Nursing shortages are notoriously cyclical, and experts predict that if, as anticipated, unemployment rates drop to 6.1% by 2015, regional nursing shortages will begin to reappear.1 Historically, the United States has relied on foreign-educated nurses (FENs) to mitigate such shortages, and between 2004 and 2008, that reliance reached its highest level yet. The percentage of FENs in the U.S. nurse workforce grew by 29% during that period, with FENs making up 8% of newly licensed employed RNs, compared with 5% before 2004.2 In some states, such as California and New York, FENs now account for about 20% or more of the employed workforce.3-5

Since 2008, U.S. recruitment of FENs has declined as a result of the economic recession and retrogression. But proposed comprehensive immigration reform legislation (the Border Security, Economic Opportunity, and Immigration Modernization Act, which at this writing has passed the Senate but has not yet been voted on in the House of Representatives) may remove current limits to the number of immigrants granted visas to enter the United States annually. If enacted, this legislation could soon clear the path for thousands of FENs to enter the United States. We estimate that approximately 13,000 nurses from India and 35,500 nurses from the Philippines may have signed contracts with U.S. recruiters and are waiting in the pipeline for visas. Furthermore, the proposed legislation includes RNs in the new temporary guest worker program, called the W-visa program, potentially providing an additional entry path for nurses who until now have used the EB-3 visa (an employment-based permanent visa, or "green card").

Whether and when there will be jobs for these nurses is an open question. But if, as predicted, unemployment rates drop and nursing shortages recur, we would be wise to prepare now by examining what we learned during the last period of high demand for FENs. If international recruitment again surges, health care leaders must ensure not only that recruitment is limited to countries with a nursing surplus, but also that the terms and conditions of recruitment are fair and that FENs are treated equitably in the workplace.

In earlier research, we surveyed FENs to examine problems they experienced during the recruitment process.6, 7 We found that about half of FENs actively recruited between 2003 and 2007 experienced one or more violations of the standards described in the Voluntary Code of Ethical Conduct for the Recruitment of Foreign-Educated Health Professionals to the United States.6, 8 Nurses from low-income countries and nurses recruited by staffing agencies were the most likely to experience abuses, while nurses who found work without a recruiter experienced the fewest problems.6 We suggested that the length of contracts (often three years) and breach-of-contract penalties (often as high as $20,000, and in some cases higher) could explain these differences, since recruited FENs may be unable to change jobs because of these constraints.6, 7

In this cross-sectional study, using the same survey data, we aimed to establish a baseline from which future improvements could be tracked. We examined FENs' perceptions of their treatment in the workplace after recruitment to the United States. Specifically, we tested associations between FEN demographics, recruitment models, visa type, and four self-reported outcomes of interest: initial hourly wages; perceived adequacy of orientation to the United States; perceived inequities (discrimination) in certain areas; and job satisfaction. FENs were asked to report their initial hourly wage in their first U.S.-based position. Perceived adequacy of orientation measured three components: clinical orientation, orientation to culture of patients, and orientation to neighborhood and community. Perceived discrimination also measured three components: respondents were asked whether their wages, benefits, and shift or unit assignments were comparable to those of U.S.-educated peers at the same facility doing the same job. (Regarding assignments, FENs also had the opportunity to identify "race, gender, nationality" as a possible cause of any inequity.) FENs were also asked to rate their job satisfaction on a five-point rating scale (1 indicating lowest satisfaction, 5 indicating highest). Our working definition of inequity was based on Whitehead's idea that inequity has a moral and ethical dimension: it refers to differences that are not only unnecessary and avoidable but also unfair and unjust.9

BACKGROUND

Research has shown that, for nurses, a safe workplace, adequate orientation, competitive wages, and job satisfaction are among the factors essential to a positive work environment.10 Positive workplace environments, in turn, predict retention among nurses,11-14 including FENs.15 Moreover, many of the identified barriers to a positive work environment for FENs, such as inadequate orientation, pay inequities, and perceived discrimination, have implications for patient safety and quality of care.16-18

Earlier studies investigating discrimination against FENs in the United States have been primarily qualitative.19, 20 An investigation by DiCicco-Bloom in 2004 that focused on 10 FENs from India found reports of discrimination in job assignments and opportunities for promotion.21 Xu has reflected on the experiences of discrimination and marginalization reported by FENs at a conference on cultural diversity in nursing.22 Other foreign-educated health care professionals have been studied as well; a study by Chen and colleagues of 25 foreign-educated physicians found reports of workplace bias and discrimination.23 And similar experiences have been reported by FENs working outside the United States. A review of 19 quantitative and qualitative studies on African nurses working in the United Kingdom concluded that many reported negative experiences, including poor pay, discrimination, and racism.24 A Canadian survey of 6,477 RNs found that, compared with their Canadian-born counterparts, FENs reported higher rates of physical, verbal, and emotional abuse, and had more difficulty securing unpaid time off, tuition benefits, and flexible schedules.25

Schumacher found that non-Canadian foreign-born RNs working in the United States for six years or less earned 4.5% less than U.S.-born nurses.26 It's of interest that this difference disappeared after the six-year mark, after which, overall, foreign-educated RNs earned 12% more than U.S.-educated nurses. Schumacher suggested that those later, higher wages might be a result of FENs being more likely to have full-time employment, more schooling, and more experience than their U.S. counterparts. But he offered no explanation for the wage differential in the early years.

Some U.S. nursing leaders have reported feeling insufficiently prepared to address the transitional needs of FENs.27 A survey of U.S. nurse executives by Davis and Kritek revealed that the most critical skill for FENs is language, but that most orientations focus primarily on clinical skills.28 And Adeniran and colleagues have argued that sociocultural differences, language difficulties, and unfamiliarity with new surroundings-rather than gaps in knowledge or clinical skills-pose the biggest challenges to FENs.29 Indeed, to address this, they developed a model program, Transitioning Internationally Educated Nurses for Success, for the Hospital of the University of Pennsylvania.

Despite early wage differentials, job satisfaction rates do not appear to be lower among FENs than among U.S.-educated RNs.30, 31 This is also supported by our analysis of findings of the 2008 National Sample Survey of Registered Nurses (NSSRN): of nurses who entered the profession between 2004 and 2008, 77% of U.S.-educated RNs and 80% of FENs reported satisfaction with their careers.2

METHODS

Data collection. One of the best data sources on FENs is the VisaScreen certificate database. VisaScreen certificates are issued by the Commission on Graduates of Foreign Nursing Schools (CGFNS) International, a nonprofit organization named by the Illegal Immigration Reform and Immigrant Responsibility Act of 1996 to review the credentials of most foreign health care professionals seeking an employment-based visa. A VisaScreen certificate is required for all U.S. employment-based visas; only a few categories of health care professionals (such as physicians; and nurses who are supervisors, students, or dependents) do not need to obtain this certificate.

In 2008, we administered a survey to a random sample of 20% of all health care professionals who received a VisaScreen between 2003 and 2007, whose e-mail addresses were available, and who had not been previously surveyed during a pilot test of the survey instrument. The survey was e-mailed to 7,740 health care professionals, of whom 79% were RNs; the other 21% included LPNs, physician assistants, physical therapists, occupational therapists, clinical laboratory technologists and technicians, speech-language pathologists, and audiologists. The survey was active for one month. (The initial invitation was e-mailed on September 4, 2008, with follow-up reminders sent on September 18, September 25, and October 2. The survey closed on October 3, 2008.) We received 1,664 responses to the survey, of which 38% (629) were from RNs working in the United States at the time of the survey. Of these, nearly 80% reported on three key variables required for this study-whether they had been actively recruited, the country in which they had been educated, and confirmation that they had not been educated in the United States-for a final sample size of 502 FENs.

The country-of-education distribution of FENs who responded to the survey was consistent with that for all FENs who received VisaScreen certificates between 2003 and 2007 and for the sample frame.

For this study, all outcomes of interest were related to FENs' employment experiences in the United States. The survey questions were developed based on information gathered from two focus groups with FENs that we conducted in 2007.7 We pilot-tested the preliminary survey on VisaScreen holders and made minor modifications before administering the final survey. In the final survey, FENS were asked questions about their initial recruitment and initial employment in the United States. In keeping with the study's timeline, the questions assumed that responding FENs would still be bound by their first U.S. work contract, as typical contracts last about three years. For most categorical questions, FENs were asked to indicate "yes" or "no" or to select from a list of options.

Data analysis. Since nearly all employment questions in the survey were categorical, statistical testing was performed using [chi]2 analysis or analysis of variance (ANOVA). With regard to perceived adequacy of orientation and perceived discrimination, we calculated the number of reported negative events per respondent and used a binary variable to indicate whether the FEN had reported at least one negative experience. We treated hourly wage and level of job satisfaction as continuous variables. During data capture, any nonresponses to categorical or continuous questions were coded as missing.

First we performed a descriptive analysis on the FEN sample using four control variables: recruitment model, visa type, country of education, and opportunity to join a union. Opportunity to join a union was used as a proxy for regional and workplace differences in wages. Our reasoning was that unions are primarily located in states with high urban populations and in hospitals, which may pay more than home health or long-term care facilities. The visa type, country of education, and union variables were cross-tabbed against recruitment model and then tested using [chi]2 analysis. Then we analyzed the outcome metrics for all respondents and by recruitment model. Categorical outcome variables were analyzed using [chi]2 analysis; continuous variables were analyzed using ANOVA. Outcome counts were also analyzed but not used in the regression analyses.

Our four outcomes of interest were analyzed using the same four control variables. The orientation and discrimination regression analyses also included hourly wages as a control. The job satisfaction regression included hourly wages as well as the orientation and discrimination outcomes. Two regression models were run for each outcome. In the first model, we categorized country of education as either "low income" or "high income" based on each country's gross domestic product (GDP) per capita. (A low-income country was so defined if in 2009 it reported a GDP of less than $20,000 USD; a high-income country was so defined if in 2009 it reported a GDP of greater than $20,000 USD.32) In the second model, we categorized country of education as "low income-not Philippines," "Philippines," "high income-not Canada," or "Canada," to help control for the two largest FEN populations: those educated in the Philippines or Canada.

Ordinary least squares and logistic regressions were run with robust standard errors. Hourly wage data were analyzed using ordinary least squares, and, for ease of interpretation, we also performed logarithmic transformations on these data. Satisfaction was analyzed by turning the five-point rating-scale question into a dichotomous variable in which a higher satisfaction score of 4 or 5 was categorized as a 1. Responses indicating insufficient orientation or discriminatory practices were analyzed using a binary variable to indicate that the nurse had had at least one negative experience. Logistic regression analyses were conducted on the resulting dichotomous variables for job satisfaction, insufficient orientation, and perceived discrimination. Because not all FENs responded to all outcome or control questions, only 302 responses were used to analyze hourly wage, orientation, and discrimination regressions; and 300 responses were used to analyze the satisfaction regressions.

FINDINGS

Sample. Of the 502 nurses in our sample, 32% reported that they found work after arriving in the United States and the remainder had been actively recruited. Of those actively recruited, 14% had been recruited by a staffing agency, 26% had been hired directly by a hospital or other health care organization, and 28% had found employment through a placement agency. A majority of respondents (68%) had been educated in low-income countries. The majority of FENs from low-income countries were educated in the Philippines (53%), while the majority of FENs from high-income countries were educated in Canada (87%). Further, when we compared our sample with the 2008 NSSRN sample of FENS, we had a higher proportion of Canadian (28%) and lower proportions of Filipino (37%) and Indian FENs (16%). The distribution of FENs by country of education and employment model differed significantly within each category (see Figure 1 at http://links.lww.com/AJN/A51).

Three types of visas were included in our analysis: EB-3 visas ("green cards"); H-1B visas; and other temporary visas, which included TN status (a three-year guest worker program recognized under the North American Free Trade Agreement for Mexican and Canadian citizens), student visas, dependent visas, and tourist visas. (Only one respondent listed lottery, so that was dropped from the analysis.) Most (96%) of the TNs in our study were Canadians. Most self-directed nurses used a temporary visa, while the majority of actively recruited nurses used a green card. H-1B visa status was held by only 3% of the sample. The low use of H-1Bs may reflect the fact that this visa requires a baccalaureate to meet both its educational and employment requirements, and entry-level nursing positions in the United States usually do not. About 31% of respondents reported having an opportunity to join a union. (For further description of the sample, see Table 1.)

Table 1. Description of the Sample

Descriptive analysis of the four outcomes.Hourly wages and job satisfaction. We found statistically significant differences in reported hourly wages across different recruitment models: self-directed FENs reported an average hourly wage of $32.31, while those actively recruited reported an average hourly wage of $28.01. For all respondents, the average reported job satisfaction score was 3.54; differences among recruitment models were not statistically significant.

Adequacy of orientation. About a third of all respondents reported that they had not received sufficient orientation to life in the United States from their employers or recruiters or placement agencies. A similar proportion reported receiving insufficient orientation to the culture of their patient populations from health care employers. About a fifth of all respondents reported insufficient clinical orientation to their new workplaces, even though very few FENs (4% to 13%) reported not having a clinical preceptor. Over 50% of self-directed FENs, FENs hired directly by a health care organization, and FENs recruited through a placement agency reported experiencing at least one insufficient orientation practice. Perceived adequacy of orientation was not correlated with recruitment model.

Perceived discrimination. FENs recruited by staffing agencies reported much higher levels of perceived discrimination with regard to salary and benefits compared with all other FENs. They were also more likely to report perceived discrimination in shift or unit assignments. Overall, significantly more FENs recruited by staffing agencies (68%) reported experiencing at least one discriminatory practice than other FENs. (For more details on this and other findings from the descriptive analysis, see Table 2.)

Table 2. Outcome Metrics by Recruitment Model

Regression analyses of the four outcomes.Hourly wages. The wages regression analysis confirmed what we found in the descriptive analysis: that FENs recruited by a staffing agency reported significantly lower wages than self-directed FENs. (For more details on this and other findings from the regression analyses, see Table 3 at http://links.lww.com/AJN/A52). The regression model suggests that self-directed nurses earned nearly 11% more than FENs employed by staffing agencies, nearly 7% more than those employed directly by health care organizations, and nearly 6% more than those with placement agency contracts (see Table 4 at http://links.lww.com/AJN/A53). Wages were nearly 14% higher for FENs educated in high-income countries compared with those educated in low-income countries, a statistically significant difference. When we omitted Canadian and Filipino FENs from the analysis, we found that FENs from high-income countries reported 28% higher wages than FENs from low-income countries, and this difference was also statistically significant. Philippines-educated FENs earned slightly lower wages than FENs educated in other low-income countries, although this difference was not statistically significant.

Job satisfaction is often used to represent a nurse's overall employment experience. Yet, after controlling for country of education, recruitment model, union status, visa type, wages, and perceived insufficient orientation or workplace discrimination, we found that (with the exception of the Canadian-educated FENs) neither country of education nor recruitment model was associated with reported job satisfaction. The best predictors of job satisfaction were workplace practices, as indicated by the outcomes for adequacy of orientation and perceived discrimination. FENs who reported receiving insufficient orientation were significantly less likely to be satisfied with their jobs than those who had adequate orientation. FENs who perceived workplace discrimination were even less likely to report job satisfaction.

Adequacy of orientation. We found that FENs educated in the Philippines or Canada were less likely to report insufficient orientation than FENs from other low- or high-income countries. We found no relationship between recruitment model and insufficient orientation.

Perceived discrimination. We found that as wages rose, FENs perceived significantly less workplace discrimination in terms of wages, benefits, and shift or unit assignments. We also found that FENs employed by a staffing agency were significantly more likely to report perceived discrimination than self-directed FENs. Respondents from high-income countries were significantly less likely to perceive discrimination than those from low-income countries.

DISCUSSION

Our study focused primarily on the perceptions of FENs, and several of our findings were alarming and merit further research. Overall, 40% of the FENs in this study perceived their wages, benefits, or shift or unit assignments to be inferior to those of their American colleagues. Our findings suggest that although not all FENs report such inequities, certain groups of FENs may be especially vulnerable-in particular, those recruited by staffing agencies and those from low-income countries. FENs recruited by staffing agencies reported the highest rates of perceived inequities in wages, benefits, and shift or unit assignments. Self-directed FENs fared best in terms of receiving the most equitable wages. Among all other FENs, those recruited directly by a health care organization reported the fewest wage and benefit inequities.

We also found strong associations between country of education and reported wage inequities. Overall, FENs from high-income countries reported earning more than FENs from low-income countries. And this wage differential was even higher after we omitted Canadian and Filipino FENs from the analysis. Philippines-educated FENs earned less compared with FENs from other low-income countries, although the difference wasn't statistically significant.

The one outcome measure for which recruitment model and country-of-education status did not appear to matter was adequacy of orientation. Not surprisingly, Canadian and Filipino FENs, whose nursing education systems are perhaps most similar to those in the United States, were the most satisfied with the amount of orientation they received, suggesting that FENs from other nations may need special support.

Regarding job satisfaction, overall, FENs reported relatively high levels of satisfaction, a finding consistent with results from the 2008 NSSRN. Yet FENs who perceived workplace discrimination were significantly less likely to report job satisfaction. Thus, while in general FENs tend to report being about as satisfied with their work as their American counterparts, there are differences among subgroups of FENs, suggesting that the measure continues to be valid and an important measure for employers to track.

These results are not definitive; there is room to challenge our findings based on variables for which we were unable to collect information. For example, while we used "opportunity to join a union" as a proxy for regional and workplace differences in wages (unions tend to be located in urban areas and in hospitals, where wages tend to be higher), it's possible that other factors are in play. For example, FENs recruited by staffing agencies and FENs from low-income countries might have fewer years of experience than FENs in other groups, which could explain why they are paid less. They might also have been assigned to areas of the country where prevailing wages are lower. Similarly, the higher wages reported by self-directed FENs might be a function of their having had more education and experience, finding work in large urban areas (because that's where relatives had settled), or having better social networks in the United States (allowing them to shop around for the best job offer).

On the other hand, the constellation of observed differences among groups of FENs merits reflection. It's plausible that these differences are evidence not only of perceived discriminatory practices, but also of a real difference in treatment that is unnecessary and unfair. For example, if differences in wages and benefits were simply a function of location or experience, it's unlikely that FENs would report perceived discrimination in those areas when asked how they fared compared with U.S. nurses in the same workplace.

Let's consider some possible explanations for these problems. First, deficiencies in oversight of the prevailing wage system have been reported by employers and labor organizations.33 All U.S. employment-based visas require that an employer pay a foreign-born worker the prevailing wage for the worker's locality or match the wage it pays other workers with the same level of experience and education. But nursing wages for FENs are complicated. The U.S. Department of Labor's Office of Foreign Labor Certification has established four wage levels for nurses and other immigrant workers based on education and experience and on job descriptions.34 Even though recruiters generally look for "the best and the brightest," job descriptions produced for the purposes of visa filing likely call for the lowest level of experience so that recruiters can offer the lowest possible wages. It's possible that many FENs are misclassified at lower wage levels than their American counterparts would be. Moreover, FENs still in their home countries likely have little access to information about the nursing marketplace or their rights under U.S. law. Thus a FEN with nine years of experience might accept an entry-level job for a number of reasons: she or he might not know the prevailing wage for someone with nine years' experience, might consider only that the offered salary is higher than would be possible in the home country, or might believe this will be her or his only opportunity to migrate.

Second, differences in the experiences of FENs compared with their U.S. counterparts could be a function of race, ethnicity, or language. Indeed, findings from several qualitative studies suggest that this may be the case.19-21, 24, 25 Another factor might be a perception on the part of some employers that, regardless of education and experience, FENs from low-income countries will take time to adapt to American culture and health care practices and therefore initially deserve lower wages. In conducting previous research, we learned that staffing agencies with domestic and international divisions typically use different contract models for each division, with Canadians treated essentially as part of the domestic workforce.7 For example, Canadian FENs often aren't asked to sign a contract or aren't subject to high breach-of-contract penalties, and might be offered better job options, such as higher-paying travel nurse assignments.

The peculiarities of international staffing agencies urgently warrant further research. To date, research on the use of supplemental nurses has focused on their education and experience and on demographic factors, including race and ethnicity.35, 36 But we found no research comparing domestic staffing agencies with those specializing in FENs. Given that the latter typically require FENs to sign multiyear contracts while they are still in their home countries and impose high breach-of-contract penalties, it's possible that positive incentives aimed at retention are less necessary-and this could potentially affect a range of wage, benefit, and other work conditions.

Limitations. As this survey was cross-sectional, no causal relationships could be determined. Moreover, the data were collected retrospectively, requiring FENs to remember what they had experienced some time ago; for some respondents, this was as long as five years before the survey. A weakness of all retrospective surveys is that respondents might not recall some aspects of their experience and might underreport adverse events. Ideally, future researchers will find ways to survey FENs closer to the time of recruitment.

This study had several limitations regarding its generalizability. Although the VisaScreen certificate database is the best source available, the sample frame included several groups of respondents that had to be removed, including FENs educated in the United States, those who hadn't indicated their country of education, and those who hadn't reported whether they had been recruited. When we compared our sample with those included in the 2008 NSSRN, we found that Canadian-educated FENs were slightly overrepresented (28%) and that Filipino- and Indian-educated FENs were slightly underrepresented (37% and 16%, respectively). Thus, our sample may underrepresent the experiences of FENs from low-income countries versus those from high-income countries. But it's important to note that the effect of this bias on our findings would be to understate their magnitude, since respondents educated in high-income countries reported less discrimination and higher wages than those educated in low-income countries.

The overall survey response rate was 21.5%-a low rate that increases the possibility of response bias present in surveys of this kind, since people with negative experiences are arguably more likely to respond. But again, the distribution of respondents might mitigate this effect. Of the 502 respondents, a disproportionate number were educated in Canada; and Canadian FENs were far more likely to report positive experiences. The higher proportion of Canadian respondents might result from the Canadian nurses' having more trust in the confidentiality of Internet surveys, whereas nurses from low-income countries might have less trust and thus be more afraid of reprisals for their answers. If that's the case, the actual percentage of FENs perceiving discrimination could be higher than our findings indicate.

Ideally we would have weighted our data according to our sample demographics. But there are no data on the total FEN population employed in the United States at the time of the survey; thus we lacked reliable data from which to develop weight variables. Similarly, our findings regarding wages must be viewed with caution because we did not collect data on geographic location, job titles, or years of experience. Wages were not adjusted to 2008 dollars because FENs could have been hired initially at any time between 2003 and 2007. We could not control for either workplace setting or state of employment, making it impossible to produce a comparison to "prevailing" wages. Future researchers should collect additional demographics about FENs, including years of experience, workplace settings, geographic locations, and job titles.

CONCLUSIONS

Despite FENs' immigration status as "skilled workers," and although nearly half of our respondents entered the United States with a permanent visa, our findings and those of other researchers indicate that large numbers of FENs experience inadequate orientation and workplace discrimination. These problems are more likely to occur with FENs recruited by staffing agencies and FENs coming from low-income countries.

Although our findings are based on FENs' self-reports and don't constitute legal evidence of discrimination, they raise both practical and ethical concerns that should matter to nurse executives and others striving to create positive workplace environments. Perceived workplace discrimination can have detrimental health consequences for the workers themselves, including hypertension, depression, substance abuse, and a variety of heart diseases.37-39 Kingma goes further, stating that "discrimination and marginalization of the international nurse threatens patient safety and disrupts the health team cooperation dynamic required to advance the delivery of care."40

To expand on our findings, future surveys should collect more information on FENs' demographics as well as their workplace experiences. Care should be taken to define and measure specific manifestations of discrimination. More metrics will assist in determining whether our findings are robust. It would also be useful to include survey questions that distinguish between the responsibilities and policies of FEN recruiters and those of the health care organizations that employ FENs; and to compare the experiences of nurses within single facilities.

The current proposed immigration reform legislation suggests that the next wave of foreign nurse recruitment may be upon us, whether or not the projected U.S. nursing shortages occur. Tens of thousands of FENs have already signed contracts and are awaiting their visas, and as that backlog clears over the next few years, our challenge will be to learn from the past. Employers, recruiters, and nurse advocates must pay better attention to the real terms and conditions of FENs' employment, looking carefully at wages, benefits, and shift and unit assignments, and working to bolster transitional orientation programs. Only in so doing can we ensure equitable treatment and improved retention of FENs.

References