The impact of the Great Recession on the ability of parents to fund their children’s college education

Abstract

In the United States, most families in the middle class or higher are expected to finance a significant portion of the cost of their children’s college education. Using data from the Longitudinal Study of American Life (LSAL), we analyse the impact of the Great Recession (GR) on the beliefs about responsibility and plans of parents in their mid- to late thirties to finance the post-secondary education costs of their children. Results demonstrate that the GR was unrelated to parents’ beliefs about their responsibility to finance their children’s post-secondary education, but it was associated with their plans for financing it. Parents who experienced a positive net impact of the GR were more likely to report having a savings plan and being able to borrow money if needed. Parents who experienced a negative net impact of the GR were more likely to report that their child needed a grant, scholarship or loan. College-educated parents were more likely to believe that parents have primary responsibility for financing their children’s education and to have a savings plan in place. However, parents who had had a student loan themselves and who had more children pursuing college were more likely to believe that parents have partial financial responsibility and that children should also contribute financially by getting grants, scholarships and loans.

Introduction

In the United States, most families in the middle class or higher are expected to finance a significant portion of the cost of their children’s college education (Bengtson et al, 2002; Massey, 2007; Steinberg et al, 2009; Putnam, 2015). When children decide to pursue careers in law, medicine and other fields with limited federal funding, it may be necessary for parents to provide part of the support of graduate or professional education. In this context, it is likely that the economic downturn associated with the Great Recession (GR) would create some difficulty for some parents in meeting these obligations. This analysis uses data from the Longitudinal Study of American Life (LSAL) to examine how the GR was associated with parents’ beliefs about their responsibility and plans to financially support their children’s college education.

Although the GR had a negative impact on a portion of adults in the LSAL – a national cohort of persons born mainly between 1972 and 1975, and aged in their mid-thirties at the time of the GR – it is important to note that some members of this same generation prospered throughout the GR. They report retaining their job or jobs, buying or continuing to make payments towards buying a home, taking vacations, buying a new automobile, and saving for the cost of their children’s education and for their own retirement (Miller and Cepuran, in this issue). Thus, our analysis will examine the perceived impact of the GR on parents’ beliefs and plans for educational funding.

The American system of higher education funding

For readers not familiar with the funding of higher education in the US, it may be useful to provide a brief summary prior to turning to the analysis of the data. Unlike Europe and many other countries, the federal (or national) government operates only four military academies and one peace academy. Slightly more than 60% of post-secondary students attend universities funded by state governments (McFarland et al, 2018; Snyder et al, 2018). For many years, state governments provided substantial funding for those institutions, especially for those colleges and universities engaged in educating teachers. In the 19th century, the Congress passed the Morrill Act to provide land to each state government to establish a college or university to advance the study of agriculture, engineering and other useful arts.

Many of the oldest and most distinguished universities in the US are private, not-for-profit institutions that were founded by various religious groups for the purpose of educating ministers to serve churches in the colonies. Harvard University, Yale University and other early colleges and universities were dedicated to the education of ministers. In most cases, the religious control of these institutions has ended or been greatly diminished, and these institutions have established large endowments from gifts from their alumni. As a result, it is possible for many of these institutions to have need-blind admissions, meaning that applicants are evaluated on their merits without regard to their ability to pay. All students selected for admission are provided with adequate support to cover the cost of that institution.

Few US institutions of higher education, however, have endowments that are large enough to cover adequately the full cost for all students selected for admission. Most educational institutions rely on a combination of federal and state funding and student tuition to cover expenses (Clotfelter et al, 1991). In recent decades, public opinion has shifted to the notion that higher education is more beneficial to the individual than society, and consequently, the government has been providing less financial aid to schools (Heller and Rogers, 2006). Due to decreased government funding, schools have had to increase tuition fees (McPherson and Shapiro, 2002; Baum and Ma, 2010). In 2008/09, the yearly cost of tuition and fees associated with higher education in the US was $28,440 for private, four-year colleges, $7,560 for public, four-year colleges, and $2,730 for public, two-year colleges (Ma et al, 2018). In addition to less government funding to institutions of higher education and rising tuition, the type of federal aid provided to students has shifted from grants to loans, which has increased the financial burden of paying for college on students (Price, 2004; Heller and Rogers, 2006; Baum and Steele, 2007).

For most students who plan to attend state universities or non-profit colleges or universities, parents and family are expected to provide a portion of the cost of post-secondary education (Henretta et al, 2012; Friedline et al, 2017). There is a universal federal student aid form that is used by all colleges and universities and it requires each applicant’s parents to report all of their income and assets. Depending on the ability of parents and family to contribute, there are federal student grant programmes and student loan funds. Students may also be expected to work part-time on campus in a library, in food service or in related support services through a federal work-study programme. Any individual student may have a financial aid package that involves some parental or family contribution, some direct grant funds, some work-study income and some funds from loans that the student will have to repay after they finish their formal schooling.

Responsibility and plans for financing children’s higher education

Although there is a general agreement in the US that parents should provide some financial support towards their children’s education (Steelman and Powell, 1993), there is less consensus on how much support they should provide, even among families of similar socio-economic backgrounds (Hamilton, 2013). In the past, nearly 75% of parents contributed financially to their dependent children’s education, and parental contributions comprised the largest source of monetary support for children’s post-secondary educational expenses (Churaman, 1992). It is unclear whether parents’ beliefs about their responsibility for financing their children’s education were impacted by their experiences of an economic downturn such as the GR.

Evidence from qualitative studies suggests that parents have been reconsidering their responsibility for incurring financial debt to finance their children’s college education. Unemployment rates among college students doubled during the GR (Hartmann et al, 2010), which led some middle-income parents to question the financial returns of a four-year college education and weigh the cost-benefit of incurring a large financial burden to pay for it (Iversen et al, 2011). In contrast, another study found that nearly all upper-middle-class parents believed in the necessity of a college education and a parent’s obligation to take on some of the financial cost; however, they also reported great variability in their ability to finance their children’s education, and parents who reported struggling financially expressed a need for the child to contribute financially as well (Holmstrom et al, 2011).

Parents’ increased general need and willingness for their children to contribute financially to their post-secondary education is evident in increases in college student employment rates and student loan debt. Some 56% of college students worked at least part-time in the late 1980s (Gleason, 1993) whereas 85% of college students at four-year public and private universities worked part-time in 2008 (U.S. Department of Commerce, 2016). Most students in the US graduating from four-year colleges in 2008 had student loan debt. Graduating seniors with student loans had an average debt of $23,340 with the highest debts reported by students at for-profit colleges followed by private non-profit colleges, and public colleges (TICAS, 2014). The average student loan debt has continued to increase with graduates of four-year colleges having an average debt of $28,650 in 2017 (TICAS, 2018). Given that large amounts of student debt can hamper students’ future accumulation of wealth, it is important to understand the factors associated with parents’ willingness and need to shift the financial burden of higher education to their children.

To date, few quantitative studies have investigated the parent and family factors associated with the degree to which parents assume responsibility for financing their children’s college education, particularly in light of the GR. Similarly, limited quantitative research has examined the extent to which these factors are associated with parents’ plans to finance their children’s college education. Prior to the GR, most parents paid for their children’s education using a combination of income, savings, and loans from private lenders and other family members (Churaman, 1992). However, it is unclear whether the rising cost of higher education in conjunction with the economic recession had an impact on parents’ plans for financing their children’s post-secondary education. Indeed, parent characteristics have been identified as an area needing more focus to help explain parental resource investment in their children’s post-secondary education (Steelman and Powell, 1991; Downey, 1995).

Parent and family socio-demographic characteristics and financing post-secondary education

Earlier research indicates that parents’ socio-demographic characteristics are associated with their willingness and ability to contribute financially to their children’s education. In general, parents with higher levels of income and education are more likely to believe that they should be primarily responsible for financing their children’s post-secondary education, and they are more likely to have the financial resources to do so (Steelman and Powell, 1991). Similarly, prior research has found that parents who are married and working tend to have more financial resources to contribute to their children’s education than their single non-working counterparts (Charles et al, 2007).

In addition to socio-economic differences in parents’ beliefs and plans for funding children’s post-secondary education, some racial/ethnic differences have been identified in extant research as well. Caucasian parents were twice as likely as their African-American, Hispanic-American and Asian-American counterparts to report that students should bear the primary financial responsibility of paying for college; conversely, racial/ethnic-minority parents were more likely to place the financial burden on the parent (Steelman and Powell, 1993). Despite differences in belief of primary financial responsibility, after taking other socio-demographic factors into account, parents made the same amount of effort to save for their children’s post-secondary education regardless of race/ethnicity (Steelman and Powell, 1993).

Another factor associated with parents’ beliefs and plans to pay for their children’s post-secondary education is the number of children expected to pursue education beyond high school. Based on resource dilution theory (Blake, 1981), parents have a finite amount of financial resources that they must spread among their children. Thus, parents with more children will have fewer resources to distribute to each child. Research on college financing using data from the 1980s provides initial support for this theory; parents with fewer children are more likely to report believing that parents have primary responsibility for financing their children’s post-secondary education while parents with more children are likely to attribute greater financial responsibility to the child (Steelman and Powell, 1989; Downey, 1995; Powell and Steelman, 1995). It is necessary to investigate whether this resource dilution hypothesis is supported across historical contexts, such as the GR.

Similarly, research on parents in the 1980s suggests that parents who received financial assistance from their own parents were more likely to attribute primary financial responsibility to the parent rather than the child, and that they were more likely to have saved more money for their children’s education than parents who received no financial assistance from their parents (Steelman and Powell, 1991). Whether this association is evident for parents who experienced the GR has not yet been investigated empirically yet.

Study objectives

In sum, the cost of a post-secondary education is becoming increasingly unaffordable in the US due to rising tuition, decreases in median family income, and the fact that student financial aid has not keep pace with tuition costs (NCPPHE, 2011). With the financial burden of funding children’s post-secondary education shifting from government to the individual, it is important to understand what factors are associated with parents’ financial decision making regarding their children’s college education. Thus, the present study seeks to understand the extent to which parent and family characteristics are associated with parents’ beliefs and decisions around financing their children’s college education in the context of an economic event such as the GR.

A primary goal of this study is to investigate whether study results from data on older cohorts (for example, Steelman and Powell, 1991, 1993; Hamilton, 2013) and smaller, qualitative samples (such as Iversen et al, 2011; Holmstrom et al, 2011) are replicated in a contemporary, nationally representative sample of US families. In addition, the present study extends the literature on how parents finance their children’s post-secondary education by examining several types of plans (saving; borrowing; expecting child to get a grant or scholarship, or take out a loan; receiving help from extended family). Importantly, this study contributes to the existing literature by providing an analysis of the relative importance of parent and family factors in coping with a negative event like the GR. The GR was the first major economic crisis experienced by this generation of adults and it is important to understand the factors that are associated with individual family reactions to the GR in a longitudinal context.

The present study seeks to answer the following research questions within the context of the GR:

1. What parent and family characteristics are associated with parents’ beliefs about their responsibility for financing their children’s post-secondary education?

2. What parent and family characteristics are associated with parents’ plans for financing their children’s post-secondary education?

Methods

These analyses use data from the Longitudinal Study of American Life (LSAL), an ongoing study that began collecting data in 1987 on a nationally representative sample of public school students in grades seven and ten (ages 12/13 and 15/16) in the US and has surveyed them annually thereafter. The sample for the present study was based on LSAL study participants with children in the year 2015 (N=1,529).

In 2014, the LSAL asked its national sample of adults aged 39 to 42 to report on plans for the financing the higher education of their children (if they were parents of minor children). The questionnaire asked about the number of children in each family and the level of education that the LSAL participant (the parent of the minor children) expected each child to complete, their ability to finance the expected levels of post-secondary education, and the various sources of funding that they anticipated. This analysis utilises the 30-year record of the LSAL and some of the measures of the impact of the GR described in earlier analyses (Miller and Cepuran, in this issue) to examine the extent to which the GR shaped parental perceptions of their ability to finance all or part of their children’s post-secondary education.

Measures

Degree of financial responsibility to fund child’s post-secondary education

Parents were asked to report on their beliefs regarding how much financial responsibility they had for funding their children’s post-high school education:

Thinking about the post-high school years of schooling for these children, do you feel that you are: primarily responsible for providing financial support for your children’s education; partially responsible but each child should provide some support through work, loans or scholarships; not responsible for the costs of your children’s education; the costs of your children’s education will be covered by other sources.

(family, trusts, savings)

The last two response categories were collapsed together to create a trichotomous variable to represent parents’ belief regarding their financial responsibility for their children’s post-high school education (1 = parent is primarily responsible, 2 = parent is partially responsible, 3 = parent has no responsibility/covered by other sources). For the logistic regression analyses, this variable was collapsed into two categories (1 = parent is primarily responsible vs 0 = partial and/or none).

Plan for meeting financial contribution

Using three response options (0 = no, 1 = yes, 2 = not sure), parents reported on their plans for financing their children’s education beyond high school:

Thinking about this level of needed resources, do you think that you have adequate resources to provide this level of support; have a plan for savings that will provide this level of support; will be able to borrow to provide the needed support; will need for your child/children to win scholarships or grants; will need for your child/children to take loans to provide the needed funds; will get financial support from other family members for these expenses; do not have a plan to reach the level of support needed.

Each plan was collapsed into two categories (1 = yes, 0 = no/not sure) for analysis. A total of seven dichotomous variables were created for each funding plan.

Race/ethnicity

Parent’s race/ethnicity was categorised as African-American, Hispanic-American, or Caucasian (reference group) using three dichotomous variables.

Highest parent education

Parents reported on their level of education and their spouse’s level of education. A dichotomous composite variable representing the highest level of educational attainment between the two parents was created (1 = a four-year college degree or more).

Marital status

Respondents reported on their marital status and responses were coded dichotomously where 1 = married or with partner.

Work status

Respondents also reported on their work force status, which was coded dichotomously where 1 = working or self-employed.

Number of children expected to attain post-high school education

Parents reported on their educational expectations for each of their children. A count variable of the number of children expected to pursue education beyond high school was created.

Ever had student loan

In 2009, parents were asked whether they had ever taken out a student loan. Responses were coded dichotomously where 1 = yes, respondent had a student loan.

Perceived impact of the Great Recession

In the 2014 LSAL survey, parents were asked a series of questions about how the GR of 2008–09 affected their life. These yes/no questions were both positive and negative and asked respondents whether they lost their job due to the GR; experienced significant reduction in work hours; experienced difficulty paying rent or mortgage; lost health insurance or other job benefits; had difficulty saving for their child’s education; bought a house; went on vacation; or similar effects. A full description of the item wording and factor loadings are provided in Miller and Cepuran (this issue). A confirmatory factor analysis of 14 items produced factor loadings that were converted into a composite variable ranging from −10 (most negative impact) to +10 (most positive impact) and this composite index is included in the model.

Analysis

It is useful to begin with a descriptive summary of the key variables and relationships before turning to multivariate models. The two primary research questions examine the impact of the GR on (1) parents’ sense of responsibility for financing the post-secondary education of their children, and (2) parents’ plans to provide the needed level of funding.

Using a measure of the net impact of the GR on each participant in the LSAL (which ranges from −10 to +10), we group the level of GR impact into five ordinal categories to provide a summary view of the respondents’ perceived impact of the GR. There is a positive (but modest) association between the net impact of the GR and parents’ acceptance of primary responsibility for financing their children’s post-secondary education (see Table 1). The gamma for this association is 0.09, meaning that 9% of the variance in acceptance of primary responsibility can be attributed to the net impact of the GR on the LSAL participant. The association between the impact of the GR and parents’ perception of the adequacy of their resources was positive with a gamma of 0.37, indicating that LSAL parents who were not harmed by the GR but prospered were the most likely to think that their resources were adequate for financing the post-secondary education of their children.

Table 1:

LSAL parents’ willingness to accept responsibility for college financing, 2015

	Responsibility for financing			Adequate resources			N
	Primary	Partial	None/ Other	Adequate	Not sure	Not Adequate
All LSAL parents	33%	61%	5%	44%	25%	31%	1,410
Impact of the Great Recession
−10 to −7 (negative)	28	66	6	13	13	73	32
−6 to −4	37	59	4	15	27	58	83
−3 to +3	31	63	6	16	27	57	374
+4 to +6	29	63	9	30	23	47	283
+7 to +10 (positive)	37	60	4	43	26	31	638
Gamma =	0.09			0.37
Participant level of education
High school or less	32	62	6	20	26	44	551
Associate degree	23	64	13	36	16	48	129
Bachelor degree	36	60	4	32	28	40	481
Master’s	34	61	5	34	22	44	227
Doctorate or profess.	42	56	2	29	25	46	100
Gamma =	0.09			ns
Participant had student loan
Had student loan	28	67	5	29	26	45	731
Did not have loan	38	56	6	34	24	42	780
Gamma =	−0.18			ns
Race and ethnicity
African-American	37	61	2	30	24	46	118
Hispanic-American	37	51	12	36	24	40	107
Caucasian	31	63	6	31	25	44	1,136
Gamma =	NA			NA

Note: NA = Not Appropriate; ns = not significant at the 0.05 level; Gamma is a proportional-reduction-of-error (PRE) statistic for measuring the relationship between two ordinal variables. In terms of its PRE interpretation, a gamma is the ordinal equivalent of a Pearson’s R² (see Costner, 1965, for an extended discussion of gamma and other PRE statistics).

The level of education completed by LSAL participants was moderately related to parent acceptance of primary responsibility for financing their children’s education (gamma = 0.09) but was unrelated to their assessment of the adequacy of their resources for financing the post-secondary education of their children (see Table 1). Other analyses of the impact of the GR found that the level of educational attainment prior to the GR was strongly related to a positive net impact of the GR (Miller and Cepuran, this issue), and we would expect that parental acceptance of primary responsibility might be related to education in roughly the same way as the net impact of the GR itself. The absence of an association between the level of LSAL participant education and a sense of adequacy of their resources may reflect differential expectations about the level of education that their children will seek and the cost of that education. College-educated parents, for example, are more likely to encourage their children to attend college and seek graduate or professional education beyond the bachelor degree than are less-well-educated parents (Bengtson et al, 2002; Massey, 2007; Putnam, 2015).

Looking at bivariate relationships, the prior experience of parents (LSAL participants) with student loans appears to colour their sense of responsibility for their children’s future post-secondary education. Parents who had previously held student loans for their own education were more likely to say that they had only partial responsibility for financing the post-secondary education of their children’s post-secondary education (see Table 1). LSAL participants who did not have loans to finance their own education were more likely to accept primary responsibility for paying for the post-secondary education of their children.

Similar bivariate analyses found that neither the race nor ethnicity of LSAL participants were related to their perceptions of responsibility or their assessment of the adequacy of the resources for financing the post-secondary education of their children.

These bivariate relationships are interesting, but in life, all of these variables occur at the same time. It is necessary to utilise a multivariate model to assess the relative influence of each of these factors on the acceptance of responsibility and assessment of the adequacy of resources. For this purpose, we turn to a set of logistic regression models. For the first study question, parents’ belief about the degree of their financial contribution was regressed on all study variables of interest (race/ethnicity, highest parent education, marital and work statuses, number of children expected to attain post-secondary education, ever had student loan, perceived impact of the Great Recession). For the second study question, the same set of study variables were used to predict each of the seven plans parents had to meet their financial contribution. To reduce the possibility of Type I errors due to multiple comparisons, Holm’s correction (Holm, 1979) was applied to the final models. This sequential method, which makes an adaptive adjustment to each p-value, is less conservative than the Bonferroni correction (Levin, 1996; McLaughlin and Sainani, 2014; Chen et al, 2017). Finally, all analyses were estimated with probability weights so that results could be generalised to the population of interest, parents in the US who were in their mid-thirties during the GR.

Within the analytic sample, there was a low amount of missing data across variables, ranging from zero to 9% (see Table 2). Missing data were imputed using multiple imputation by chained equations procedures in Stata15 (StataCorp, 2017) to create 25 complete data sets. A set of auxiliary variables, not a part of the analysis, was used to help with the precision and efficiency of the imputation model and meet the missing at random (MAR) assumption underlying imputation procedures (Graham et al, 2007; Acock, 2012). The auxiliary variables included the total number of children parents had; the number of children not yet in post-secondary school; the parent’s average maths, reading and science achievement score; parent encouragement of academic and educational attainment by their children; and parent’s depressive symptomatology.

Table 2:

Weighted descriptive statistics of non-imputed predictor variables

Predictors	N	M or %	SD	Min	Max	% Missing
African-American	1,394	8.61%		0	1	0%
Hispanic-American	1,394	7.87%		0	1	0%
Caucasian	1,394	83.52%		0	1	0%
Parent has college degree	1,379	61.20%		0	1	1.08%
Married	1,393	84.12%		0	1	0.07%
Working	1,393	88.33%		0	1	0.07%
No. of children expected to attain education beyond high school	1,393	1.69	1.25	0	8	0.07%
Parent ever had student loans	1,264	44.82%		0	1	9.33%
Great Recession impact	1,305	4.88	4.82	−10	10	6.38%

Results

The sample was 84% Caucasian, 9% Black, and 8% Hispanic (see Table 2). More than half of the parents in the sample who reported responsibility for funding their children’s post-secondary education had at least a college degree (61%), were married (84%), and employed (88%). Parents expected between one and two of their children (M = 1.62, SD = 1.25) to attain an education beyond high school. Fewer than half of the parents had ever taken out a student loan (45%), and the average LSAL parent experienced little negative impact from the GR (M = 4.88, SD = 4.82).

Study question 1: Parent’s degree of financial responsibility

What are parents’ beliefs about their responsibility for financing their children’s education beyond high school? The majority of the parents (64%) believed that it is partially the parents’ responsibility to fund their children’s education whereas only a third (31%) of the parents believed it is all the parents’ responsibility. In contrast, only a small minority of families (5%) believed it is not the parents’ responsibility to fund their children’s education. Of note is that parents who believed it is all the responsibility of the parent to fund their children’s education reported more positive GR impact than parents who believed it is partially or none of the parents’ responsibility (see Appendix A).

The results from the logistic regression adjusted with the Holm’s correction indicate that three predictor variables – parents with at least a college degree, the number of children expected to get additional education beyond high school, and whether the parent ever had a student loan – significantly predict parents’ beliefs regarding their degree of financial responsibility for funding their children’s education (see Table 3). The odds of parents believing it is all (versus partially or none of) the responsibility of the parent to fund their children’s education beyond high school increases by 66% for parents with at least a college degree. In contrast, the odds of parents believing it is all (versus partially or none of) the responsibility of the parent to fund their children’s education decreases 18% with each additional child. The odds of believing it is all (versus partially or none of) the parent’s responsibility decreases by 40% for parents who have ever had a student loan. None of the other variables in the model, including the impact of the GR, were significant.

Table 3:

Predicting parents’ degree of financial responsibility using 25 multiply imputed data (N=1,394)

	All parents’ responsibility vs partial and/or none
	OR	SE	p-value	95% CI
African-American (vs Caucasian)	1.52	0.41	0.12	0.90	2.57
Hispanic-American (vs Caucasian)	1.41	0.35	0.17	0.87	2.28
Parent has college degree	1.66	0.28	0.00	1.19	2.30
Married	0.81	0.16	0.27	0.55	1.18
Working	1.33	0.31	0.22	0.84	2.10
No. children expected to attain education beyond high school	0.82	0.05	0.00	0.73	0.93
Parent ever had student loans	0.60	0.10	0.00	0.43	0.83
Great Recession impact	1.01	0.02	0.70	0.98	1.04
Constant	0.50	0.14	0.02	0.28	0.88

Notes: OR = odds ratio; SE = standard error; CI = confidence interval.

Bolded coefficients indicate what remains significant after the Holm’s correction (Holm, 1979).

Study question 2: Plans for meeting financial contribution

Adequate resources

Do parents have adequate resources to fund their children’s education? About 44% of parents reported that they do not have adequate resources to fund their children’s education while 32% reported that they do have adequate resources and 24% reported not being sure. Parents who reported having adequate resources reported more positive GR impact than parents who reported not being sure and not having adequate resources. Parents who reported not having adequate resources reported being the most negatively impacted by the GR (see Appendix B). Results from the logistic regression analyses indicate that each additional positive impact of the GR, increased parents’ odds of reporting adequate resources to fund their children’s education by 16% (see Table 4).

Table 4:

What are your plans and resources for meeting those needs/funding students’ higher education? (N=1,394)

DV: Adequate resources	Yes vs No/Not sure
	OR	SE	p- value	95% CI
African-American (vs Caucasian)	1.31	0.40	0.37	0.72	2.39
Hispanic-American (vs Caucasian)	1.61	0.45	0.08	0.94	2.77
Parent has college degree	1.09	0.20	0.64	0.76	1.55
Married	1.01	0.23	0.97	0.64	1.58
Working	0.61	0.14	0.03	0.39	0.96
No. children expected to attain education beyond high school	0.92	0.06	0.19	0.81	1.04
Parent ever had student loans	0.73	0.12	0.06	0.52	1.01
Great Recession impact	1.16	0.02	0.00	1.11	1.21
Constant	0.36	0.12	0.00	0.19	0.69

Notes: DV = dependent variable; OR = odds ratio; SE = standard error; CI = confidence interval.

Bolded coefficients indicate what remains significant after the Holm’s correction (Holm, 1979).

What are parents’ plans for funding their children’s higher education?

Having a plan for savings.

Some 47% of parents had a savings plan, 42% did not have one, and 11% were not sure (see Appendix C). Logistic regression results indicate that parents with a college degree had 95% higher odds of having a savings plan in comparison to parents without a college degree. Each additional positive impact of the GR increased parents’ odds of having a savings plan by 15% (see Table 5).

Table 5:

What are your plans and resources for meeting those needs/funding students’ higher education? (N=1,394)

DV: Have a plan for savings	Yes vs No/Not sure
	OR	SE	p- value	95% CI
African-American (vs Caucasian)	1.85	0.53	0.03	1.05	3.26
Hispanic-American (vs Caucasian)	1.56	0.43	0.11	0.90	2.69
Parent has college degree	1.95	0.33	0.00	1.40	2.71
Married	0.94	0.20	0.76	0.62	1.42
Working	0.99	0.21	0.97	0.65	1.51
No. children expected to attain education beyond high school	0.90	0.06	0.09	0.80	1.02
Parent ever had student loans	0.68	0.11	0.01	0.50	0.92
Great Recession impact	1.15	0.02	0.00	1.11	1.19
Constant	0.41	0.11	0.00	0.24	0.69
DV: Able to borrow	Yes vs No/Not sure
	OR	SE	p- value	95% CI
African-American (vs Caucasian)	1.12	0.32	0.70	0.64	1.97
Hispanic-American (vs Caucasian)	1.28	0.35	0.37	0.75	2.19
Parent has college degree	1.11	0.18	0.54	0.80	1.53
Married	1.31	0.26	0.18	0.88	1.93
Working	1.64	0.36	0.02	1.07	2.53
No. children expected to attain education beyond high school	1.05	0.07	0.47	0.92	1.21
Parent ever had student loans	1.05	0.16	0.73	0.78	1.43
Great Recession impact	1.09	0.02	0.00	1.05	1.12
Constant	0.29	0.08	0.00	0.16	0.51
DV: Need child to win grant or	Yes vs No/Not sure
scholarship	OR	SE	p-value	95% CI
African-American (vs Caucasian)	2.38	1.09	0.06	0.97	5.85
Hispanic-American (vs Caucasian)	1.20	0.36	0.55	0.66	2.16
Parent has college degree	0.72	0.13	0.08	0.50	1.03
Married	0.64	0.17	0.10	0.38	1.09
Working	1.44	0.37	0.16	0.87	2.39
No. children expected to attain education beyond high school	1.29	0.09	0.00	1.13	1.49
Parent ever had student loans	1.73	0.30	0.00	1.23	2.42
Great Recession impact	0.90	0.02	0.00	0.86	0.94
Constant	3.03	1.14	0.00	1.45	6.34
DV: Child needs to take out loans	Yes vs No/Not sure
	OR	SE	p- value	95% CI
African-American (vs Caucasian)	0.40	0.11	0.00	0.23	0.69
Hispanic-American (vs Caucasian)	0.65	0.18	0.11	0.38	1.11
Parent has college degree	0.86	0.14	0.36	0.62	1.19
Married	0.89	0.19	0.60	0.58	1.37
Working	1.52	0.34	0.06	0.98	2.36
No. children expected to attain education beyond high school	1.28	0.08	0.00	1.14	1.44
DV: Child needs to take out loans	Yes vs No/Not sure
	OR	SE	p-value	95% CI
Parent ever had student loans	2.44	0.38	0.00	1.79	3.32
Great Recession impact	0.89	0.02	0.00	0.86	0.92
Constant	0.83	0.25	0.53	0.45	1.51
DV: Financial support from other family members	Yes vs No/Not sure
	OR	SE	p-value	95% CI
African-American (vs Caucasian)	0.76	0.22	0.35	0.42	1.35
Hispanic-American (vs Caucasian)	1.37	0.42	0.31	0.75	2.50
Parent has college degree	1.31	0.27	0.19	0.88	1.94
Married	0.60	0.14	0.03	0.38	0.95
Working	0.78	0.19	0.30	0.49	1.25
No. children expected to attain education beyond high school	0.92	0.06	0.24	0.81	1.05
Parent ever had student loans	1.02	0.18	0.93	0.72	1.43
Great Recession impact	0.98	0.02	0.33	0.95	1.02
Constant	0.52	0.15	0.03	0.29	0.93
DV: Have plan	Yes vs No/Not sure
	OR	SE	p-value	95% CI
African-American (vs Caucasian)	1.49	0.49	0.23	0.78	2.83
Hispanic-American (vs Caucasian)	0.98	0.30	0.95	0.54	1.78
Parent has college degree	1.90	0.41	0.00	1.24	2.91
Married	1.13	0.28	0.62	0.70	1.83
Working	0.80	0.20	0.39	0.49	1.32
No. children expected to attain education beyond high school	1.07	0.09	0.41	0.91	1.25
Parent ever had student loans	0.79	0.15	0.22	0.53	1.16
Great Recession impact	1.04	0.02	0.06	1.00	1.08
Constant	2.76	0.93	0.00	1.43	5.35

Notes: DV = dependent variable; OR = odds ratio; SE = standard error; CI = confidence interval.

Bolded coefficients indicate what remains significant after the Holm’s correction (Holm, 1979).

Able to borrow.

About half (53%) of the parents reported being able to borrow money, while 26% said they were unsure, and 22% reporting that they would not be able to borrow money (see Appendix C). Parents’ odds of reporting that they would be able to borrow money increased 9% for each additional positive impact of the GR experienced by the parent (see Table 5).

Need child to win grant or scholarship.

The majority of parents (73%) indicated that their child would need to win a scholarship or grant. About 14% of parents said their child would not need to take out a loan while 13% of parents were unsure (see Appendix C). The odds of needing a child to win a grant or scholarship increased by 29% for each additional child expected to get an education beyond high school. Similarly, the odds were 73% higher for parents who had ever had a student loan. In contrast, each additional positive GR impact decreased the odds of parents needing the child to win a grant or scholarship by 10% (see Table 5).

Child needs to take out loans.

Roughly half of the parents (55%) reported needing their child to take out a loan. About 24% said that their child would not need to take out a loan while 22% were unsure (see Appendix C). In comparison to Caucasian parents, African-American parents had 60% lower odds of reporting that their child needs to take out a loan. In contrast, the odds of needing a child to take out a loan was 1.28 times higher for each additional child expected to attain education beyond high school and 2.44 times higher for parents who had ever had a loan. The odds of needing a child to take out a loan decreased by 11% for each additional positive impact of the GR experienced by the parent (see Table 5).

Financial support from other family members.

Most parents (63%) did not expect financial support from other family members to help fund their children’s education beyond high school. In contrast, 22% said that they would have financial support from family members and 15% were unsure (see Appendix C). None of the variables in the model significantly predicted access to financial support from other family members (see Table 5).

No plan.

The majority of parents reported having a plan to fund their children’s education (64%) while 18% said they did not have a plan and 18% said they were unsure if they had a plan (see Appendix C). The odds of having a plan increased by 90% for parents with a college degree in comparison to parents without one. The GR did not have a significant impact on whether parents had a plan (see Table 5).

Gender of parent.

Although prior research has not found any gender differences in parents’ financial contributions towards their children’s higher education (Flint, 1997), we ran t-tests to check whether this null finding was replicated in our sample. We found only one significant difference; compared to fathers, mothers had higher odds of reporting that their children needed to win a grant or scholarship. For the sake of brevity, we did not include parent gender in our models. Future research may examine whether this association is replicated in other samples, and if so, investigate why mothers are more likely to report that their children need to win a grant or scholarship.

Discussion

The present study extended prior research on college financing in the US using contemporary, nationally representative data. Specifically, this study examined the extent to which parent and family characteristics were associated with parents’ beliefs and plans around financing higher education in the context of the GR. Results demonstrated that although parents’ beliefs about their responsibility to fund their children’s post-secondary education were impervious to the effects of the GR, their plans to do so were strongly linked to their experience with the GR. Conversely, parent and family characteristics, such as educational attainment, number of children expected to pursue higher education, and parents’ experiences funding their own post-secondary education, were significantly associated with both parental beliefs and financial plans.

Impact of the Great Recession

Study results indicate that the GR was not associated with parents’ beliefs of their responsibility to finance their children’s college education. Instead, parents’ beliefs about financial responsibility were associated with parent and family characteristics. Parents with higher levels of education were more likely to accept all responsibility for funding their children’s post-secondary education while parents who had borrowed loans for their own education and who expected more of their children to pursue post-secondary education were less likely to accept all financial responsibility. These results replicate findings found in earlier cohorts demonstrating the link between parents’ acceptance of primary financial responsibility and parents’ socio-economic background and own financing experiences (Steelman and Powell, 1991). These results also provide empirical support for the relevance of the resource dilution theory to parents’ beliefs about their level of responsibility for financing their children’s post-secondary education.

In contrast, parents’ perceived impact of the GR was consistently associated with their plans for financing their children’s post-secondary education. Parents who experienced a more positive impact of the GR were more likely to report having adequate resources and the ability to pay for their children’s post-secondary education. Specifically, parents who experienced fewer negative consequences from the GR were more likely to have a savings plan in place and were more likely to report that they could borrow funds to pay for their children’s education if needed. In contrast, parents who suffered substantial negative consequences from the GR were more likely to report that their children would need to win some scholarship assistance or take out loans to help finance post-secondary education costs.

Parent and family characteristics

Several parent and family characteristics were significantly associated with their plans for financing their children’s post-secondary education. In alignment with results based on older parent cohorts (Steelman and Powell, 1991, for example), parents with higher levels of education reported having a savings plan and having a plan in general to fund their children’s college education. Indeed, prior research indicates that although parents tend to misjudge the actual cost of higher education, parents from higher socio-economic backgrounds tend to have a better understanding of college costs (Horn et al, 2003). As a result, of having greater knowledge of costs, it is likely that they would put a greater emphasis on having a plan in place to finance their children’s post-secondary education.

Similar to results from a study of parents in the 1980s and 1990s (Steelman and Powell, 1991), parents who had taken out educational loans for themselves were more likely than parents who never had educational loans to report that their children would need to contribute financially to their own education through grants, scholarships and loans. Moreover, compared to the impact of the GR, parents’ own loan experiences were more strongly associated with parents’ expectations that children contribute financially to their post-secondary education through grants or loans. This finding suggests that parents’ expectations and values regarding how families should finance children’s post-secondary education are transferred intergenerationally, and that this transmission occurs across different economic contexts including an economic recession.

Parental expectations for children to contribute financially to their schooling were associated with the number of children expected to pursue post-secondary education as well. Indeed, the present findings support the resource dilution hypothesis and align with prior research indicating that children with many siblings received less financial support for college from their parents (Steelman and Powell, 1989; Downey, 1995; Powell and Steelman, 1995). Parents in our study were more likely to report that children needed to get grants, scholarships or loans as the number of children expected to attain higher education increased. Moreover, the number of children expected to pursue a post-secondary education was more strongly associated with the need for children to contribute financially to their education than the impact of the GR, which suggests that the resource dilution theory is robust across historical contexts. Taken together, these results indicate that parent and family characteristics were more strongly correlated with parents’ plans for financing their children’s education than the impact of the GR.

In comparison to parents’ level of education, experience with student loans, and number of children expected to pursue education beyond secondary school, none of the other parent and family characteristics were associated with beliefs and plans to finance their children’s education. One exception was that African-American parents were less likely than Caucasian parents to report that their children would need to take out loans even after taking other parent and family characteristics into account. This finding corroborates prior research demonstrating that Caucasian parents are more likely than racial/ethnic-minority parents to report that children should be responsible for funding their college education (Steelman and Powell, 1993). It is not possible to determine the reason for this racial/ethnic difference in the present study, but this is one potential area for exploration in future research. Similarly, none of the variables in our analyses were associated with parents’ plans to acquire financial support from other family members even though over 20% of parents reported it as one of their plans for financing their children’s college education. Given the importance of extended family among certain groups (such as immigrant, low-income, ethnic minority), future research should investigate what factors are associated with parents’ acceptance of extended family as a source of financial support for their children’s post-secondary education.

Study limitations

These findings should be considered in light of the limitations of this study. First, all measures are based on parents’ self-report and perceptions rather than objective measures of the impact of the GR on their finances and their ability to fund their children’s post-secondary education, which potentially may have biased results. However, objective measures of income are not synonymous with beliefs of income adequacy (Hazelrigg and Hardy, 1997; Mirowsky and Ross, 1999). Extant literature indicates that subjective measures of financial status are better predictors of satisfaction of overall quality of life (Ackerman and Paolucci, 1983).Thus, it may be the case that parents’ perceived responsibility and ability to fund their children’s post-secondary education is more important than objective measures of how much the GR impacted their financial situation for explaining their financial behaviours. This should be empirically tested in future research.

Another limitation of the present study is that we were unable to examine whether parents’ perceived responsibility and funding plans differed by their child’s gender. It was not possible to examine this question with the LSAL data because parents were not asked to report on their funding beliefs and plans for each child, but this could be a line of research that is pursued with alternative data. Currently, there is conflicting evidence as to whether parents socialise their children differently based on their gender and whether those differences affect their children’s outcomes. There is some evidence that parents may parent differently based on gender role stereotypes (Eccles et al, 1990), but a meta-analysis of studies based on parents from Western countries indicates that differences in parents’ differential socialisation of boys and girls are generally small and have non-significant effects on children’s outcomes (Lytton and Romney, 1991). It would be informative to know whether these socialisation differences extended to parents’ plans for funding their children’s post-secondary education and if, in turn, this leads to differences in pursuing post-secondary education. At least one study found that the child’s gender does not have any bearing on parental beliefs about their level of responsibility or savings plans for financing their children’s education (Steelman and Powell, 1991).

Finally, the processes underlying the association between parents’ experience in the GR and plans to finance their children’s post-secondary education are unclear. Future research should investigate the mechanisms that explain how parents’ perceptions of the impact of the GR influence their plans to fund their children’s post-secondary education.

Despite these limitations, the present study contributes to this area of research in several important ways. Most notably, this study is one of the few empirical studies to highlight the relative importance of parent and family characteristics for understanding how parents approach financing their children’s post-secondary education. In addition, this study uses contemporary, nationally representative data to replicate results based on older cohorts of parents and to expand our understanding of parents’ approach to financing post-secondary education beyond having a savings plan to include borrowing and relying on financial contributions from children and extended family. Finally, although this study is based on parents in the US, many of the issues discussed in this analysis are relevant to European and other nations who are transitioning from systems of publicly financed higher education to mixed systems that include parent and family financing at least in part. There is a small but growing literature on the impact of tuition systems and parent and family obligations to participate in funding a portion of the costs of higher education in European countries and other nations (Greenaway and Haynes, 2003; Heller and Rogers, 2006; Hübner, 2012; Bruckmeier and Wigger, 2014; Dearden et al, 2014; Denny, 2014; Goksu and Goksu, 2015; Marginson, 2018; Murphy et al, 2018). Future research should examine whether the associations found in the present analyses are replicated for parents living in other countries with different educational contexts and funding expectations.

Appendices

Appendix A: Non-imputed, weighted descriptive statistics of predictor variables by parents’ degree of financial responsibility

	Parent’s degree of financial responsibility
	All parents’ responsibility					Partially parents’ responsibility					Not parents’ responsibility
	N	M or %	SD	Min	Max	N	M or %	SD	Min	Max	N	M or %	SD	Min	Max
African-American	431	10.18%		0	1	889	8.43%		0	1	69	2.70%		0	1
Hispanic-American	431	9.08%		0	1	889	6.38%		0	1	69	17.43%		0	1
Caucasian	431	80.74%		0	1	889	85.19%		0	1	69	79.87%		0	1
Parent has college degree	427	64.09%		0	1	881	59.52%		0	1	68	43.16%		0	1
Married	431	81.75%		0	1	888	86.09%		0	1	69	74.99%		0	1
Working	431	90.55%		0	1	888	87.89%		0	1	69	81.46%		0	1
No. children expected to attain education beyond high school	431	1.49	1.08	0	4	888	1.82	1.32	0	8	69	1.49	1.25	0	4
Parent ever had student loans	388	37.47%		0	1	809	48.83%		0	1	64	40.66%		0	1
Great Recession impact	405	4.91	5.08	−9	10	833	4.59	4.86	−10	10	64	4.13	4.38	−8	10
Full analytic sample
N	1,389
%	31.03%					64%					4.97%
% Missing	0.36%

Appendix B: Non-imputed, weighted descriptive statistics of predictor variables by adequate resources

	Parents have adequate resources to meet financial contribution
	Yes					No					Not Sure
	N	M or %	SD	Min	Max	N	M or %	SD	Min	Max	N	M or %	SD	Min	Max
Black	434	8.08%		0	1	593	9.08%		0	1	327	7.96%		0	1
Hispanic	434	8.91%		0	1	593	7.07%		0	1	327	7.55%		0	1
White	434	83.02%		0	1	593	83.85%		0	1	327	84.49%		0	1
Parent has college degree	430	66.07%		0	1	588	56.93%		0	1	323	60.78%		0	1
Married	434	87.56%		0	1	592	81.58%		0	1	327	84.69%		0	1
Working	434	87.15%		0	1	592	88.52%		0	1	327	89.74%		0	1
No. children expected to attain education beyond high school	434	1.66	1.12	0	6	592	1.84	1.34	0	7	327	1.51	1.22	0	8
Parent ever had student loans	393	38.80%		0	1	538	47.35%		0	1	298	47.24%		0	1
Great Recession impact	403	6.73	3.94	−9	10	555	3.32	5.06	−10	10	310	4.55	4.86	−9	10
Full analytic sample
N	1,354
% of Sample	32.05%					43.80%					24.15%
% Missing	2.87%

Appendix C: Weighted descriptive statistics of non-imputed plans to meeting financial contribution and demographics

Study question 2: dependent variables	N	M or %	% Missing
Plan to meeting financial contribution
Have a plan for savings	1,356		2.73%
No		41.74%
Yes		47.42%
Not sure		10.84%
Able to borrow	1,336		4.16%
No		21.63%
Yes		52.69%
Not sure		25.67%
Need child to win scholarship/grant	1,350		3.16%
No		14.15%
Yes		72.81%
Not sure		13.04%
Child needs to take out loans	1,342		3.73%
No		23.55%
Yes		54.92%
Not sure		21.54%
Financial support from other family members	1,332		4.45%
No		62.76%
Yes		21.77%
Not sure		15.47%
Have plan	1,297		6.96%
No		17.96%
Yes		64.30%
Not sure		17.73%