Determinants of Employment among Well-Educated Refugees Before and After the 2007 U . S . Economic Recession

Little is known about how the overall employment conditions in a country impact the likelihood of employment of newly arrived refugees. In the current study, we compare employment and determinants of employment of highly educated Middle Eastern refugees to Michigan that arrived before and after the 2007 recession. We also look at self-reported barriers to employment. Results show that the general downturn of the economy made it substantially more difficult to secure employment, even for well-educated refugees. Thus, before the economic downturn, 22.9 % of refugees were unemployed as compared to 55.1 % once the recession had set in (p < 0.01). There were also substantially more self-reported barriers to employment after the economic downturn. The study points to the importance of understanding both individual characteristics and the general employment conditions in the new host country when studying variation in refugee employment success.


Introduction
Refugees face broad as well as specific barriers to employment and under employment [1][2][3][4][5][6][7] . Being employed increases both economic and social integration and, for refugees, offers the opportunity to gain self-esteem, facilitate new social contacts, and to learn or improve English language skills [8][9][10][11][12][13][14][15][16][17] . Lack of employment is also a significant risk factor for the development of poor mental and somatic health [8,[10][11] . Thus, securing employment for refugees has implications for social, vocational, as well as health benefits [8][9]10,17] . To date, there are limited studies in terms of determinants of refugee labor market participation combined with studying their interaction with the general labor conditions in the host country. For example, economic conditions and whether a country is in recession or not [18] . Typically, studies compare refugee experiences with immigrants from other countries or with host country citizens [19] . Most research has focused on limited cross sectional convenience samples of refugees, and the study population has chiefly been a heterogeneous mix of global refugees [17][18]20] . In addition, these studies did not account for ethnic and cultural comparisons and this makes inference from prior studies difficult to apply to educated refugees as a whole [10,20] . Parallel to these weaknesses, there is a lack of control for the facilitating and barrier determinants, which makes cause-and-effect conclusions difficult to come to in terms of well-educated refugees.
Studies to date suggest that individual (human capital) factors are important determinants influencing the employability of refugees [21] . These main factors are language proficiency [23] , acculturation [22] , education [6.23-24] and work experience in the country of resettlement [10] . Additionally, studies show that refugee women typically have lower levels of labor market participation than men [23,25,26] , and those aged 40 and over tend to have lower employment rates [27] . Factors further contributing to "de-skilling" are an individual's personality as well as work ethic, and self-determination [8,11,16,28] . Recent reports also reveal that psychological and physical trauma that refugees faced when fleeing their native countries may also affect refugee employability opportunities [8,11,23,29] .
2 Structural conditions at the time of arrival in the host country have also been reported to highly influence access to employment opportunities for refugees [2,23,30] . For example, due to economic downturns, labor market conditions, and restraints, skilled refugees are often accepting of or are inherently forced to accept lower status and lower paid jobs where language proficiency and work experience are less important, and thus the chances of improving language skills are further limited [10,22] . Moreover, according to Taylor, refugees are disadvantaged through "the decreased availability of post-arrival services,'' largely due to reductions in public financing of refugee programs and services [31] . Additionally, host country attitudes and discrimination toward refugees have proved to also play a role in affecting a refugee's chance of obtaining employment [8,[32][33][34] . According to Colic-Peisker and Tilbury, "racism has more impact on one's labor market marginalization than sexism and ageism" [27] . Moreover, pre-migration tertiary education may have limited effects on the employability of refugees, because of post-migration structural constraints, such as difficulties in the recognition of foreign diplomats and the recertification of credentials [4,10,28] .
This study focuses on the influence of self-reported facilitating and barrier level factors in the employment outcome of well-educated Iraqi refugees in the United States. An exploration of self-reported facilitating and barrier level factors in an already high unemployment market may further the understanding of the downward mobility of these otherwise highly qualified refugees. That is, one would expect that in financially constrained times, refugees would have a harder time securing a job. Using data from

Methods
The overall aim was to determine the relative importance of individual as opposed to labor market conditions in terms of employment success of well-educated Iraqi professionals. A formal collaboration was established between the International Society of Iraqi Scientists, headquartered in Michigan, USA and Wayne State University (WSU), Detroit, MI, USA. The vast network of Iraqi scientists was leveraged to advocate for the completion of the survey to its membership base. Direct e-mail requests for electronic survey participation were sent to the international membership base with a link to Survey Monkey. The members were asked to forward the link of the survey to everyone in their e-mail address book who is eligible to participate. The study was also announced in several Iraqi news web pages. The electronic surveys were collected over a four-month period, from July to October 2009. Voluntary completion of the survey constituted informed consent. Approval for this study design and analysis was obtained from WSU Human Investigation Committee.
A total of 1,061 surveys of highly educated Iraqi immigrants/refugees residing throughout the world were obtained through Survey Monkey. Of those, 68 surveys were discarded because they lacked complete information and or were ineligible for participation because they were over 65 years of age. The balance of 993 surveys was further stratified; 842 immigrants/refugees that resided outside the US were also excluded, leaving a pool of 151 subjects. Within this group, 3 additional participants were excluded, due to disability, leaving a homogenous study population of 148 refugees for analysis. This final study group represented about 14 % of the total respondents to the electronic questionnaire. This study sample was further classified into two groups: those arriving before and those arriving after the year 2007, which marks the initiation of the great recession.
Logistic regression was used to examine independent factors as predictors of employment. Analysis was conducted and all independent variables were accounted for (age, gender, marital status, level of education, pre and post-2007 immigration period and perceived barriers and facilitators to employment).

Questionnaire
The electronic survey consisted of 35 questions. The measures included basic demographics such as age, gender, marital status, pre-migration occupation held in Iraq, and level of educational attainment. The survey also included questions about current employment status, whether the employment status was in a professional or non-professional classification, and the perceived barriers and facilitators to employment. The barriers were self-reported and included categories of financial independence/not being ready to work, employer discrimination, poor language skills, lack of professional knowledge/skills, costs of necessary training, not knowing how to find and apply for a job, a lack of jobs available, and poor health. The facilitators involved in securing a job in the U.S. consisted of self-response to categories of spousal support, personal effort, and help from a relative/friend, special support, language ability, community support, training course, and U.S. process and system.

Analysis
Two of the variables, age and year of arrival into the U.S were dichotomized. Age was dichotomized into younger and older than 45 years of age, in order to gain a better understanding of how age may affect employment conditions. The year 2007 was chosen as the cutoff point to look at the effect of the beginning of the U.S. economic recession impacting employment among two separate waves of immigration into the U.S. Data were analyzed using SPSS version 20.0. Chi -square tests were used to test for group differences in demographic variables across refugee groups (pre 2007 vs. post 2007). Binary logistic regression was used to study the relationship between the two separate time period intervals of refugees against perceived barriers and facilitators, as well Lett Health Biol Sci | volume 1: issue 1 www.ommegaonline.org 2 as basic demographics affecting employment status. Table 1 shows the demographic characteristics of both groups (pre and post-2007 Iraqi refugees). There were significant differences between the pre and post-2007 refugee groups on the following independent variables: age, gender, marital status, education attainment, current employment status, and period of time in securing current employment from the time of entry into the U.S. Post-2007 refugees were generally younger, single; to add more females were represented than there were in pre-2007 group. The post-2007 group had more individuals who possessed a bachelor's degree or equivalent than the pre-2007 group. However, the post-2007 group had fewer individuals who possessed either an MS or PhD or MD degree. The post-2007 group was also far more likely to report not attaining any type of employment, and they were less likely to be employed in their original profession or have secured a skilled/technical job than their pre-2007 comparison group. Interestingly, the post-2007 group was far more successful at securing some type of employment within their first year of arrival to the U.S. compared to the pre-2007 group. Inversely the pre-2007 group was more successful than the post-2007 group at obtaining employment over the 2-3 year period. There was one significant difference reported between the pre and post-2007 groups regarding barriers to obtaining employment. The barriers reported were financial independence and not ready to work ( Table 2). There was a significant statistical difference with one of the facilitators of employment between the pre and post-2007 groups. The facilitator was the support of their friends to obtaining employment (Table  3). Table 4 shows binary logistic regression in which male refugees were nearly three times as likely to be employed compared to their female counterparts (OR 3.03; 95 % C.I. 1.367 -6.740). Pre-2007 Iraqi refugees were over four times more likely to have secured employment as compared to post-2007 Iraqi counterparts(OR = 4.23295 % C.I.1.849 -9.687).

Discussion
Our findings support our hypothesis that post-2007 educated and skilled Iraqi refugees face more multi-faceted barriers to employment (unemployment and underemployment) than their pre-2007 comparisons' potentially due to the severe downturn of the U.S. economy. Results from this study showed that post-2007 well-educated Iraqi refugees faced much higher unemployment and underemployment than their pre-2007 counterparts. Despite their prior academic and professional achievements, the post-2007 group was less able to secure suitable professional employment in alignment with their pre-migration experience as compared to those arriving before 2007. Additional findings revealed that the post-2007 group were more likely to be deskilled in getting the job although they held high degrees of education, and although this is in tandem with previous studies [8,11] , these studies did not take into account the economic conditions at the time of arrival for recent refugees. This may suggest that labor market conditions in the post-2007 economic climate were not receptive to employment for well-educated recently arrived refugees. Additionally, our study showed that only one facilitating factor-friend(s) support played a significant role in securing a job. This is in contrast to other studies [2,[22][23][24][25] that showed multiple factors impacting the likelihood of employment, and those routinely contain language ability.
Interestingly, in terms of employment, both refugee groups reported that language proficiency was not a barrier for them. This is in contrast to a previous study which showed that language proficiency and job training programs have been reported to be critical factors for securing a job [28] . This is also in contrast to a study conducted by Wright, et al [35] which found that English proficiency (worse English) significantly predicted refugee unemployment 2 years post-arrival. This may indicate that a shift toward other unreported or undiscovered factors being more important for their chances in securing a job. The plausible explanation for this is that the study subjects were well educated, and as such may have had a more comfortable proficiency in English than participants in previous studies. It is also possible that other studies did not differentiate as to educational levels between comparison groups. Since many prior studies have shown that multiple barriers are often reported by refugees attempting to secure employment [8,11,16] , we predicted that barriers overall would be more pronounced in a recessionary period like post-2007. This was not the case. Although our study subjects reported multiple barriers to securing employment; only one of them -financial independencewas statistically significant between the pre and post-2007 groups. These results appear to be new and in contrast to previous studies which suggest that refugees need to be armed with multiple assets to successfully compete for employment opportunities. This may be attributable to various reasons; what appears to stand out is that the post-2007 group being generally much younger may have not had enough work experience in the donor country. Although generally educated, they are not as well educated as their pre-2007 counter parts. This factor may somehow indirectly contribute to their lower unemployment rate in an already highly competitive work environment in a region which already has an overall high unemployment rate [8,11,16] . Lastly, the overall small sample size in the study population may not be representative of the global population studied.
It may give the perception, that current factors such as English as Second Language (ESL), job readiness that focuses on resume writing and job interviewing preparation, did not play a large role in securing employment. It may point in the direction of in-dividual factors that were more instrumental to obtaining employment -essentially friend(s) and personal efforts, particularly within the first year of arrival. This may also warrant additional attention around leveraging identified existing facilitators and seeking out additional job promoting factors to support the employability prospects of refugees. This approach becomes even more relevant and potentially cost effective in a weak labor market.
Overall, immigrant and refugee resettlement has the potential to deliver significant economic benefits to society [36] . As a whole, skilled and educated refugees constitute a rapid availability of various technical labor skills, many of which are in current shortage, such as engineers, pharmacists, physicians, etc. Collectively, skilled refugees will pay taxes, both directly and indirectly, by filling gaps in the labor workforce resulting in overall higher output. Resources that affect broad level factors have not been able to keep pace with the continually increasing influx of refugees, who need greater and more intensive interventions to adapt and integrate into the workforce. Potentially further adding to the challenges faced by educated refugees, Iraqis specifically, is that the local economy may have already absorbed as much as it can handle over the last 2 decades. Also, the recent acute and chronic load of recently arriving refugees may not be able to be supported or sustained by the local economy due to the local struggling labor market in general. Lastly, research examining the effectiveness of current refugee programs and services could increase the current cost effectiveness and future allocation of limited resources, by reducing or eliminating programs and services lacking efficacy and focusing on programs and services shown to work around pairing refugees with employment opportunities.

Study, limitations and strengths
This study has several methodological limitations. First, the study used a limited cross sectional convenience sampling of willing participants who had access to internet services. The overall sample size was relatively small. It is not clear if the data is representative of the pre and post-2007 refugee population in general. In parallel to this, the study questionnaire was limited to broad and general questions around employment; it did not delve into the use of public sponsored programs or the efficacy of supportive services aimed at supporting employment for post-2007 refugees.
The study's strengths resided in the exclusive inclusion of study subjects residing in Michigan. This was planned in order to compare a homogeneous ethnic/minority population in two separate waves of refugee immigration; pre and post-2007, and also to study the subjects residing in the same demographic labor market. At the same time, this inherent strength limited the data analyzed to participants residing in Michigan only, as such it was limited to the homogenous labor and economic climate of southeastern Michigan. Large, random sampled, cross sectional studies are needed to elicit and dissect mitigating micro and macro-economic level factors and detailed employment statistics to provide greater understanding of the factors of employability for pre and post-2007 Iraqi refugees.

Conclusion
This comparative study showed, generally, that post-2007 refugees were more likely to have a difficult time securing a job as compared to refugees that arrived before the 2007 recession. More focus and attention is needed to help maximize the facilitators, as well as to reduce the barriers that post-2007 refugees face in the employment process. We suggest that future strategies around skilled/professional refugees need to focus on individual employability factors as well as measures to overcome both personal and structural barriers to labor market participation. More research is needed in this area as refugees could play an important role in the economic development of countries, especially when they are highly educated and skilled.