Common use of Survey Description Clause in Contracts

Survey Description. I designed the survey to find evidence of differences in beliefs and future expectations for ISA participants and non-participants. Each ISA applicant received a survey invitation email which explained that only Purdue University students were being invited and that I was conducting the survey to learn about how their “experiences, attitudes, expectations, and beliefs influence how [they] pay for college.” I did not explain that I was specifically studying the ISA program as I did not want students to be thinking about the ISA program when answering the questions. All students who completed the survey were immediately given a $20 Amazon gift card code. Approximately 60 percent of ISA applicants chose to complete the survey. To account for potential non-response bias, I use propensity score 10A large fraction of students go directly to graduate school, some take internships, and others report that they are still seeking full-time employment. weighting to weight the data by the inverse probability of responding. Very few observables affect the likelihood of responding to the survey and the weighting method does not change the results. After answering a few standard demographic questions, the students were asked to rate a set of 12 statements about debt aversion. ▇▇▇▇▇▇▇, ▇▇▇▇▇▇▇▇ and ▇▇▇▇▇▇▇▇ (2011) show that framing a financial contract as a debt or as a loan affects the student’s reported willingness to enter into the contract. My hypothesis is that students with greater aversion to debt will be more likely to choose to participate in the ISA than students with less debt aversion. The debt aversion questions used in this survey were developed and tested for reliability by ▇▇▇▇▇▇ and ▇▇▇ (1995). Students were asked to report that they either strongly agree, somewhat agree, neither agree or disagree, somewhat disagree, or strongly disagree with each statement. Of the 12 statements, 6 are “pro” statements and 6 are “anti” statements. Values of 5 through 1 are assigned to the responses for ease of reporting with the scale reversed for “anti” statements. Therefore, for all statements, higher values indicate greater aversion to debt. The questions are reported in Table 4 along with the mean score (on a 5 point scale) for each question for participants and non-participants. The final column reports the p-value for a t-test of the equality of means. As shown in Table 4, there is no evidence that ISA participants have greater debt aversion than ISA non-participants. In fact, question 1 is the only question in which there is a statistically significant difference between ISA participants and non-participants, and surprisingly suggests non-participants have greater debt aversion. Perhaps it is not surprising that the students in this study have similar views on debt as they all have federal student loans (if they did not, they would not have been eligible for an ISA through this program). Debt aversion seems to have little impact on ISA program participation among this student population. Financial experience and sophistication may impact which students decide to participate in the ISA program. To look for evidence of this, I ask survey participants to report their experience with checking and savings accounts, the stock market, car loans, and credit cards. Students were also asked a question to asses their knowledge about the power of compound interest that was introduced by ▇▇▇▇▇▇▇ and ▇▇▇▇▇▇ (2015): Suppose you owe $1,000 on your credit card and the interest rate you are charged is 20% per year compounded annually. If you did not pay anything off, at this interest rate, how many years would it take for the amount you owe to double? The correct answer is about 3.6 years. Students who understand simple interest but ignore or do not understand interest compounding would arrive at an answer of 4 to 6 years. Selecting an answer of less than 2 years or more than 6 years demonstrates a misunderstanding of how interest accrues. As reported in Table 5, about 40 percent of the students selected the correct answer of 2 to 4 years and a majority of the students with an incorrect answer selected 4 to 6 years. These are higher percentages than ▇▇▇▇▇▇▇ and ▇▇▇▇▇▇ (2015) found in a nationally representative survey of adults. The largest difference is that ISA participants are 9 percentage points more likely to answer that they “do not know” how long it would take for the amount to double. ISA participants and non-participants have similar financial experience in terms of types of accounts. Participants are 8 percentage points less likely to have a credit card and have .18 fewer cards than non-participants on average. There are no statistically significant differences in employment experience. The ISA should be more attractive to students who expect to have lower salaries after graduation and less labor force participation. One important source of differences in labor force participation is the timing of marriage and children. Given the gender difference in how children affect labor force participation, I would expect that women who anticipate having more children or having children earlier would find the ISA more attractive than men with similar family expectations. Survey participants are asked to report their “best guess” of what their annual salary would be if they were to accept a full-time job soon after graduation. Betts (1996) found that fourth-year student knowledge of salaries in their own field were quite accurate, and the survey responses are consistent with this. Questions which ask the student to report a percentage chance are elicited by moving a slider between 0 and 100. Table 6 suggests that ISA participants have lower starting salary expectations than non- participants. This is as expected, though the difference is only about $3,000. The entire distribution of starting salaries seems to be lower for ISA participants. The difference in expected salaries shrinks over time as the non-participants expect a slightly lower rate of salary growth, though this difference is not statistically significant. There is no difference in expected labor force participation between participants and non-participants. For both men and women, it is the non-participants who believe they are more likely to get married. Also surprising is that for women, it is again the non-participants who expect to have children sooner and have more children. Neither of these findings are consistent with the hypothesis that the ISA should be more attractive to those who expect lower earnings because of family responsibilities. Students who expect to live in cities with a high cost of living and higher wages after graduation should be less interested in participating in the ISA. The survey presents students with a list of 12 cities presented in a random rank order. Students asked to imagine that they receive a job offer from a company with locations in each of the 12 cities and the company asks the student to rank the cities by where they would most like to work to where they would least like to work. Students are told that the salary does not depend on the location assigned. Students can drag the cities up or down the rank order list to reorder them. A rank of 1 is the most preferred location and a rank of 12 is the least preferred. Table 7 presents the average rank for each city for participants and non-participants. For only 1 out of the 12 cities is there a statistically significant difference between the average ranking for participants and non-participants. It seems most likely that this is just due to chance rather than something special about Pittsburgh, though I construct a number of location characteristic measures to look for other differences. The 12 cities were selected to be able to test for the importance of specific characteristics. Those who only ranked high-population metro areas (Chicago, Washington DC, San ▇▇▇▇- cisco, Boston, and Phoenix) in their top 3 are defined as have a large city preference. Those who only ranked low population metro areas (Peoria, Fort ▇▇▇▇▇, Evansville, Topeka, and Terre Haute) in their top 3 are defined as having a small city preference. Students who prefer Indianapolis to Pittsburgh, Fort ▇▇▇▇▇ to Peoria, and Terre Haute to Topeka (regardless of where each of these cities appears in the rankings) are defined as have an Indiana preference. Students who have the opposite ranking for all three of those city pairs are defined as hav- ing an outside Indiana preference. An eastern preference is defined as ranking Boston and Washington DC above both San Francisco and Phoenix. The opposite ranking is defined as a western preference. Finally, to directly test if higher salaries affect ISA participation, I average the median household income for the three highest ranked metro areas for each student and report the average for participants and non-participants separately. Across all these location preference variables, I find no statistically significant evidence for differential selection into the ISA. Participating in an ISA is also a form of insurance against low earnings and therefore students with higher risk aversion may be more likely to participate. The survey contains a single two-part question designed to elicit the magnitude of risk aversion:11 Suppose that a distant relative left you a share in a private business worth one hundred thousand dollars. You are immediately faced with a choice: (a) cash out now and take the $100,000 or (b) wait until the company goes public in one month which would give you a 50 percent chance of doubling your money to $200,000, and a 50 percent chance of losing one-third of it, leaving you with $66,000. (a) pay 12 monthly installments of $100 each (b) pay $1,200 one year from now Which would you choose?

Survey Description. I designed the survey to find evidence of differences in beliefs and future expectations for ISA participants and non-participants. Each ISA applicant received a survey invitation email which explained that only Purdue University students were being invited and that I was conducting the survey to learn about how their “experiences, attitudes, expectations, and beliefs influence how [they] pay for college.” I did not explain that I was specifically studying the ISA program as I did not want students to be thinking about the ISA program when answering the questions. All students who completed the survey were immediately given a $20 Amazon gift card code. Approximately 60 percent of ISA applicants chose to complete the survey. To account for potential non-response bias, I use propensity score 10A large fraction of students go directly to graduate school, some take internships, and others report that they are still seeking full-time employment. weighting to weight the data by the inverse probability of responding. Very few observables affect the likelihood of responding to the survey and the weighting method does not change the results. After answering a few standard demographic questions, the students were asked to rate a set of 12 statements about debt aversion. ▇▇▇▇▇▇▇, ▇▇▇▇▇▇▇▇ and ▇▇▇▇▇▇▇▇ (2011) show that framing a financial contract as a debt or as a loan affects the student’s reported willingness to enter into the contract. My hypothesis is that students with greater aversion to debt will be more likely to choose to participate in the ISA than students with less debt aversion. The debt aversion questions used in this survey were developed and tested for reliability by ▇▇▇▇▇▇ and ▇▇▇ (1995). Students were asked to report that they either strongly agree, somewhat agree, neither agree or disagree, somewhat disagree, or strongly disagree with each statement. Of the 12 statements, 6 are “pro” statements and 6 are “anti” statements. Values of 5 through 1 are assigned to the responses for ease of reporting with the scale reversed for “anti” statements. Therefore, for all statements, higher values indicate greater aversion to debt. The questions are reported in Table 4 along with the mean score (on a 5 point scale) for each question for participants and non-participants. The final column reports the p-value for a t-test of the equality of means. As shown in Table 4, there is no evidence that ISA participants have greater debt aversion than ISA non-participants. In fact, question 1 is the only question in which there is a statistically significant difference between ISA participants and non-participants, and surprisingly suggests non-participants have greater debt aversion. Perhaps it is not surprising that the students in this study have similar views on debt as they all have federal student loans (if they did not, they would not have been eligible for an ISA through this program). Debt aversion seems to have little impact on ISA program participation among this student population. Financial experience and sophistication may impact which students decide to participate in the ISA program. To look for evidence of this, I ask survey participants to report their experience with checking and savings accounts, the stock market, car loans, and credit cards. Students were also asked a question to asses their knowledge about the power of compound interest that was introduced by ▇▇▇▇▇▇▇ and ▇▇▇▇▇▇ (2015): Suppose you owe $1,000 on your credit card and the interest rate you are charged is 20% per year compounded annually. If you did not pay anything off, at this interest rate, how many years would it take for the amount you owe to double? The correct answer is about 3.6 years. Students who understand simple interest but ignore or do not understand interest compounding would arrive at an answer of 4 to 6 years. Selecting an answer of less than 2 years or more than 6 years demonstrates a misunderstanding of how interest accrues. As reported in Table 5, about 40 percent of the students selected the correct answer of 2 to 4 years and a majority of the students with an incorrect answer selected 4 to 6 years. These are higher percentages than ▇▇▇▇▇▇▇ and ▇▇▇▇▇▇ (2015) found in a nationally representative survey of adults. The largest difference is that ISA participants are 9 percentage points more likely to answer that they “do not know” how long it would take for the amount to double. ISA participants and non-participants have similar financial experience in terms of types of accounts. Participants are 8 percentage points less likely to have a credit card and have .18 fewer cards than non-participants on average. There are no statistically significant differences in employment experience. The ISA should be more attractive to students who expect to have lower salaries after graduation and less labor force participation. One important source of differences in labor force participation is the timing of marriage and children. Given the gender difference in how children affect labor force participation, I would expect that women who anticipate having more children or having children earlier would find the ISA more attractive than men with similar family expectations. Survey participants are asked to report their “best guess” of what their annual salary would be if they were to accept a full-time job soon after graduation. Betts ▇▇▇▇▇ (1996) found that fourth-year student knowledge of salaries in their own field were quite accurate, and the survey responses are consistent with this. Questions which ask the student to report a percentage chance are elicited by moving a slider between 0 and 100. Table 6 suggests that ISA participants have lower starting salary expectations than non- participants. This is as expected, though the difference is only about $3,000. The entire distribution of starting salaries seems to be lower for ISA participants. The difference in expected salaries shrinks over time as the non-participants expect a slightly lower rate of salary growth, though this difference is not statistically significant. There is no difference in expected labor force participation between participants and non-participants. For both men and women, it is the non-participants who believe they are more likely to get married. Also surprising is that for women, it is again the non-participants who expect to have children sooner and have more children. Neither of these findings are consistent with the hypothesis that the ISA should be more attractive to those who expect lower earnings because of family responsibilities. Students who expect to live in cities with a high cost of living and higher wages after graduation should be less interested in participating in the ISA. The survey presents students with a list of 12 cities presented in a random rank order. Students asked to imagine that they receive a job offer from a company with locations in each of the 12 cities and the company asks the student to rank the cities by where they would most like to work to where they would least like to work. Students are told that the salary does not depend on the location assigned. Students can drag the cities up or down the rank order list to reorder them. A rank of 1 is the most preferred location and a rank of 12 is the least preferred. Table 7 presents the average rank for each city for participants and non-participants. For only 1 out of the 12 cities is there a statistically significant difference between the average ranking for participants and non-participants. It seems most likely that this is just due to chance rather than something special about Pittsburgh, though I construct a number of location characteristic measures to look for other differences. The 12 cities were selected to be able to test for the importance of specific characteristics. Those who only ranked high-population metro areas (Chicago, Washington DC, San ▇▇▇▇- cisco, Boston, and Phoenix) in their top 3 are defined as have a large city preference. Those who only ranked low population metro areas (Peoria, Fort ▇▇▇▇▇, Evansville, Topeka, and Terre Haute) in their top 3 are defined as having a small city preference. Students who prefer Indianapolis to Pittsburgh, Fort ▇▇▇▇▇ to Peoria, and Terre Haute to Topeka (regardless of where each of these cities appears in the rankings) are defined as have an Indiana preference. Students who have the opposite ranking for all three of those city pairs are defined as hav- ing an outside Indiana preference. An eastern preference is defined as ranking Boston and Washington DC above both San Francisco and Phoenix. The opposite ranking is defined as a western preference. Finally, to directly test if higher salaries affect ISA participation, I average the median household income for the three highest ranked metro areas for each student and report the average for participants and non-participants separately. Across all these location preference variables, I find no statistically significant evidence for differential selection into the ISA. Participating in an ISA is also a form of insurance against low earnings and therefore students with higher risk aversion may be more likely to participate. The survey contains a single two-part question designed to elicit the magnitude of risk aversion:11 Suppose that a distant relative left you a share in a private business worth one hundred thousand dollars. You are immediately faced with a choice: (a) cash out now and take the $100,000 or (b) wait until the company goes public in one month which would give you a 50 percent chance of doubling your money to $200,000, and a 50 percent chance of losing one-third of it, leaving you with $66,000. (a) pay 12 monthly installments of $100 each (b) pay $1,200 one year from now Which would you choose?