Beyond Taylorism, Part II

When we realized ____, we did what most large organizations do when they find themselves falling behind the competition: we worked harder. We deployed more resources, we put more people to work, and we strove to create ever-greater efficiency within the existing operating model. Like obnoxious tourists trying to make themselves understood in a foreign country by continuing to speak their native tongue louder and louder we were raising the volume to no good end.

Fill in the blank with “our education system wasn’t preparing students for the demands of a more competitive economy”, and the rest of the paragraph is a pretty apt description of how policymakers at all levels of government have approached education reform since the release of A Nation at Risk. The passage is actually from Team of Teams, describing how the Joint Special Operations Task Force’s initial reaction to the nimble, decentralized enemy they encountered.

It didn’t take long for Gen. McChrystal to understand that achieving victory would require much more than either adding more resources or operating with more efficiency — it would require a complete shift in how his organization operated.

Jal Mehta’s The Allure of Order shows how the last few decades of education reform (for the most part) followed the default approach described in the quote above. Inflation-adjusted spending on schools and the number of employees staffing them have both skyrocketed relative to the modest growth in the number of students they serve, but we’ve only seen marginal improvements in student outcomes.

Similar to the Task Force in 2004, our schools need to change how they operate in order to meet the complex challenges they face.

Most public school systems are set up as a reductionist hierarchy: teachers report to an assistant principal, assistant principals report to the principal, and principals report to a superintendent. Administrators devise how schedules and resources should be allocated. Teachers specialize in their roles and are expected to carry out the designs of those further up the chain.

This would make sense if educating children was an easily repeatable and complicated process, like assembling an automobile, but it’s not. Again, Team of Teams illustrates the difference between the complicated and the complex:

“The workings of a complicated device like an internal combustion engine might be confusing, but they ultimately can be broken down into a series of neat and tidy deterministic relationships; by the end, you will be able to predict with relative certainty what will happen when one part of the device is activated or altered.

Complexity, on the other hand, occurs when the number of interactions between components increases dramatically — the interdependencies that allow viruses and bank runs to spread; this is where things quickly become unpredictable. “

I’d argue that running a successful school is more like the latter description than the former. Students in great schools aren’t treated like cogs moving along an assembly line — they’re a part of several networks of relationships with peers, teachers, coaches, administrators, guidance counsellors, and more.

Unfortunately, in many public schools, teachers and administrators operate in a series of silos. This may make it easier for the science department to get better at running lab sessions, but it also makes it more difficult for staff across departments to collaborate and adapt in a fluid manner.

Seeing the shortcomings of the Task Force’s structure, Gen. McChrystal decided to act:

“We had to tear down familiar organizational structures and rebuild them along completely different lines, swapping our sturdy architecture for organic fluidity, because it was the only way to confront a rising tide of complex threats. Specifically, we restructured our force from the ground up on the principles of extremely transparent information sharing (what we now call “shared consciousness” and decentralized decision-making authority (“empowered execution”). We dissolved the barriers — the walls of our silos and the floors of our hierarchies — that once made us efficient… We abandoned many of the precepts that had helped establish our efficacy in the twentieth century, because the twenty-first century is a different game with different rules.

Mehta’s recognizes in The Allure of Order that “we need to design a new type of system, indeed to remake the sector as a whole.” He would prefer that a redesign focus on strengthening the professional practice and expertise of teachers, but Gen. McChrystal’s experience suggests that we may want to consider a different course.

What would it looks like for our schools to eschew their silos and hierarchies and become transparent, decentralized organizations? Even if we could imagine what this would look like, would it even be possible to make that transition?

Stay tuned…

Beyond Taylorism, Part I

Three books on three seemingly different topics changed how I think about public policy.

My reading list this month came together serendipitously. A tweet led me to buy The Conservative Heart, I ordered Team of Teams after listening to a podcast, and purchased The Allure of Order after reading Neerav Kingsland’s blogs about it.

At a first glance, these books don’t look like they’d fit together. The Conservative Heart is about communicating a conservative anti-poverty agenda. Team of Teams is about the lessons Gen. Stanley McChrystal learned while combating the insurgency in Iraq. The Allure of Order recounts a century of education reforms.

All three books, with their own perspectives, share an important insight: complicated problems are different than complex problems, and they require very different approaches.

What is the difference between a problem that is complicated and one that is complex? Arthur Brooks breaks it down in The Conservative Heart:

“Complicated problems are extremely difficult to understand, but they can be resolved with sufficient money and brainpower. And once you find the solution, the problem is permanently solved. You can replicate the solution over and over with a high degree of success. Designing a jet engine is a complicated problem. Figuring out how to build the first jet engine took sophisticated tools, computing ability, and expert engineers. But once engineers figured out how to do it—and designed a jet engine that worked—they could replicate the process and make jet engines routinely.

Complex problems are very different. They initially seem simpler to understand but can actually never be “solved” once and for all. One example is a football game. You know exactly what success looks like—it’s when your team wins. (In my case, it’s when the Seattle Seahawks win.) But there are so many trillions of combinations of things that can happen on the playing field, so many variables and ambiguities, that even the best data and strategies are dwarfed by the uncertainty that remains.”

At the turn of the century, businesses were grappling with complicated problems. Frederick Taylor is largely responsible for helping manufacturers treat organizational management like an engineering problem. Best practices were standardized across organizations while managers planned out the most effective ways for workers to carry out discrete tasks.

Known as Taylorism, the scientific approach to management spread far beyond the factory floor. Team of Teams describes how Taylorist principles have guided the complicated operations of militaries for centuries. In The Allure of Order, we learn how Taylor influenced several waves of reform that focused on measuring results to drive greater efficiency.

There’s just one problem: our problems (particularly in education) are increasingly complex. And Taylorism is lousy when it comes to addressing complexity.

Thankfully, we have a framework that can help us move beyond Taylorism. Team of Teams illustrates how the Joint Special Operations Task Force was able to transform how they operated in order to defeat a dynamic and decentralized enemy. More importantly, it outlines how leaders can move beyond running an efficiency-focused organization to leading one that can adapt to meet the needs of complex problems.

In future posts, I’ll dive in to the details of how this approach could help us address the problems raised in The Allure of Order. At it’s heart, it’s about replacing a top-down, meddling mindset with one of my favorite McChrystal phrases:

“Eyes on — hands off”

Deal on charters in detail

The CT Mirror reports that Governor Malloy and Democratic legislators struck a budget deal that would allow two new charter schools to open as well as supporting the growth of existing charter schools.1

On paper, this charter growth will cost the state $12.4 million in FY16, but it actually took an additional $23.5 million increase in FY16 ECS funding to make it happen. In other words, for every $1 in new funding to support growth in charter schools, legislators insisted that $2 went to support traditional public schools via ECS grants.

As I’ve noted, it’s much easier for legislators to give districts more money instead of addressing the inconsistent distribution of current ECS funding. The question I’d like to address: how well does this budget distribute the new ECS funding?

Looking at towns’ ECS funding gap compared to their increase in ECS funding by FY17, we see that there’s a trend, but it’s not very clear. Towns with larger gaps will see a larger boost in their ECS grants, but towns with similar gaps are getting very different increases:


If we take a slightly different look at the changes, this time in percentages, the distribution of new ECS funding starts to make a little more sense:


This chart shows a tighter correlation between the percentage of the fully-funded ECS grant towns currently receive and the percentage increase in their ECS grant by FY17. Towns that currently get the smallest percentage of their ECS grant target amount will receive the largest percentage increases in funding, while towns getting a higher percentage will see a smaller increase in funding.

Of course, there are exceptions.

By FY17, the ECS grant will be 77.9% funded, but 47 towns with more than 78% of their target ECS grant will see an increase in funding. This comes at the expense of towns that are still significantly under-funded compared to their ECS target. For example, Milford won’t see their ECS grant increase by a single penny, even though they get less than 40% of their target ECS grant.

To put it in perspective, this is what the distribution of ECS grant “percent funded” will look like by FY17 (the first chart has a normal x-axis, while the second uses a logarithmic scale to better illustrate the left side of the distribution):



To answer my original question, how well is this new ECS funding being spent?

The new funding will do some good for under-funded districts, but the following problems persist:

  • Over-funded towns with low levels of poverty will continue to get more than their fair share in ECS grants.
  • The ECS formula still doesn’t fully account for the impact of concentrated poverty.
  • Choice programs, including charters, magnets, and vo-tech schools, are still funded via separate line items.

Much more work remains to be done.

  1. According to another CT Mirror article, the additional funding will allow current charter schools to add approximately 700 students across the state. The new charters slated to open this fall in Stamford and Bridgeport would serve 168 and 250 students, respectively, in their first year of operation. 

Hard Choices

There are few people in Connecticut defending the current system of distributing Education Cost Sharing (ECS) grants. Sure, there’s a formula that is supposed to guide annual appropriations, but there are two big problems. To start, the legislature hasn’t ever appropriated the level of funding the formula calls for. Given the economic tumult of the past decade and Connecticut’s somewhat anemic growth rate, it’s not hard to understand why this is the case.

Less understandable is how the legislature chose to allocate ECS funding. According to testimony submitted by Connecticut Voices for Children, the line item for ECS grants is about 77% of what it ought to be if it were fully funded. As I’ve noted in a previous post, some towns get much less than what they should expect at this funding level and some get more. A lot more.


The chart above illustrates how close each town is to their targeted grant at a 77% funding level compared to the town’s level of student poverty1. It’s pretty clear that there are many towns with low levels of poverty getting much more than they should be getting in ECS funding. Here’s the same chart with a logarithmic scale on the y-axis to make it easier to see which towns we’re talking about.


Looking at this problem in real dollars shows a slightly different picture. A few towns (West Hartford, Danbury, Milford) are underfunded by millions and millions of dollars, while the big urban centers (Hartford, Bridgeport, New Haven) are getting at least $10 million more than they are supposed to receive.


I’m not too worried about the big cities getting relatively more funding at this point. Our state’s education funding formula should gives more weight than it currently does to students with additional learning needs and to towns with high concentrations of poverty. These cities have both, so I think this divergence from their adjusted ECS grant is justifiable.

It’s much more difficult to come up with a good reason for overfunding districts with low levels of poverty. In effect, our state is providing extra subsidies to school systems in some of the most affluent communities in the state at a time when many school systems are struggling with their towns to meet the needs of their students.

###What did the Appropriations Committee do to address this problem?

The Connecticut General Assembly’s Appropriations Committee attempted to address this issue in their version of the Governor’s budget bill, H.B. 6824. While the Governor’s initial budget would flat-fund ECS grants for FY16 and FY17, the Appropriations budget would direct an additional $3.7 million in FY16 and $10.6 million in FY17 to some of the most underfunded towns.

Unfortunately, all other towns were held harmless, meaning that overfunded towns will continue to enjoy their ECS windfall at the expense of districts with greater financial needs. An updated version of the chart we started with doesn’t show a very different picture by FY17.


###Making hard choices

What would happen if our legislators decided to end the overfunding of districts with lower levels of poverty? Would we be able to end some of the most egregious examples of under-funding in the current system of ECS grants?

First, we would need to decide which overfunded towns should lose their windfall. As noted earlier, I don’t think it’s wrong to overfund the big cities given the challenges they face. Instead, I’ll pick a limit of 25% Free/Reduced Price Lunch (FRPL) resident students to decide which towns should lose their ECS windfall. More than a third of Connecticut students qualify for FRPL, so I think it’s a fair cutoff to determine “lower poverty” towns.

So how many towns are overfunded and have fewer than 25% of their resident students in poverty? Ten? Twenty? Nope - the answer is 88. In other words, more than half of Connecticut’s 169 towns are lower-poverty towns getting more ECS funding than they ought to - about $59 million extra each year.

That $59 million in overfunding to towns like Greenwich, Fairfield, and Guilford could go a long way towards bringing some of the most underfunded towns closer to their ECS target grants.

Let’s say we wanted to bring every town up to at least 75% of their adjusted ECS grant2. This would allow the state to continue providing additional funding to the state’s highest-poverty districts while ensuring that no other districts were terribly far off of their adjusted ECS grant. This would cost just under $45 million, but it would still leave a few districts with ECS grants that are millions of dollars less than they should receive.



We could use some of the $14 million left over3 to ensure that no town was more than $7 million away from their adjusted ECS grant. This would impact four towns (Danbury, Hamden, Stratford, and West Hartford) and cost about $8.3 million.



This would leave $5.7 million that could be used in a variety of ways to help improve the quality of education in Connecticut’s most challenging areas instead of subsidizing towns with relatively low levels of student need.

I can understand why the Appropriations Committee chose to address towns with severely underfunded ECS grants. It’s always easier to give towns more grant money than it is to take it away. It’s easy to say that as long as ECS isn’t fully funded, that no town should see a cut in their ECS grant.

Those easy choices have consequences. The ECS formula, while flawed in many ways4, does call for state aid to be distributed according to town wealth and student need. Ignoring the formula’s prescribed distribution has placed the state in the position of benefiting pretty affluent towns at the expense of the underfunded towns. In my example above, the towns identified as underfunded had nearly a quarter of their students qualifying for Free/Reduced Price Lunch (22.6%). In the overfunded, low poverty towns, that rate is only 10.9%.

We elect leaders to make hard choices, but they’ve been taking the path of least resistance when it comes to ECS. There’s no way to get around it: the current distribution of ECS grants benefits our state’s wealthiest towns, forcing towns with more students in poverty to rely on local property taxes to make up the difference.

Every year our legislators use the underfunding of ECS as an excuse to continue this practice is another year of additional burden to the students and taxpayers in these communities. It’s time to stop taking the easy road and do what’s right for our state’s towns and students.

  1. All of the charts in this post use the size of a town’s name to represent the number of public school students that live there. That is to say, towns with more students will appear in a larger font than towns with few students. 

  2. Town’s adjusted ECS grant = Town’s fully funded ECS grant * funding ratio. In the current budget, ECS is only allocated 77% of what the formula calls for, so a town’s adjusted ECS grant would equal 77% of their fully funded ECS grant. 

  3. $59 million in overfunding - $45 million to bring every town up to at least 75% of their target grant = $14 million left over. 

  4. More on this in future blog posts, but to keep it short, I think the ECS formula should weight more heavily for poverty, provide a weight for concentrated poverty, and include all public schools. Yes, that means I think we should fund districts, magnets, and charters the same way. 

More data please

Two weeks ago, Rob Stellar wrote about the low rate of college completion among Hartford Public School graduates on TrendCT. More importantly, he raised some important questions that may better explain what’s going on here:

“What is preventing so many students who graduate high school from enrolling in college? Is it for financial reasons or lack of access to other resources/college prep programs? How well do Hartford high schools prepare students for college? For the students that do complete college, what type of degree did they obtain? What sets these students apart from their peers who do not finish college?”

Rob doesn’t presume to know the answers to these questions - he suggests that more data is necessary to understand the complexities of this problem. A reasonable suggestion to help identify specific ways to help students in Hartford complete college.

A recent CT News Junkie op-ed intended to counter Rob’s CT Trends piece by claiming that we need “no more data” on the subject, pointing to examples of how proficiency rates data and graduation standards may have been compromised in Hartford.

However, the answer to incomplete or flawed data shouldn’t be no data - it needs to be better data.

A recent article in the New York Times emphasizes the need for analysts to ask “What did I miss?” when evaluating easy-to-measure data. At the end of the article, the author notes an under-reported aspect of Bill James’ Moneyball approach to baseball analytics:

“The numerical revolution he spearheaded was never about putting traditional experts out of business. It was about acknowledging our ignorance, then gathering and testing data, whatever its source.”

This is exactly how we should be trying to understand how well our students, teachers, and schools are doing and what we can do to improve them. Let’s collect more data, test it, and use it to improve what adults are doing to improve student outcomes.

Many schools are already implementing a smart use of data to improve instructions and operations, but there’s still room to improve and include more sources of data. Student surveys, growth models, and more accessible data for researchers are just a few ways we could achieve a more robust educational dataset in Connecticut.

As we work towards building more robust datasets in the field of education, we also need to make sure we’re engaging in rigorous, transparent analysis. We also need to acknowledge the limits of what data can reveal to us. As noted statistician John Tukey observed:1

“The data may not contain the answer. The combination of some data and an aching desire for an answer does not ensure that a reasonable answer can be extracted from a given body of data.”

Let’s collect more educational data and apply it appropriately - it’s the best way to answer how we can improve the way our public schools serve students.

  1. Quote is from Jeff Leek’s The Elements of Data Analytic Style - I highly recommend this e-book to any data analyst!