that can be learned in a set of specific courses and that they have no intention of or need to complete a degree.6  From this point of view, low completion or even transfer rates are more reflections of the diversity of roles taken on by these institutions than they are an indication of institutional failure.7
    Because of the ambiguity about the nature of a successful community college experience, we shall analyze five different outcome variables.  In the first, we count the number of degree credits earned by the students, regardless of whether they are taken in a community or a senior college.  Second, we consider simply whether or not the student has earned an associates degree. Third, we measure whether the student has transferred to a four-year program or institution (regardless of whether they ever finish that degree).  Fourth, we create a category in which we pool those who earned an associate degree and those who transferred.  Finally, we look at students who have transferred to a BA program and try to determine the characteristics that promote successful completion of that degree.  We address the problem of variation in students’ educational goals by controlling for stated aspirations.8
    Table 7 displays the distribution of credits earned (a student earns credit if he or she completes a course with a passing grade). The foreign born students earn significantly more credits than the native born.  Only 30 percent of the foreign born who graduated from foreign high schools earn only 20 credits or less, compared to over 48 percent of the native born.  In contrast, 48 percent of the foreign born (foreign high school) earned 60 credits (enough for an associates degree) while less than 30 percent of the native born accumulated 60 credits.
    Table 8 presents degree and transfer data.  On all of these measures, the foreign born (foreign high school graduates) are the most successful, while the native born are the least successful.  The experience of the foreign born with US high school degrees lies in the middle.  The most dramatic difference among these three nativity groups is in the rate of associate degree attainment rather, than in the transfer rate.  On the other hand, of those who did transfer—that is, those who ever enrolled in a bachelor's degree—42 percent of the foreign born (foreign high school) actually earned a BA degree, while only 35 percent of the native born earned that degree.9
    One interpretation of these results is that immigrants who graduated from high school abroad arrive with a reasonably strong underlying level of education, but with language deficiencies.  As we have seen, these immigrants have relatively high scores on  the math assessment test.  They then use the community college to strengthen their language skills.  Once that is achieved, they are able to accumulate credits and they are more likely to earn degrees.   This suggests that many of these immigrants did not come from the poorest social classes in the sending countries.  The 2-year programs allow them to overcome one particular weakness in their preparation.  Natives who have a similar type of education may be more likely to start directly in a four-year program.
    On the other hand, overall the data still show a low rate of degree completion.  Even though over three quarters of the students who enter the two-year program state that they aspire to at least a BA, eight years after enrolling, only one fifth have enrolled in a bachelors program.  Although the foreign-born in general do have more educational achievement, their chances of transfer are not much higher than those for natives.10
    Are these results the direct influence of some factors associated in particular with nativity, or are they the result of the characteristics that the immigrants happened to have?  Tables 9-12 present multivariate analyses of the determinants of credits earned, AA completion, transfer to a BA program, and BA completion (for those who transferred).  These analyses control for other factors that might influence credit accumulation or the probability of graduation or transfer.  In these regressions we use the same control variables that we used in the analyses of whether the student started in a two- or four-year program.  These included measures of alternative commitments (part-time status, work and childcare responsibilities), SES (parent's education, and whether the student was in SEEK or College Discovery), stated aspirations, and measures of educational preparation (GED and test scores).  Some CUNY institutions give both two and four year degrees.  In these analyses of educational outcomes, we include a dummy variable that indicates whether the student went to one of these “hybrid” institutions.  Analysts argue that combining two and four year programs in one institution facilitates transfer by obviating the need to apply to a new institution and move to a new location (Dougherty 1994).
    The results of this multivariate analysis are partially consistent with the two-variable comparisons presented in Tables 7 and 8.  Immigrants who attended high school abroad accumulate more credits and are more likely to complete an associates degree than natives. They are not more likely to transfer, and if they do transfer, they are not more likely to earn a BA than natives.  But once we add the controls, immigrants who graduated from high school in the US appear to have as much educational success (indeed more) than the those who completed high school abroad.  That is, the foreign-born-US-high-school graduates earn more credits and are more likely to complete an associates degree or to transfer, than natives.  The lower achievement of this group suggested by Tables 7 and 8 appear to result from their characteristics rather than an unmeasured difference between the graduates of US and foreign high schools.
    Interestingly, race and ethnicity appear to have little effect on these measures of educational success.  Hispanics do earn fewer credits and blacks are slightly less likely to transfer, but these variables do not influence the probability of completing and associates degree.  The most significant result is that blacks and Hispanics who do transfer do have a lower probability than whites of earning a BA.
    These Tables provide some interesting results concerning the effects of attending a school that combines two- and four-year programs (hybrid colleges).   As expected, enrolling in a hybrid college will increase the probability of transfer, but it will actually decrease the probability of completing an associates degree.  Moreover, once a student at a hybrid college has transferred to a four-year program, he or she does have a higher probability of earning a BA.

The Relationship Between Immigrant and Native Enrollments at CUNY
    Does the dramatic increase in immigrant enrollments at CUNY have an impact on the enrollments of native-born students?  Discussions of immigration policy often focus on the potential competition between immigrants and natives.  While much of this discussion is focused on employment, social services have also received some attention. Analysts have asked whether the arrival of immigrants in some sense crowds out natives from education or other services.  It is very difficult to make this type of argument in the case of CUNY.  Indeed, CUNY’s overall enrollments are considerably lower than they were in the mid-1970s (prior to the implementation of tuition).  Clearly the university has the capacity to expand if students are available.  To the extent that the Schmidt recommendations reduce enrollments in the four-year schools, there will be even less potential conflict between those groups.  Indeed, CUNY four-year school administrators who want to prevent a decline in their enrollments will be particularly interested in recruiting more foreign-born students.  And there is nothing in our analysis of the patterns of enrollments in the two- and four-year schools which would suggest the development of any conflict or competition between immigrant and native students.

Conclusion
    Perhaps the most striking conclusion from this paper is that the immigrant enrollment at CUNY has grown so rapidly.   Certainly, enrollments at CUNY reflect the immigrant character of the city’s population: the foreign-born are actually over represented among the university's students relative to their share of the population.  While the foreign born share of the population grew from just under 30 percent in 1990 to just over 40 percent in 1999, that share of CUNY enrollments grew from 33 percent in 1990 to 48 percent in the seven years ending in 1997.  Comparable data for 1999 undoubtedly would show a continued growth in immigrant enrollments at CUNY.  Thus in under a decade, the university has undergone a profound change in the nature of its students body.
    What about the relative roles of two- and four-year programs?  The new CUNY policy will, if fully implemented, increase the importance of the two-year programs.  Between one fifth and one third of the students in the four-year programs fail at least one of the  assessment tests.  But the ESL exception will mitigate the effect of that policy, especially for the foreign born who attended foreign high schools.  It seems likely that the policy will actually increase the concentration of native-born blacks and Hispanics in the community colleges, and as we have shown, these students are already over represented in the community colleges.  Immigrants, on the other hand, especially those with a US high school education, are, contrary to our expectations, somewhat more likely to enroll in a four-year rather than a two-year program.
    What about educational outcomes?  Just over a fifth of the two-year entrants earn an associates degree and about same number transfer to a four-year program.  Less than 10 percent earn a BA within eight years of initial enrollment, although over 70 percent stated that they aspire to complete at least a BA.  But immigrant two-year entrants do appear to have higher levels of educational achievement than natives who enter the same programs.  Immigrants earn more credits and are more likely to complete an associates degree, although they are not more likely to transfer or to complete a BA once they have transferred.  Our explanation for this is that, for many immigrants, underlying educational preparation is considerably better than these assessments suggest, although this is to some extent “masked” by problems with language skills.  But their experience in the community colleges gives them a chance to strengthen those skills, and eventually the influence of their stronger educational preparation does assert itself as indicated by greater credit accumulation  higher graduation rates.  According to this story, the community colleges do play a classic adjustment role, by providing a chance for immigrants to catch up by strengthening their weak English language skills.  But this effect is not strong enough to propel them to a higher probability of transfer.
    We also examined the enrollment patterns of racial and ethnic minorities in CUNY.  We found Hispanics and blacks particularly concentrated in community colleges, and implementation of the CUNY remediation policy would further concentrate them in the two-year programs.  But contrary to the consensus in the research literature, we found that minority status was less important than nativity in determining educational achievement.  Blacks and Hispanics did have a lower probability of earning a BA if they did transfer, but that was the only strong effect of race and ethnicity on any of our outcome measures.  Thus these groups certainly have a low chance of earning a BA, both because they enroll disproportionately in two-year programs, and if they do transfer, have less chance of finishing.  In another surprising result, we found that after controlling for test scores and personal characteristics, for the most part, there was no special Asian effect.
    High levels of immigrant enrollments are not new at CUNY, but the rapid growth of those enrollments during the last decade is unprecedented.  This increase in the number of foreign-born students has undoubtedly increased enrollments in the university as a whole, and should therefore probably be seen as a positive development.  Certainly the trend will create more pressure on the services in the university designed to strengthen language skills and provide developmental education.  But this service is a increasingly a core activity of urban public universities, and community colleges in particular.  One trend that we have identified that certainly deserves more attention is the continued and, we predict, growing concentration of native blacks and Hispanics in the two-year programs.  Of course enrollment in community colleges provides important opportunities for many students, but as minorities become more incorporated into the overall society and economy, we would expect to see their distribution among different educational institutions to reflect that distribution in the society as a whole.  At least we would like to know more about why these groups continue to be over represented in the two-year programs, even after controlling for test scores and other demographic characteristics.  In the end, it is important that the effort needed to adjust to the rapid increase in the immigrant enrollments not divert attention from the continued educational problems faced by many native groups.

Bibliography
“An Assault on Open Admissions,” New York Times, 19 March 1998, A20.
Bailey, T., & Averianova, I.E. (1998, October). Multiple missions of community colleges:  Conflicting or complimentary?  Community College Research Center, Teachers College,  Columbia University.
Cameron, S.V. and Heckman, J.J. (1993). The nonequivalence of high school
equivalents. Journal of Labor Economics, 11(1) Part 1, 1-47.
CUNY.  (1995).  Immigration/Migration and the CUNY Student of the Future.  Report prepared by the City University of New York.
Dougherty, K.J. (1994). The Contradictory College: The Conflicting Origins, Impacts, and Futures of the Community College. Albany, NY: State University of New York Press.
Hirschman, C., and M. G. Wong. (1986).  The Extraordinary Educational Attainment of Asian-Americans: A Search for Historical Evidence and Explanations. Social Forces, 65(1), 1-27.
Kane, T.J. & Rouse, C.E. (1995). Labor-market returns to two- and four-year college.  American Economic Review, 85, 600-614.
Lavin, D. E., and D. Hyllegard.  1996.  Changing the Odds: Open Admissions and the Life Chances of the Disadvantaged.  New Haven and London: Yale University Press.
MacDonald, H. (1994).  Downward Mobility: The Failure of Open Admissions at City University. City Journal, Summer, 10-20.
National Center for Education Statistics. (1999). National Postsecondary
Student Aid Survey: 1995-96. Washington, D.C.: U.S. Department of Education.
Data file.
Ogbu, J. U.  (1991).  Minority Coping Responses and School Experience.  Journal of Psychohistory, 18, 433-56.
Rong, X. L., and L. Grant. (1992). Ethnicity, Generation, and School Attianment of Asians, Hispanics, and Non-Hispanic Whites. The Sociological Quarterly, 33 (4), 625-36.
Rouse, C.E. (1995). Democratization or diversion?: The effect of community colleges on educational attainment.  Journal of Business & Economic Statistics, 13, 217-224.
Sue, S., and S. Okazaki. (1990).  Asian-American Educational Achievements: A Phenomenon in Search of an Explanation.  American Psychologist, 45 (8), 913-20.
Vernez, G. & Abrahamse, A. (1996).  How immigrants fare in U.S. education.  New York: Rand.
Schmidt, B.C., Badillo, H., Brady, J.V., MacDonald, H., Ohrenstein, M., Roberts, R.T., and Schwartz, R. (1999).  The City University of New York: An Institution Adrift.  Report of the Mayor's Task Force on the City University of New York, June 7, 1999.