Abstract
This study analyzed the impact of school nurses on economic efficiency. Public schools must provide resources such as books, teachers, and buses necessary to accommodate all enrolled students even if all students do not come to school every day and even if some leave school before the end of the school day. To the extent that resources are allocated but unused, there will be an increase in what is termed X-inefficiency. To investigate this phenomenon, data related to student absences and student checkouts were collected from 22 schools with a total student population just over 10,000 students during a 20-day period in the fall of 2001. This study found that fewer children checked out of school for medical reasons when a full-time school nurse was available at the school. Additionally, this study supports the contention that full-time school nurses might have a positive impact on the percentage of children with health care insurance coverage. However, a large number of students did check out for what appeared to be less than valid reasons. Thus, although the presence of a school nurse reduced the X-inefficiency, school policy and procedures allowed a degree of X-inefficiency to remain.
INTRODUCTION
Government programs and agencies exist to provide services that are expected to achieve designated social purposes. As the scope and costs of operations at all levels of government continue to grow, the public is becoming more insistent that public programs and agencies be evaluated in terms of how well they are achieving their missions and objectives. The public often questions whether the government is efficient. In response, public policy-makers and administrators are attempting to improve the efficiency and effectiveness of government. In the arena of public economics, Joseph Stiglitz (2000) has captured the motives behind these attempts in three central questions: “Does the government do too much? Does it do well what it attempts to do? Could it perform its economic role more efficiently?” (p. 4). In Alabama, efforts to ensure that state agencies and programs achieve the expected levels of efficiency and effectiveness were most recently seen in 1999, when the legislature included performance measures in the state education budget (Act 2000–594).
MISSIONS AND PERFORMANCE MEASURES FOR ALABAMA SCHOOL PROGRAMS
Performance-based budgeting for the public schools in Alabama means that measures of performance for the various school programs must be related to the missions of those programs, which in turn are expected to achieve the overall objectives of the state’s public education system. In Alabama, a Circuit Court has ruled public education a right and a constitutional function of government. Additionally, this court has ruled that the education provided must be equitable and adequate ( Alabama Coalition for Equity v. Guy Hunt, 1993). The Alabama legislature, therefore, must provide the resources necessary for an adequate education for all students. The debate about what constitutes an adequate education continues. But in simple terms, the meaning of adequacy would include providing an adequate number of current and usable textbooks for each student, sufficient teachers to maintain desired class size ratios, and a curriculum that is challenging and valid ( Alabama Coalition for Equity v. Guy Hunt, 1993). In addition, schools must also transport all students who need transportation and must provide counseling programs, library resources, and other supportive services. Under certain conditions, schools also must provide lunch and breakfast programs.
For the Alabama legislature, the allocation of fiscal resources for education should be based upon providing an adequate education for all students. Because certain parents choose to send their children to private, parochial, or other schools, all students may be defined as all students who are enrolled in public schools. Alabama’s Foundation Program requires the state legislature to allocate the resources such as teachers, buses, books, and supplies needed to provide for an adequate education for all students (Act 95–314). In essence, they must provide for all students, even if all students do not come to school every day. Given the general parameters of a production function model, which includes students attending school as the primary input, the only way for efficiency to be maintained is to have as many students come to school as are enrolled and to have them remain in school all day. To the extent that students either fail to come to school or leave school before the end of the school day, efficiency is decreased.
The School Nurse Program in the Education Production Function
In 1998, the Alabama legislature passed a law that required the employment of at least one registered nurse in every school system (Act 98–672). In addition, the law provided a phase-in provision that required local school boards to hire additional registered nurses each year until the school system ratio of nurses to students was 1:2,000 in 2010. However, there were no performance measures for the school nurse included in this legislation. The performance of these nurses did not drive the funding for the school nurse program, nor did it appear that the legislature had any clear performance goal in mind.
In 1999, the Alabama legislature added performance measures to its funding legislation for many state agencies and programs. The funding for school nurses was accompanied by two performance indicators: (a) the number of state-funded nurses in public schools will increase from approximately 128 to 140, and (b) the number of incidences in which nonqualified school personnel administer medication or first-aid treatment will be reduced (Act 2000–594). These two indicators illustrated the problems encountered when implementing performance-based budgeting. The first indicator did not relate to performance but merely to providing the service, in this case nurses. The second indicator would be problematic to measure and presupposes that some reduction was desired from a known level. Because this type of data has not been collected in Alabama, it would be impossible to determine if there were a reduction. Further, even if the data were available and showed that there were 3,000 incidents of nonqualified school personnel administering medication or first-aid treatment this year and this number were decreased to 2,500 next year, is it reasonable to assume there was more efficiency?
Hypotheses
For this study, the following hypotheses were evaluated:
Schools with a full-time school nurse will have higher average daily attendance than schools without a full-time nurse.
Schools with a full-time school nurse will have higher all-day attendance than schools without a full-time nurse.
Schools with a full-time school nurse will have a lower percentage of students checking out for medical reasons than schools without a full-time nurse.
Schools with a full-time school nurse will have a higher percentage of students enrolled in health care insurance coverage than schools without a full-time nurse.
Definition of Terms
The following terms were operationally defined for this study:
All-day attendance: The number of students who are present in a given school from the beginning of the day through the end of the school day.
Average daily attendance (ADA): The calculation of attendance based on the number of students who are present in schools during the initial morning roll call.
Average daily membership (ADM): The average number of students enrolled during the first 40 days.
Medical checkouts: The number of students who check out of school before the end of the day because of medical reasons such as not feeling well, illness, or injury.
Performance measurement: A method of measuring an organization’s progress in achieving predetermined goals that have been developed based on the mission of the organization. Performance measurement includes identifying the mission, goals, outcome indicators, activities, inputs, outputs, and efficiency for an organization (Walters, 1998).
Efficiency: Defined as the cost of resources (inputs) required to produce one unit of output (Walters, 1998).
Effectiveness: Defined as comparing output or outcomes to expectations (Hedley, 1998). In other words, did the organization accomplish what it intended to accomplish?
Otis-Lennon School Ability Test (OLSAT): A measure of school ability. The national grade percentile rank and stanine indicate how well a student performed on the OLSAT compared with students in the norming group who were in the same grade.
MEASURING PERFORMANCE OF PUBLIC ORGANIZATIONS
In the literature of public economics, it is recognized that applying the economic concept of allocative efficiency to evaluations of public agencies and programs is hampered because of the nature of the intended and actual outcomes of the production functions. First, the benefits are often highly intangible, so the quality and quantity are virtually impossible to ascertain. Second, because most public organizations and programs are intended to produce services that are not sold at prices, and moreover, because these services are not comparable to services produced and sold in private markets, it is virtually impossible to place monetary values on the outcomes (Hyman, 1999).
In a more pragmatic approach to applying the economic concepts of efficiency to public agencies and programs, the term efficiency in public administration literature usually corresponds to the economic concept of technical efficiency or productivity (Leathers, 1979). It refers to the ratio of output to input. For example, an educational program with a given number of teachers, administrators, and facilities achieves maximum efficiency when the greatest amount of education is gained.
There have always been conflicts between the goals of efficient and responsive government (White, 1999). The call for government reform may be more strident in public education than in any other policy area. In recent years many politicians have been elected using education as their number-one policy issue. Whereas the public is demanding reform, it is also demanding more accountability, especially from its public schools (McLaughlin & Wand, 1995; Walters, 1998). Parents and the general public want safe schools, high standards, and higher test scores. However, for the most part, the public is not interested in increasing taxes for education until they can be assured of greater accountability. Education and other governmental agencies have not been as active in developing meaningful and effective measures of performance as the private sector (Kaplan & Norton, 2001b. The response to this call for accountability both in general and within the education community has been a tendency to measure what is easy to measure (Lefkowith, 2001).
X-inefficiency in Public Organizations
In 1966, Harvey Leibenstein introduced the concept of X-inefficiency as another form of nonallocative inefficiency (Leibenstein, 1966). Until then, economists had assumed that firms always minimized costs. Leibenstein challenged this assumption by postulating that suboptimal decisions and less-than-rational decisions would affect the traditional models. In simple terms, such decisions could cause a firm to produce with higher costs than necessary or to produce less output than it is capable of producing (Babilot, Frantz, & Green, 1987). The estimation of X-inefficiency is based on comparing the actual output with the maximum possible output for a given set of inputs (Vitaliano, 1997). From this it is clear that if inputs (in this case students in attendance compared with enrollment) are decreased after outputs (funding levels) are determined, there will be an increase in X-inefficiency (C. G. Leathers, personal communication, 2002).
Research has shown varying levels of X-inefficiency in public organizations, including schools (Ruggiero & Vitaliano, 1999; Vitaliano, 1997; Wang, 2000). For all children to be ready to learn and to make optimal use of their learning opportunities, they must attend school and be prepared to learn. When students miss full days or partial days, efficiency is reduced. Any child who is not well or becomes ill during the day may no longer be able to focus on learning. In such cases, the child is focused on health issues, and the primary resource for the child is the school nurse. This problem is most acute in schools with high proportions of children living in poverty, in which case the school nurse may serve as the primary health professional for the child (Laflin & Hanna, 1992; Wold, 2001).
This study compared schools with a full-time school nurse to schools without a full-time nurse using a variety of performance measures related to efficiency. The performance measures focus on financial objectives and theories of economic efficiency and were developed using the Balanced Scorecard methodology (Kaplan & Norton, 2001a).
Efficiency is a function of the inputs or assumptions about the inputs in the model. In this case, the state must provide the resources to educate all students. As a consequence, the number of students enrolled was used to generate the other inputs in the model. Thus, the expected number of students enrolled has an impact on the number of teacher units funded, as well as other elements of education. To the extent that students do not come to school or leave school before the end of the school day, efficiency is decreased.
METHODS
Data Collection Instruments
Data were collected using a variety of formats including archival records, on-site structured interviews with school principals, daily checkout forms, and surveys of parents. These data were collected during fall 2001. The data on student absences were collected on a daily basis for the 20 days investigated in this study. The selection of a 20-day data collection period was based on both the burden of data collection on the participating schools and the need to collect sufficient daily data to generalize to the entire school year. The middle of the fall semester was determined to be typical of most school days. The 20 school days represent 11.4% of the total school days in Alabama (20 of 175). Thus, it may be considered that the 20 days are a purposeful but representative sample of all academic school year days. The specific data for the numbers of students leaving each school were collected using a checkout form that included date, time, and reason that the student left school.
Following the 20 days of data collection, structured interviews were conducted with each school principal to verify the school enrollment, time of dismissal, and the procedures that each school followed when permitting students to check out of school because of health-related complaints. Following the interview, actual numbers of absences for the 20 consecutive days of the study were verified through school records.
To obtain information on the number of students who had health insurance, a simple yes-no question was developed and distributed to each school. Principals were asked to distribute this questionnaire to all students so that parents could provide information on student health insurance. The number of parents responding affirmatively to the question as a percentage of total respondents was used as the estimate of the total percentage of students who had health insurance.
Selection of Experimental Schools
In fall 2001, there were 473 school nurses employed in the 128 school systems in Alabama. There were 1,471 schools in the 128 school systems, only 35 of which employed a full-time school nurse. For this study, only elementary schools were selected to avoid confounding variables (e.g., truancy) found in high schools. There were 943 elementary schools (schools that contained grades K-5), but only 17 of these employed a nurse full time. Eleven schools with a full-time school nurse were selected for this study based on whether the nurse had been employed full time at that school for at least 1 year prior to the study date.
Selection of Control Schools
After the 11 experimental schools were selected, demographic data were analyzed to aid in the selection of schools for the control group. The demographic data used were the percentage of students receiving free or reduced-price lunches, percentage minority, ADM, and estimates of average school ability as measured by the OLSAT. Based on the data for the experimental group, a control group of 11 schools was selected that matched as closely as possible the demographic variables in the experimental group. The sample included approximately 5,000 students from the experimental group and 5,000 from the control group.
Research Design
The basic research design for this study was a post-test-only design. Because no pretest data were available, a matched control group was used to control for possible variability on the initial level of the performance measures under investigation in this study. To the extent that the schools were similar on the demographic variables, there appeared to be no compelling reason to assume that the pattern of student absences would be significantly different without the intervention of a full-time school nurse.
Each student in all 22 schools was asked by the school principal to have the health care questionnaire completed by a parent and returned to the school. ADA was calculated based on the number of absences per day for students. The ADA was calculated as the number of student days (enrollment times 20) minus the number of student days absent divided by the number of student days. The resulting percentage indicates the average percentage of enrollment that is present for instruction each day of the study.
Statistical Analysis
To analyze the differences in ADA, average all-day attendance, percentage of enrollment checking out for medical reasons, and percentage of students with health care insurance between schools with a full-time school nurse and those without, independent-samples t tests were conducted. For all statistical tests, the probability of type I error was set to 5% (α = .05). For all statistical analyses conducted within this study, the appropriate effect sizes were also calculated. Because statistical significance is a function of sample size, any size difference, given a large enough sample, will result in a statistically significant difference. Using the general conventions from Cohen (1969), a small effect was defined as .2, a medium effect as .5, and a large effect as .8 (p. 12). The use of effect sizes, in conjunction with statistical tests of null hypotheses, provides a measure of the usefulness of statistically significant differences, particularly when sample sizes are large.
RESULTS
Hypothesis 1
This hypothesis was whether schools with a full-time school nurse would have higher ADA than schools without a full-time nurse. The ADA was calculated as the number of student days (enrollment times 20) minus the number of student days absent divided by the number of student days. The resulting percentage indicates the average percentage of enrolled students present for instruction each day of the study. Table 1 shows the results of an independent-samples t test between schools with a full-time school nurse and those without a full-time school nurse.
The ADA between schools with a full-time school nurse and schools without a full-time school nurse was not statistically significantly different, t (20) = 0.68, α > .05, d = .3 (Table 1). In addition, the effect size indicates a small effect for the presence of a full-time school nurse. These data indicated that both groups reported ADA for the 20-day period of the study consistent with the ADA reported for all public schools in Alabama (i.e., 95.4–95.7%). For the past several years, the state ADA averaged 95–96%. Both groups in this study fell within that range. Some might argue that increasing the attendance rate from the current 95–96% attendance rate would be a very difficult task because of the variety of reasons that students miss school (e.g., illness, appointments, parental choice). Some might argue that children who qualify for free or reduced-price meals at school attend school more regularly than children from more affluent households, in that most of their nourishment during the week might come from meals served at school. Because the control schools selected for this study were matched with the experimental schools based on demographic data including the percentage of students receiving free or reduced-price lunches, there was little variability in the data.
Hypothesis 2
Hypothesis 2 was whether schools with a full-time school nurse would have higher all-day attendance than schools without a full-time nurse. All-day attendance was calculated in the same manner as ADA except that all students who checked out during the school day were also subtracted from the total student days.
The average all-day attendance between schools with a full-time school nurse and schools without a full-time school nurse was not statistically significantly different, t (20) = 0.53, α > .05, d = .3 (Table 2), when total checkouts were included in the calculation. Once again, there was a small effect for a school nurse in the all-day attendance rate, and the effect is virtually identical to the effect on ADA. Interestingly enough, although the differences between the groups were not statistically significant, these data did isolate the fact that large numbers of students checked out of both groups of school within the last 30 minutes of the school day for various nonmedical reasons.
Hypothesis 3
Hypothesis 3 was whether schools with a full-time school nurse would have a lower percentage of students checking out of school for medical reasons than schools without a full-time nurse. Using the enrollment for each school, the percentage of enrollment checking out during the 20 days of the study was determined and compared between groups.
The percentage of student checkouts for medical reasons in schools with a full-time nurse (M = 11.1%) was statistically significantly lower, t (20) = 2.27, α <.05, d = −.77 (Table 3), than schools without a full-time nurse (M = 15.7%). In addition, the effect size of −.77 indicates that the presence of a full-time school nurse has a large impact on the percentage of enrolled students who check out for medical reasons. This hypothesis was supported by the data in this study.
Hypothesis 4
This hypothesis was whether schools with a full-time nurse would have a higher percentage of students enrolled in health care insurance coverage than schools without a full-time nurse. Control and experimental schools included a total enrollment of 10,262 students. School principals were asked to distribute the Health Insurance Questionnaire to all students with an encouragement to return them to school before the end of the 5-day week. This effort resulted in a return of 4,628 responses from parents, representing a 45% return rate. It should be noted that the form, in addition to asking whether the child was covered by health insurance, also allowed parents to indicate if they would like more information about health insurance. Consequently, there is no obvious reason to assume that parents who did not have health insurance would be less likely to return the questionnaire than those with insurance, because returning the form would aid parents in obtaining insurance. Thus, even though the response rate was lower than expected or desired, the results were considered representative of the entire student population under study.
The percentage of students enrolled in health care insurance coverage was not statistically significantly different, t (20) = 1.7, α > .05, d = .69, between schools with a full-time nurse and those without a full-time nurse (Table 4). However, when considering the effect of a full-time school nurse, it can be seen that the schools with a school nurse have a medium effect on the percentage of students within the school who have student health insurance. Therefore, although it may be that the relatively small school sample size in this study might have contributed to the lack of statistical significance, the effect of the presence of a school nurse should not be discounted. According to state and national sources, school nurses have played an integral part in helping families apply for and receive health care insurance coverage since the Children’s Health Insurance program began in 1998 (Council of Chief State School Officers, 2001; F. Shinbaum, personal communication, 2002). From these data, it was evident that a full-time nurse was able to devote more effort to assisting families with these needs than other school personnel. The nationwide effort to provide all children with health insurance apparently has had a positive impact on the school children in Alabama. The percentages of students with health insurance are relatively high (86% and 90%), but this study lends support to the notion that a full-time school nurse may play an important role in meeting this national goal.
Structured Interviews With School Administrators
Structured interviews were conducted with each principal or, in a couple of cases, his or her designee regarding specific information relating to checkouts by students. When asked who decided when a parent should be called to pick up a child with health-related complaints, all control schools responded that the secretary, the teacher, or a parent volunteer made such decisions. In contrast, all 11 experimental schools required the school nurse to assess the child and determine whether a parent should be called to pick up the child.
CONCLUSION
It is evident from this study that students leave school for various reasons. This study demonstrated that a full-time school nurse reduces the number of students who leave school because of a medical complaint. What also was evident from this study was that school-level policy enforcement, or lack thereof, allowed many students to leave school unnecessarily. The data on student checkouts revealed that a large percentage of students in both groups checked out for less-than-valid reasons, many within the last 30 minutes of the school day. Regardless of the reason that students leave school, their departure before the end of the school day results in an increase in X-inefficiency. To the extent that other extraneous reasons are eliminated through school policy and enforcement, X-inefficiency will be reduced. Thus, this study demonstrated how performance measures might be used to uncover sources of X-inefficiency. It would appear that more stringent policies relating to excusing checkouts might be one strategy that could be employed by school administrators to decrease this type of X-inefficiency. Another source of X-inefficiency exists if students with health-related complaints in schools without a full-time school nurse must remain in school without any treatment or intervention. One could reasonably infer that such students might not be able to concentrate on the educational program because of their discomfort and thus might not obtain the full benefit of the educational program. In an era of restricted funding and scarce resources, measures that would decrease inefficiency in government should be vigorously pursued.
IMPLICATIONS FOR SCHOOL NURSING PRACTICE
From this study it is possible to infer that positive externalities may result from the presence of a full-time school nurse. For example, children enrolled in schools with a full-time school nurse may have the added benefit of additional health care assessment, treatment, and early referral compared with children enrolled in schools without a full-time school nurse. Assuming that students who attend school for an entire school day have greater opportunities for learning and acquiring skills and knowledge, then all-day attendance of students would be an additional positive externality.
Findings of this study reinforce the need for individual school nurses to be proactive and creative in defining their role through the selection of performance measures. There is no indication that the public wants less accountability in public organizations, including schools. The mission of public schools is to educate students. Although school nurses may provide direct educational services to individual students or groups of students, most school nurse services could be described as indirectly related to the education mission. When selecting performance measures for school nursing, it is easy to think of those that are easy to measure, such as the number of students screened for spinal disorders or the number of presentations made to teachers and students. However, such measures may not provide a relevant link to the organization mission and may not be the best choice of performance measures. Performance measurement should not attempt to measure everything that is measurable. Whatever measures are selected will become very important. School nurses need to seek measures of performance that tie back to the overall mission of education, whether these measures are direct or indirect. When performance measures are selected with careful thought, the measurable influence and impact that a school nurse provides for students and staff may extend well beyond the expectations of educators and school nurses.
RECOMMENDATIONS FOR FUTURE RESEARCH
Because of the complexity and “noise” found in the data regarding all-day attendance, additional study is recommended in this area. Additional study could indicate policy implications such as checking attendance at the beginning and at the end of the school day to get a clearer picture of the all-day attendance of students. Additional policy implications may address implementation of stronger policies and procedures regarding acceptable reasons and limitations for early checkouts. Although many factors contribute to whether a child has health care insurance, it appears evident from this study that full-time school nurses contribute to enrolling children and alerting families to the health care options that may be available. Additional study in this area might target other factors. Finally, this study did not attempt to link the presence of a school nurse with any measures of student achievement. This linkage may well exist but was not investigated as part of this study.