Paradoxes of Efficiency in Education (Part 2)

Introducing an innovation to increase efficiency ending up with more inefficiency is a paradox. Most obviously, it occurs in transportation: fuel efficient cars that get more miles per gallon of gas than before end up multiplying demand for more such vehicles  putting more cars on roads spouting gases into the air. And in medicine–see Part 1. Such paradoxes of efficiency occur as well in education.

Are there jobs in which there are few gains in productivity–that is, workers produce more at less cost–yet wages of these “unproductive” workers rise over time. None?

Wrong!

Think of a string quartet playing to a live audience 300 years ago. The number of musicians and the time they needed to play a Beethoven sonata in the late 18th century  haven’t changed, yet today’s quartets playing the same sonata in the same amount of time  make far more than those four musicians centuries ago. Why is that?

Economists William Baumol and William Bowen used that example in their 1966 article to make the point that in certain people-dependent, service-driven, labor-intensive occupations such as the arts, health care, and education there are few productivity increases (e.g., musicians are not paid to play Beethoven sonatas faster nor are teachers paid more to teach history faster). Traditionally, factory and business employers raise wages for workers as a result of new technologies and managerial techniques that increase employee productivity–making and selling more products at less cost than before. Higher wages usually follow gains in worker productivity.

No so for labor-intensive work. The technology used to play for live audiences, teach students, and care for patients leaves hardly any room for labor-saving innovations to increase efficiency since the product is actually the labor of the musician, teacher, and care-giver. But to retain experienced, talented, and hard-working artists, teachers, and health care workers and to keep them playing, teaching, and taking care of the elderly and ill, salaries go up over time. This paradox of efficiency is called “Baumol’s Cost Disease.”

In applying Baumol’s Cost Disease to education, additional paradoxes of efficiency turn up, for example, around class size and applying technologies to teaching and learning.

Class size

A most common move to increase efficiency of teaching in both K-12 and higher education is to make classes larger–one way of skirting “Baumol’s Cost Disease.” As one researcher in the late-1920s put it: Larger classes make for fewer teachers and lower building costs. Increasing the size of classes, then, offers an obvious and tempting means to immediate educational economy.

How large or small class size should be–without losing “effectiveness” however measured–has continually been contested. Researchers at all levels of schooling have done comparison studies since the early 20th century (see here here, here, and here) to determine exactly what class size is both efficient and effective. Research findings in the early 21st century remain contested for the simple reason that while the costs of providing teachers in classrooms drops as a district increases class size, questions of teacher effectiveness in achieving desired student outcomes arise time and again (see here and here).

Historically, class sizes in urban schools in the early 20th century ran to 50-plus students and have fallen each decade until they range in the 20s to 30s.  In most K-12 instances, both parents and teachers sought smaller class size, often to sizes below 20 students per class because they believed that smaller classes would give teachers time to build relationships with students, work with individuals, and boost students’ academic outcomes. As class sizes dropped, costs for hiring additional teachers and finding space for those smaller classes rose. And that is the paradox of efficiency that policymakers have found themselves in repeatedly when it comes to class size.

Another bind in which policymakers find themselves is not knowing what the exact number of students per class is best insofar as measures of academic achievement. A tradeoff between efficiency and effectiveness continues to plague policymakers although most practitioners and parents urge smaller class size than now exists.

In higher education, increasing institutional efficiency in the face of rising tuition costs and low faculty teaching loads was to have more lectures for hundreds of students in first two years of college and reserve seminars for 15-25 students in the final years of an undergraduate’s career and graduate school. And now to increase efficiency, offering  online courses is believed to reduce costs and provide access to professors.

Increasing class size through scheduling large lectures in introductory courses, reserving seminars for advanced students and increasing online instruction–are efficiency measures that often end up in students frustrated in having little contact with professors, questioning the effectiveness of larger classes and online learning while turning off students to the course content they seek to learn. So decreasing or increasing class size in both K-12 schools and higher education lead to greater inefficiencies.

New technologies

And, of course, there is increased use of new technologies, the all-purpose solution to inefficiency.  There is little doubt, even among skeptics, that computerization of the administrative side of schooling–personnel actions, budgeting, purchasing, and collecting student data–have been streamlined and desks piled high with records and folders have decreased yet the numbers of administrators both in K-12 schools and in higher education have increased. Similarly, on the curricular and instructional side of schooling, the paradox of efficiency has become apparent. Examples abound.

*Online courses in K-12 and higher education have been touted as being cost-efficient and cost-effective for decades but proliferation of such courses have raised questions among current and prospective students if that kind of instruction and teacher contact is what they want. The belly-flop of Massive Open Online Courses (MOOCs) and its high attrition has added to those doubts as has the continuing spread of online classes and numbers of students who fail to complete courses. Technologically induced efficiency leads to larger inefficiencies in dropouts and frustrated teachers and students.

*With standardized tests increasingly being taken online, districts have spent additional funds beyond buying devices and new software (and maintaining both before they sink into obsolescence) and more dollars to train students and teachers to use online tests (see here and here).

*”Personalized learning” is the Holy Grail of efficient teaching–each lesson adapted to the strengths and weaknesses of each student. This has been the dream of school reformers for decades. With the flow of tablets, laptops, and phones into schools for instructional use, the day of efficiency has dawned. Here is where the paradox kicks in.

In the quest to make teaching and learning faster and better, a wealth of technological devices and software have been mobilized and put into classrooms. In the name of efficiency and effectiveness, current students have far more access to technologies than students 35 years ago when they were initially introduced. Yet test scores and other measures of academic achievement have not climbed as more machines and software have spread through U.S. classrooms. Nor have the amounts of money being spent on these new technologies decreased as they have become ubiquitous. Another instance of the quest for efficient teaching and learning leading to inefficiency.

 

 

Advertisements

13 Comments

Filed under Reforming schools, school reform policies

13 responses to “Paradoxes of Efficiency in Education (Part 2)

  1. Reblogged this on 3-Star learning experiences and commented:
    Larry Cuban writes in his latest blog on “Paradoxes of Efficiency in Education “:
    In applying Baumol’s Cost Disease to education, additional paradoxes of efficiency turn up, for example, around class size and applying technologies to teaching and learning.
    In the quest to make teaching and learning faster and better, a wealth of technological devices and software have been mobilized and put into classrooms. In the name of efficiency and effectiveness, current students have far more access to technologies than students 35 years ago when they were initially introduced. Yet test scores and other measures of academic achievement have not climbed as more machines and software have spread through U.S. classrooms. Nor have the amounts of money being spent on these new technologies decreased as they have become ubiquitous. Another instance of the quest for efficient teaching and learning leading to inefficiency.

  2. Great argument! I think there are some important nuances to extend the argument.
    1. While we know a great deal about how to measure learning, use of the typical state norm-referenced test to measure learning of curriculum standards only represents a variety of policy-level misconceptions about the nature of these tests and the standards and the relationship of both to the enacted curriculum. So as a practical matter, we don’t have adequate measures of learning in common use. Thus, it makes no sense to discuss cost-effectiveness.
    2. As you argue so well, measures of efficiency can be profoundly misleading in education. I would argue that this problem has not been solved in a century of research on instruction. Thus, we have no valid way to make a cost-efficiency argument for learning. However, there are many non-instructional measures of efficiency that are probably valid for non-instructional tasks performed by teachers and schools, and these are prime targets for automation through technology. Familiar examples include computer-administered testing (justified by lower cost of administration when compared to paper-based tests), e-books (if justified by lower cost than print books), automated record keeping and reporting (if justified by reduced demand on teacher time), and the like. Note that none of these automation strategies should be expected to make any claim for improved learning and instruction.
    3. We do know that direct instruction is effective when properly designed and delivered, by various mixes of teacher and computer delivery (see the 2018 meta-analysis by Stockard, et al, in Review of Educational Research). This allows for an argument that technology used in this way can improve learning for students, especially for those who otherwise would not have access to properly designed direct instruction (this rules out most MOOCs). This is a cost-benefit argument.
    4. Unfortunately, many who embrace online instruction uncritically do so based on a simple cost comparison, without any consideration of effectiveness. As you argue, this kind of “cost saving” is highly misleading, whether or not technology is involved, and can lead to unintended consequences.

  3. Laura H. Chapman

    I love the example of highly skilled performances of music of a certain kind and the economic concept of Baumol’s Cost Disease.” Ballet provides another example. Rube Goldberg solutions to problems illustrate some possibilities if you chase effectiveness without concern for efficiency.

    I am reminded of the bizarre federal definition of “student growth” as “a change in student achievement for an individual student between two or more points in time” (Fed. Reg., 2009, p. 59806). “Achievement” actually referred to scores on standardized state tests, administered in a specific subject to all students at a given grade level, and constructed to yield a distribution of scores approximating a bell curve.

    Those tests and scores, never designed to evaluate teachers, were nevertheless used for that purpose. So, the follow-on teacher ratings were defined like this. An “effective teacher” is a teacher whose students achieve acceptable rates (e.g., at least one grade level in an academic year) of student growth. A “highly effective teacher is a teacher whose students achieve high rates (e.g., one and one-half grade levels in an academic year) of student growth…” (Fed. Reg., 2009, p. 59805).

    I think Robert Linn came up with one of the first graphs to illustrated the problem with NCLB’s focus on “growth in scores” with hypothetical “trajectories” over time. Before Arne Duncan’s Race to the Top, education was being reduced to a race–with time for learning, rate of learning, and tested course content to be covered (distance) all that mattered. Start behind… and you cannot catch up unless you learn more and faster. I think Robert Linn was among the first to illustrate that problem (see Figure 4 https://onlinelibrary.wiley.com/doi/abs/10.1111/j.1745-3992.2005.00021.x ).

    Later on, I tripped on a report from scholars associated with MetaMetrics. They were chasing the idea of extracting greater efficiencies in teaching reading. They hoped to set “growth velocity standards for learning.” They described their theoretical mapping of “aspirational trajectories toward graduation targets” in reading skills as analogous to “modifying the height, velocity, or acceleration respectively of a projectile launched in the physical world.” I kid you not. (Williamson, Fitzgerald, & Stenner, 2013, p. 63). They appeared to be seeking greater precision in setting targets and cut scores for grade-to-grade progress in meeting the CCSS. https://www.studentachievement.org/wp-content/uploads/Text-Complexity-Ed-Researcher.pdf

    I am not an expert in quantitative research. Even so, there seems to be a paradoxical end game with a system of teacher evaluation dominated by the production of gains in test scores. If the system works as intended, the worst teachers will be fired for failing to produce their expected gains, and the remaining teachers will change their practice in order to improve their evaluations.

    Over time, variation in teacher ratings will decline. More teachers will receive high ratings. These high ratings will be criticized as too high (perhaps because teachers are insufficiently rigorous) and so the bar will be raised.

    At some point, teaching practice will be so improved that NO significant variations show up on tests and the rating system is worthless. You cannot do the statistical gymnastics for value-added scoring if students’ scores go up, but measures of variation in teacher practice go down. The whole evaluation system is undermined.

    This sequence of events is the logical outcome of the circular reasoning that operates in current reforms. Value-added measures have been thoroughly discredited. But they are still used in Ohio and other states. I love this brief: VAMs Are Never “Accurate, Reliable, and Valid” http://journals.sagepub.com/doi/pdf/10.3102/0013189X16651081

    Really a great pairing of the two posts. Thank you.

    • larrycuban

      I appreciate the comment and links you provided, Laura, about misapplication of efficiency and effectiveness criteria to teaching and learning.

  4. There is a lot of BS circulating around and within the hyper-focus on learning efficiency, standards-based learning, standardized test scores, and teacher accountability.

    Any school system that is struggling is only as strong as its weakest links. If the weakest links cannot be strengthened, few gains can be made. Each school district has its own weak link(s); analyze the list below and work on that

    Standards/Curriculum
    Best practices/pedagogy
    Teachers (qualifications/experience/skill sets)
    Support staff (aids/counsellors/social workers)
    Administration
    Teacher evaluation/support (mentoring)
    Schedule/Interruptions
    Grading policy
    Instructional technology
    Funding/resources/allocation of
    Physical learning conditions
    Promotion policy
    Discipline policy
    Academic requirements
    Academic options
    Pathways to success
    Class size
    Student motivation
    Student behaviors/attitudes
    Student attendance

    • larrycuban

      Thanks for taking the time to comment.

    • This is really helpful to increase efficiency. The paradox is the most helpful to increase the learning capability of the students who want to be more efficient. In this present days technology is essential to learn something clearly. A student need to give more effort in their study. But there are many institute in Urban area they do not know the importance of technology and their classes are not longer as their requirement. Most of the students depend on self study. How to improve their education system? How to increase the use of new technology in Urban area and the class?
      For the technology system they do not study properly and can not do any work which will be efficient for the world. The technology system is essential from school life but many student do not know what is the technology?

  5. Pingback: The efficiency paradox | Kate's Crate

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s