Teachers as health workers; a consideration of the economic impacts using unique survey data

Abstract

Teachers face a multiplicity task in undertaking their work. Some of these tasks are not primarily educational and may be described as health related. In the provision of these services there is a net transfer of economic benefit from the Education Sector to the Health Sector. Disentangling the net amount of this transfer is difficult and depends largely upon the substitutability between teacher-initiated health-work and “market” based health work which might reasonably be carried out by health professionals.

The paper, belatedly published below, was originally presented at the Oxford Educational Research Symposium 2016 and is one of the first to attempt to disentangle these links and place a value on the extent of the net subsidy provided by teachers. It uses a unique data set compiled by MacDonald (et al) and combines this with an economic impact model developed at the University of Queensland, Faculty of Business, Law and Economics. The results of this study will influence both teacher training and remuneration and well as highlighting the important role played by teachers in student health. Readers should remember that financial estimates are in 2016 Australian dollars

Due to space considerations endnotes and model description have been omitted.  A copy may be obtained on request from the author

1.0 Introduction

The modern teacher has several other roles other than being an educator beyond that of an educator. They have a multiplicity of roles. While a teacher spends most of the day in the classroom, the actual teaching component is only part of the job. Teachers in modern classrooms are no longer lecturers, they are facilitators, their main task is to set goals and organise the learning process accordingly(http://www.sens-public.org/article667.html?lang=fr)

 An effective teacher understands that teaching involves wearing multiple hats to ensure that the school day runs smoothly, and all students receive a quality education.  In conducting their work, teachers become aware of heterogeneity among students, not just in terms of learning ability but also in terms of mental and physical health. Alperstein, G and Egan, M (2012) Education and Student Health; The Big Picture, The Department of Education, Employment and Workplace Relations (DEEWR), Canberra)  Common health related problems among students include, mild physical and intellectual disabilities, issues relating to nutrition and personal care issues. There is likely to be school based or generalised procedures and guidelines in place to assist teachers in dealing with these issues. For example, the Victorian Education Department website advises:

“It is never the job of a teacher to diagnose medical conditions. Do not make assumptions, but once a diagnosis is made by appropriate professionals you are able to seek further targeted assistance if necessary” (Victorian Department of Education, 2014).

However, there are likely be student-specific and unexpected health related issues that teachers face in and out of the classroom that require judgement on a case by case basis. This paper is not about the specific instances where teachers may be required to assist in health related issues but rather examines the economic implications of educators devoting some of their time to what is essentially non-teaching related issues. It highlights several issues.

• How much of the paid for time of the teacher devoted to issues that are primarily health related as opposed to those that are primarily education related?
• To what extent does this diversion of education time impact on the overall effectiveness of the provision of educational services?
• Are all, or the majority, of in-classroom health issues better left to health professionals of an appropriate skill level?
• What are the economics of the current system of mixing education and health responsibilities within the portfolio of the classroom teacher?

He paper uses a unique data set compiled by the University of Queensland, School of Education and combines this with a non-linear economic impact model (QNLMRM) operated with the Australian Institute for Business and Economics (AIBE)

1.1 Some Economic Considerations

The UQ study has established that teachers routinely undertake activities that may be classified as health related. The extent of this health work (or the propensity of each teacher to undertake health work) varies by school type, student characteristics and by the characteristics of the individual teacher.  Like all human activity these have economic consequences which may be quantified along as appropriate data are available and some modest simplifying assumptions are made to the likely consequences of teacher behaviour in the health related area.  For example, it is valid to assume that the health interventions are neutral or net- positive (that is the intervention is not harmful or reduce the welfare of the student).   Even with this simplifying assumption, the economic factors that come into play through health related teacher activity are complex. There are several potentially counterbalancing economic trade-offs, based upon the concept of opportunity costs and cross subsidization.  By way of definition, opportunity costs in economics may be defined as.

Opportunity Cost = $Return of Most Lucrative Option – $Return of Chosen Option          (1)

It defines the cost of devoting resources to one activity in terms of the potential loss of benefits from using the same resources in an alternative use( http://www.education.vic.gov.au/school/teachers/health/Pages/default.aspx?Redirect=1)  For example an opportunity cost of a teacher devoting part of work time to health related activities is the value of that time devoted to normal teaching activities.    Cross subsidization is a slightly more complex term and its outcome in terms of net benefit is difficult to judge. In the case in point, a teacher responding to a health issue reduces the need for intervention by school of government based health services, either proportionally or more than proportionally (if the early intervention reduced the eventual health consequences). In this situation, the employer of the teacher has cross subsidized the health providers but at the cost of loss work time frothier employee.  As a further complication there are opportunity costs associated with various forms of cross subsidisation.( http://www.education.vic.gov.au/school/teachers/health/Pages/default.aspx?Redirect=1

 In the case of teachers engaged in health work there are a few potentially counterbalancing flows.

1. The element of teacher time diverted from “education” to health work represents an economic transfer from the employers of teachers (State Government, Private organisations) to providers of health services( Within Schools or to Health authorities)- if the students would have sought alternative help from heath providers and/or if the teacher intervention prevented a subsequent and possible more costly call on the health services at a later date (i.e. early prevention effect) in other words there is both a current and future net transfer from education to health. This in effect is the shadow price for teacher services devoted to health related activities. There may also be a private beneficial transfer between the school and parents who are either neglectful or unable to fully cater for the health benefits of their children.
2. Teacher productivity impacts. These are complicated. If substantial amounts of teacher time are diverted from planned class activities, there are likely to be impacts on the productivity of the teacher (in terms of quality delivery of teaching services) and these will impact disproportionately on “healthy” students who did not require attention but have now had less formal learning time. In other words, students that do not require health assistance are cross subsidizing those that did. A further complication is that the well placed health activities of teachers may help them connect better with some students and improve teaching productivity for the class over time.
3. Teacher production frontier implications. Following on from point (2) to this point it has been assumed that teachers are fully occupied in providing educational services and that any diversion from that schedule to undertake non-teaching activities will represent an economic cost. In economic terms we assume the teacher is on their production function.. This may not be the case. If a teacher is below their production function (and fulfilling all required duties), then any additional tasks that do not detract from their primary role is a net benefit, divisible between all stake holders. Finally, the issue is further complicated by synergy effects- teachers are often in the “right place” to administer health work, they have the trust of students and by administering this health work they may enhance the value of the current and future educational services. In economic parlance- this will either move them closer to their production frontier or move the frontier itself to the right.
4. Within School transfers. School management is responsible for all aspects of school operations, they may arbitrage between those resources primarily dedicated to teaching and those dedicated to school welfare including health issues to achieve overall cost minimisation

2.0 Quantifying some economic considerations

Economic quantification, especially where both private and social costs are involved requires some simplification to be used. There is also likely to be some data related issues and two important issues: economic value/cost and its distribution. For example, it is relatively straight forward to calculate the potential value of health–related work by equating it to the opportunity cost in terms of lost paid-for teaching time. However, it is much more difficult to disaggregate that benefit between the potential beneficiaries. For example, the potential beneficiaries of such a transfer include

• The student
• Parents
• Other teachers who may teach the student concerned
• The school
• Governmental health services

The exact distribution of these benefits between the above will depend upon the nature of the health issue, family circumstances and the relativities between school provided and public provided health facilities and in the absence of longitudinal unit record data would prove very difficult to estimate with any degree of accuracy.  Consequently, this section limits itself to examining the potential transfer of economic benefit from schools to the above in aggregate. This, by itself represents important new information which will have implications for teacher training, teacher remuneration and the case for widening the considerations of the social and economic benefits of schools

 2.1 Quantifying the shadow benefit to society of Teachers providing un-costed for health work

The methodology used below is in three stages.

1. Arithmetic estimation of the wage costs of time diverted to health related issues across Queensland Schools- This will provide the most important element of the direct costs. Which can then be used to shock the economic model through the Education sector
2. Consideration of other benefits (such as early prevention benefits) for possible inclusion in the direct costs
3. Application of the direct costs to the Queensland using the Non Linear Multi-Regional Model (QNLMRM) developed at the University of Queensland.. This model will allow the full economic benefits of the implicit health sector subsidy to be quantified

2.2 The direct (first round) benefits

Calculation and monetisation of the direct benefit is based on the following assumptions and uses the survey data across the 386 surveys to generalise about the Queensland teacher workforce in aggregate. This of course has its dangers. To this point no formal testing of the randomness of this survey has been taken and results achieved should be considered under this caveat. In defence of our use of these data it is recognised that due to extensive (in contemporary terms) unionisation among the teacher workforce, the wages paid across the sector, irrespective of type (State, Catholic or Independent) are constrained within a relatively narrow band.

• On average (mean), teachers (n=386) reported spending 592 minutes per week (9 hours 52 mins) undertaking health work with their students or 380.8 hours per year
• Current average wage of teachers is set at $86,032 per annum. This is based on a Band 3, Step 4 salary for QLD teachers as of September 2015 – basically, an experienced teacher that doesn’t hold any additional position of seniority within a school. Teachers with 7 or more years teaching experience made up 75.9% of our sample
• These data suggest an average wage per hour based on 36.5 hours of $45.32
• If all that time were devoted to teaching the economic value (and hence the shadow price of health work diversions) . would be $17261 per year, per teacher.
• Queensland schools employed the equivalent of 51,258 full-time teaching staff in 2011. Of this total, 28,332 were employed in primary schools and 22,927 in secondary schools (OESR, 2012). However, the number of numbers of active FTE (Full time Equivalent) .teachers in Queensland, making allowance for non-active teachers  at approximately  was approximately 45,000
• This suggests a very large potential total diversion cost estimate (shadow cost of teacher administrated health work) of $776, 745,000

This amount would form the direct exogenous shock to the QNLRM to estimate the total societal benefit from the additional health services provided by teachers. However, it can’t be assumed that all activities nominated as health related are totally divorced from “active teaching. Some of the actions are synergistic to the teaching role and may even be a net benefit to the overall teaching performance. As well, some tasks are of a nature that they could not realistically be transferred to health professionals, in that they may be one off task and they arise at random times (although there would be a concentration of incidents in the morning and before school) and require immediate attention. While these incidents may represent a cost in terms of lost teaching, they represent part of the pastoral care expected by teachers and not a function where health professionals could easily substitute for teachers. Effectively there is no shadow price for these activities.

The issue of shadow costing then becomes how much of the total amount of time teachers devote to health related work represents a cost to the potential economic value of their teaching if it was conducted in a manner not interrupted by health related work.

Data does not allow a disaggregation of the health related work of teachers by nature of the task. As a result, our preliminary evaluation is based on discussions with teachers and those conducting the survey. This suggest that the net reduction in teacher effectiveness is conservatively estimated at 25%. On this basis the direct exogenous loss from teaching (direct subsidy to the health industry) is $194.2 Million per annum, conversely this represents the first round social gain to society through the subsidy to the health sector.  It should be stressed that this represents a conservative estimate of the net transfers currently taking place from the Education sector to the health sector and does not consider the accumulating value of early intervention services provided by teachers on the future costs of health problems or the social value to parents and society in general.

3.0 Total Economic Benefit

Economic impact assessment intends to explain how economic flows from one industry sector (or a project or activity within that sector) affects other sectors in the economy. I teachers did not perform some health related activities then burden would need (now or in the future) to be taken up by components within the health industry, this represents a direct subsidy (transfer from the education sector) and is analogous to a direct or exogenous injection of funds into the health sector. It therefore can be modelled under standard economic impact methodology

The economic modelling undertaken in this report makes use of a specifically constructed non-linear (marginal coefficients) Input-Output table for Queensland and the rest of Australia, The Queensland Non-Linear Multi-Regional Model (QNLRM) This type of model is well suited to modelling economic impacts in the Education and Health areas  by allowing for the generation of non-proportional or non-linear results. That is, it allows us to investigate economic impacts that are expected to differ in magnitude and importance across time as the economy changes.

This model has been successfully used in several applications, both in a Queensland context and in a “Rest of Australia” context. The model is based on the non-linear properties of the IO-9 software developed by Guy West from the Centre of Economic Policy Modelling at the University of Queensland. The model is regularly updated both by the addition of newly available data and by recalibration of the elasticity estimates that are used as external inputs into the model. The latest update of the model occurred in 2015, when data on Retail services and power usage were added to the IO table and the table constrained to the latest estimates of State GSP.

 In the economic modelling lexicon, non-linear modelling sits somewhere between traditional (average coefficients) input-output modeling and computable equilibrium modeling (CGE). For a modeling exercise such as this it has advantages over both alternative methods:  For example, the Non-Linear Input-Output Model (NLIO) removes one of the major limitations of standard input-output analysis by removing the assumption of linear coefficients for the household sector and allowing marginal income coefficient adjustment.  This is because, as is widely known, the household (consumption) sector is a dominant component of multiplier effects in an input-output table.  As a result, using marginal income coefficients for the household sector will provide a more accurate and empirically more valid, estimate of the multiplier effects, which in turn, provides results closer to those of a computable general equilibrium (CGE) model.  NLIO has the advantage over CGE in that it retains the flexibility and relative data parsimony of an IO model in comparison with CGE, which is structured to estimate elasticities internally and therefore, has much higher base data requirements. As well, NLIO is better suited to the evaluation of partial equilibrium (one off projects) or events such as changes in one part of an industry sector as well as being better suited to regional evaluation.

The non-linear component is introduced into the model by the interaction of estimated and externally imposed elasticity coefficients upon the multipliers, particularly the employment and factor income multipliers. In essence the quantification of economic impact relies on the quality of input data and the ability of the modelling process to correctly interpret that data.  The Queensland Non-Linear model (QNLMRM) has performed well, in terms of accuracy of results and model diagnostics for several impact studies

17 Intermediate Sectors

• 4 Primary input sectors
• 3 Final demand sectors

The sector composition is shown below in Table 1

Table 1 Sector Composition of QNLMRM

Sources of 2 Digit ANZSIC Australian and New Zealand Standard Industrial Classification (ANZSIC) 2006 (Revision 2.0) 2013 1292.0

This structure was designed to maximise detail of each sector and its interaction with the rest of the Queensland economy and operates a few default values that estimate the average breakup of each sector into intermediate goods, primary inputs and final demand sectors.

 The specific advantages of this type of method are outlined more fully in the appendix to this report, but primarily, this technique removes some of the well-known problems of traditional Input-Output modelling, whilst maintaining the flexibility and relative data parsimony that have made Input-Output analysis such a popular and continuing method of economic analysis. 

The primary economic impact measures derived from the model are:

• Gross Output/regional Turnover – refers to the value of increased production from an additional economic activity. Within this gross value is included the value of raw materials that, in most cases, have already been counted as part of gross output from earlier production.  Therefore, there is a tendency for Gross output figures to include some double counting.  As a result, more concentration is placed upon incremental (additional output created), or value added.  Nevertheless, the concept of Gross output should not be abandoned because it is a good indicator of the level of turnover in the economy and hence a good measure of the total level of economic activity.
• Value Added – refers to added or net output. Value Added is equivalent to the Gross State Regional Product as used by the Australian Bureau of Statistics.  It is the measure usually preferred when measuring economic impact.  It measures the added value placed on intermediate products (raw materials) from the productive   It is made up of margins, wages, profits and transfers.
• Factor Income – relates to the share of value added (and gross output) which is directly paid to individuals or firms in the form of wages and or profits. It is a percentage of value added and cannot exceed value added.
• Jobs – relates (usually) to the amount of labour required for the level of production. Depending upon the type of activity, job numbers measure either the use of existing labour (continuing jobs) or hiring new staff.  Full Time Equivalent (FTE’s) employment refers to the number of full time person-years of employment generated by a particular project or event.  This alleviates the overstating of the level of job growth due to the stimulus. Note that employment outcomes relate to the whole period of the project and are not annual impacts.

To estimate the full social benefit of the health transfer from teachers we apply estimated health subsidy ($194.20) M as an exogenous shock to the economy through the Health Care and Social Assistance sector with the QNLMRM this produced the following results

 Table (1) Estimated total annual economic impact of Teacher Health transfer on Queensland Economy  $Millions

Source QMLMRM (2015)

Specifically, this suggests that expenditure by the health work subsidy provided by teachers in Queensland produces a total economic effect per annum

• Average annual Turnover or Gross Output gain of $ 414.09 million
• Annual Value added or net additions to GSP of $232.22 million
• Annual- gains to factor income of $135.63 million.
• Annual support for approximately 1032 FTE jobs

In terms of our specific example, the social value of the health transfer, if injected directly into health sector budget would produce a net addition of $232.22illion to Gross State product (GSP) and support over 1000 jobs (1032). If teaching effectiveness was not affected by this diversion in tasks, the amounts shown above would represent a net welfare increase to the state.

Is it Efficient for Teachers to Do Health Related Work?

While there is no doubt this type of transfer does occur and produces significant economic benefits in terms of net transfers, Is this an efficient use of society’s resources? The concept of opportunity costs is important here. The opportunity cost of teachers doing health work is the loss of education time, with the net trade-off being the benefits of the health work versus the cost of the loss of education time.  To arrive at an initial approximation the initial direct exogenous transfer of $194.2 was applied to the QNLMRM through the Health Sector (as above in table 1) and the Education sector and differenced. The results are shown in table 2

Table (2) Net Welfare Transfers from Teacher Health Related work

Source QMLMRM (2015) and data from UQ survey

The results shown in difference table 2 show (relatively small) positive gains in all components, indicating that there has been an across the board gain in welfare from teacher initiated health work after discounting for economic loss of standard teaching time. These net gains would be distributed across the economy but would be most concentrated in the Public Administration and Safety and Health care and Social Administration Sectors.  The result indicate that the current system is efficient compared to the alternative of teachers doing no health work and those tasks being taken over by health professionals.  However, several caveats attach to this conclusion. First the results are driven by the fact that in the QNLMRM, value added multipliers are slightly larger for Health Care and Social Assistance than they are for Education and Training[xi].  Consequently, it is inevitable that any transfer of income from Education to Health will be net positive.  Within economic models, the Education and Training and the Health Care and Social Services are considered in aggregate. We have not attempted to isolate those smaller subsectors of each industry which deal with the school sector in both an educational and health related way. It is possible that consideration of these subsectors in isolation may give a different result[xii]  Secondly, some of the important potential transfers, including the present value of early teacher detected health issues have not been quantified and included in the analysis. As these would only further add to the value of the current system their inclusion would only strengthen the initial conclusions, however further quantification of the benefits would strengthen the case for improved teacher training and raise the profile of the pastoral care work carried out by teachers.  Thirdly, we have not solved the distribution issue. In the analysis the educational transfer is assigned in bulk to the Health Care and Social Assistance sector. Earlier we identified several stakeholders who would benefit from teachers undertaking health related work. Benefits to parents and to the budgetary position of the host schools have not been considered. It may be that some School administration is saving on operational costs by expecting Teachers to undertake essentially unpaid or under-paid tasks as well as opening themselves up to litigation risks.  These considerations open considerable scope for further research.

Conclusions

The UQ study confirm that teachers engage in considerable amounts of health related work of varying degrees and intensity during their working week; approximately 10 hours per week or 360 hours per year. Teacher also serves as an effective early warning system for students that have or in danger of developing physical and mental health issues.  Assuming the basic unit of output for a teacher is instructional services to healthy students, health related activity disrupts these tasks by at least the amount devoted to the health work plus any additional disruption time. This has economic consequence for the State, the School and for Parents and Friends.  Importantly it also represents a windfall gain to health providers who may otherwise have been required to provide the health service input- ranging from, other teachers, the school nurse, paramedics, doctors and hospitals.  In other words, a cross-subsidization is occurring between education providers and health providers. In this paper we use data from a unique survey that yielded information on the extent and type of health related interventions undertaken by teachers, combined with aggregate wage costs (on costs (fixed costs of employment) were not considered as thee are considerable variations among school types and because these costs are fixed and are not dependent upon undertaking health related activity) and applied as a direct subsidy to the Health care and Social Assistance sector. The Queensland Non-Linear Multi-Regional Model was used to estimate the total economic impact of the transfer.  It was found that the welfare benefits to the State economy were considerable at approximately $230 million per annum net value added.  This subsidy also had corresponding income and employment benefits to the State economy.  However, to this point we have simply valued the current system under which teachers cross subsidized. The next question to ask was, is the current system efficient? There are several efficiency tests available to compare outcomes. Here we chose the efficiency test of relative opportunity cost. Put simply, which was the least cost alternative; maintain the current system or transfer all teacher related health activity to the health sector? By comparing alternative impacts of the direct value of the transfer between the donor sector (Education and Training) and the host sector we found that the current system had small net welfare benefits to the Queensland economy over the alternative of the cessation of teacher initiated health work. A list of caveats to this conclusion emphasised that this conclusion is most likely understated.  Finally, the thorny question of the distribution of benefits was essentially avoided. If teachers, through their actions are generating net private and social benefit then the law of marginal productivity wage payments would suggest that part of these benefits be returned to the teachers in the form of increased wages.

References

Alperstein, G and Egan, M (2012) Education and Student Health; The Big Picture, The Department of Education, Employment and Workplace Relations (DEEWR), Canberra

Coates, D (1998) “Education Production Functions Using Instructional Time as an Input” paper presented to Public Choice Society, New Orleans,

http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.501.1725&rep=rep1&type=pdf

Fraser Institute (2007) “Why do Private Schools Exploit Their Employees? http://www.nedc.com.au/files/Resources/Teachers%20Resource.pdf

Hagan, P. and Mangan, J. “Guidelines for Economic Modelling Using IO Analysis” Government Statistician’s Office 1997, Brisbane

Hanushek, E. (2007) “Education Production Functions” Palgrave, Harvard. http://hanushek.stanford.edu/sites/default/files/publications/Hanushek%202008%20PalgraveDict.pdf

Heald, D (1996) “Contrasting Approaches to the Problem of Cross Subsidy” Management Accounting Review 7, 53-72

Investia (2012) http://www.investopedia.com/terms/o/opportunitycost.asp

National Eating Disorder Collaboration (2012) “eating disorders in Schools” http://www.nedc.com.au/files/Resources/Teachers%20Resource.pdf

Queensland Teacher’s Union  (QTU)

http://www.actu.org.au/media/349581/queensland-teachers-union.pdf

Severson, H (2007) et.al “Pro-active, early screening to detect behaviourally at risk students: Issues, approaches, emerging innovations and professional practices” Journal of School Psychology 45, 193-233

Szucs, E (2009) “The Role of Teachers in the 21^st Century” Sens Public Review Web, http://www.sens-public.org/article667.html?lang=fr

Todd, P and Wolpin, K (2003) “On the Specification and Estimation of the Production Function for Cognitive Achievement” The Economic Journal, 113 F3-F33

Victorian State Education (2014) “Student Health and Well Being” for a discussion of health issues among students commonly faced by Victorian Teachers.  http://www.education.vic.gov.au/school/teachers/health/Pages/default.aspx?Redirect=1

See for example, KPMG (2009) “The Economic Impact of the Ten Days on the Island Festival”

 Note that expenditure on such items as airfares have been assigned to Queensland as a proxy for “national” expenditure 

 or Economic Modelling Using IO Analysis” Government Statistician’s Office 1997, Brisbane