By Anthony Cody.

On Sunday I spent several hours transcribing Bill Gates’ Oct. 7 speech on the subject of the Gates Foundation’s K12 education strategy. The Gates Foundation had published the text of the speech here, but I found that his actual remarks, while roughly similar, were not the same. I think it is important to understand what he actually said, so here it is, taken from the video posted here, minutes 1:22 to 1:54.

I have posted my response to this speech here. 

Bill Gates:

As Melinda said, I get to update you on our K12 education strategy. It’s not our whole education strategy, but it’s a very core piece. The idea that the kids who do make it to college get there with a sense of confidence, the right grounding, that they’ve met high standards, and that they can go ahead and not have to go into remedial classes – that’s very, very critical. This is an area that we’ve always viewed as a very key priority.

As Melinda said, we’re always trying to learn, to see what’s going to work, and as part of that, earlier this year I welcomed a guest to my office, a man named Lyon Terry, who teaches fourth grade at Lawton Elementary here in Seattle. Lyon is the 2015 Washington state teacher of the year. He’s here with us today, and I’d like him to stand up so we can honor his accomplishment (applause).

When I met Lyon, he shared a few of his secrets. Here is a good one. At the beginning of every year, he draws an arrow on a piece of paper, pointing up and to the right. He labels that “the Learning Line.” He puts a dot at the bottom, and labels it “birth.” He then puts another dot a little higher up where a fourth grader would be, and another dot higher still where a high school graduate would be. What Lyon said next was really striking. “I put a dot for myself which is a little further up the line but not at the top.” And he tells his class, “this is your life trajectory. You’ll be traveling up the learning line every day of your life.” I love a lot about this approach. What stands out for me is that he doesn’t put himself at the top. He knows that he’ll never be done learning. And I think that probably would be shared by most teachers. They always want to do better. They want to keep learning. They want to be a better teacher. They want to engage students who are bored. They want to learn how to explain fractions in a way that helps kids get it. They want to keep moving up this line, so they can help their students also move up the learning line.

One problem we see in the teaching profession is that too often, teachers have to move up completely on their own. They don’t get the feedback or tools they need to improve their practice. So they move up slowly, or not at all. That’s not only a loss for the students; it’s also frustrating for the teachers. And so the work we’re doing in K through 12, at the core of it, is the goal of helping teachers move up that learning line, faster, and in concert with their colleagues, so that every teacher is as good as they can be.

This particular focus is something we set out on seven years ago, and if I have to say where we are on this learning line, I’d say we’re not even halfway towards that goal – maybe not even a quarter of the way. But we’re here to keep moving up. I believe we are working on the right problem. Our vision is that students deserve high standards, they deserve the most effective teacher they can get, every teacher deserves the tools and support to be phenomenal, and all the students deserve to learn in a way that’s tailored to their needs. So these are the advances that we’re backing that we think will transform America’s schools, and make them back into the creator of equity that we all believe they should be.

We decided to focus on what goes on inside the classroom, and focus on the teaching profession and how we could facilitate improvement there. The evidence is very strong about the importance of an effective teacher. If you take two classrooms from within the same school, and you have a teacher in one classroom who’s in the top quartile – not at the top, but just in the top 25%, and another teacher who’s at the top of the bottom quartile, the 75%, and you look at their students’ achievement over the course of the year, their scores will be ten percent different by the end of the year. And that’s a very dramatic difference. If you go three years in a row having that top line, you would completely close the income inequity of learning in the entire country. And so making sure there’s more of those top quartile teachers, and that we’ve moved people up from that bottom quartile, that, if it’s done at scale, can have dramatic effects.

What is it that is going on with those good teachers? I was surprised that the field’s understanding of exactly what great teachers are doing was not as strong as I might have expected. There are many aspects that go into teaching. Getting the class’s attention, how you explain things, how you motivate learning what are sometimes very abstract concepts, how you deal with the fact that the kids are at very different levels, as was mentioned, that kids easily get bored, perhaps more so than ever. So we decided to invest in learning about good teaching. We decided that a goal would be to raise the learning effectiveness and reduce the number of dropouts. And if we look at this nationwide, test scores as a whole have not been going up. For example if you look at SAT, ACT, National Assessment of Educational Progress (NAEP) – particularly if you take out some segments. If you take out Asian Americans, if you take out some of the charter schools, then you can see that at best, these numbers are flat. But inside those numbers there are some points of light. There are some places where progress is being made. For example, in Kentucky, ACT scores have been going up, both in low income and high income schools for the last twenty years. The graduation rate has gone from 80% to 86%, well above the national average of 81%.

In Denver, the percentage of kids who score over 21 on the ACT has gone up by half in the past five years, from 16% to 24%. In many charter environments, including in Los Angeles at Green Dot, the schools have really high college readiness rates – over four times the District schools that have similar students. And in Washington, DC, we see students taking the fourth grade math test are outgaining their peers in every state for six years in a row.

Of course none of those places are satisfied with their results, but their progress is promising, and they all have something in common. They have invested in supporting teachers. They’ve invested in helping the teachers excel. They use multiple measures, backed by evidence. They train and certify classroom observers. They provide teachers with instructional tools aligned to the Common Core standards. And they focus their feedback and evaluation on activities that help the teacher get better in the classroom. For example, Denver uses a measure that combines teacher observation, student perception surveys, and evidence on how much the students are learning. It’s not just a system for taking the teachers and sorting them into groups. It’s a framework for moving up the learning line together. The principal visits the class, discusses it with the teacher, and decides where the teacher stands. If they’re not satisfied there are resources that can be brought in to get additional coaching from fellow teachers. And so there are clear paths for growth that are laid out.

The approach in Denver is advanced compared to what most teachers get in the rest of the United States. But most other places around the world, from China to the Netherlands, have been doing this thing for the past decade. And in that past decade, when the US was one of the leading school systems in the entire world, at teaching kids math, at teaching them to read and write – we have fallen behind. We are now in 14^th place. And almost all of those systems, that are not only above us, but used to be below us, they’ve moved ahead, a key element they have is how they help their teachers improve.

The idea here is not only to help design these systems, but also help them to be sustained, help them to constantly improve, and to help them to spread out. Our findings are grounded in the work we started in 2009, called Measures of Effective Teaching. In fact we talked about it so much, it’s got an acronym, MET. We always talk about the MET study, and it was pretty fundamental to us to really see, to videotape 13,000 different videos of classes, and to see what goes on with great teaching. The simple insight there is that the level of engagement with the kids, the level of interactivity, of drawing them in, is far higher in the classrooms where you’re getting more learning taking place. And so that was a great grounding, to say how do we build these improvement systems. The peer review and coaches in the classroom, those have proven to be really important elements.

Some people really welcome that into their classroom. But others are reluctant, when they know the system is new, they worry about the consequences that are related to it. Some people are reluctant to have this take place. In Colorado, where they’ve really persevered, and gotten the system to be a very strong system, most teachers, both by survey results and by specific discussions we’ve had, they crave the feedback. One of the coaches says, “It doesn’t matter whether teachers have taught for two years or for 25 years, they say that this kind of peer feedback is the most meaningful professional development they’ve had.”

This idea of student surveys – we now have thirty different states using that as part of the system. Students answer questions like “If I don’t understand something, my teacher tries to explain it another way,” or “My teacher doesn’t move on to the next lesson until we understand the first lesson.” Now these aren’t a perfect measure, but if they’re designed properly, they’re incredibly effective. You have a lot of students who are in there, and well-designed surveys are not only giving you a sense about the quality of teaching, they provide helpful feedback. They are diagnostic. And the perceptions of the students really do match extremely well with the learning outcomes. That was one of the things MET did – it took these different ways of evaluating, and made sure that you had very strong correlations so that as the more diagnostic efforts were taking place, those would drive student learning.

One of the elements, of course, is taking in test scores. And that, of all the evaluation elements, is perhaps the most controversial. In some ways it’s the easiest to do. It doesn’t require training peers, it doesn’t require teaching you how to understand the surveys. I personally believe they are a critical element of these systems. But they are not what tells you what skills you need to improve. I never met a teacher who said “Yeah, I got those test scores, and now I know what I need to change.” They are simply a number.

Now in this debate, you might think that it’s an all or nothing thing – that you have to just use test scores or that they shouldn’t be any part of the system, and I think that is a false choice. The role of test scores is very important. They’re very normative. They really show you how you compare and it’s a key outcome, so I’d say that for most subjects, they should be part of the mix. Today we are seeing a pretty good balance. No state uses test scores for more than 50% of the evaluation. Eight don’t use them at all, and most are somewhere in between zero and 50%.

The area that I think we need to invest the most in is this idea of classroom observation. When it’s done well, this is a very critical component. In a recent study researchers at Harvard, gave the teachers video cameras and allowed them to record as many of their lessons as they wanted. They could choose which ones they sent to the principal for him to look over and for them to discuss. And what we saw was that when you put the teachers in charge of deciding which lessons to seek feedback on, it redefines the power dynamic between them and the principal. The teachers get to lead the discussion and focus on what they think they need to improve. In most districts teachers only get two or three observations per year. But in this study, teachers voluntarily recorded thirteen lessons and sought feedback on them. And that’s a great improvement in terms of connecting up with the teachers’ desire to excel.

Another key thing we believe is that you’ve got to have high standards. Clear academic standards so that if a student does well, by the end of high school they have the skills that make them ready for college if that’s where they want to go, or the workforce, if that’s their choice. And this is where we’ve been involved with the Common Core standards. Today we have 42 states and the District of Columbia using these standards, and we view them as quite fundamental. It’s unfortunate that many of the attacks against the Common Core have not really focused on what the Common Core is and have to some degree drowned out the facts. We do need to have a system that defines excellence, and that system needs to be very thoughtfully designed, things like how do you take all your reading and writing experiences across your different classes and really see what you’re missing so that different teachers can engage on your deficits. The previous standards did not focus on that. The math standards need to be ordered in a logical fashion, and there’s been a lot that’s been learned on how to do that well, like it has in phonetics. And most state standards did not get those orderings well.

Standards are both about being well-designed and setting a very high standard. If you don’t get enough so that you end up in a remedial math class, what we’ve seen is over 70% of the kids who aspire to go to college don’t pass that entrance math exam and get stuck in a remedial math class, those 70% end up dropping out of school. And they’ve wasted money, they’ve wasted time, they have a negative self image. It’s really, in some ways, the ugliest piece of where the education system is letting kids down. Kids should know that they’ve learned enough in high school so that remedial math won’t be necessary.

Now the rollout of Common Core, there’s a lot that’s been learned about this. How we can take the time to prepare teachers. How we can help people understand where the new cutoff levels are, and those have to be articulated. There’s a lot of work about the rollout, including a mea culpa, where in some places our foundation and perhaps others were naïve about those rollouts. But in no way should that mask the value and the impact that these standards are already having. One aspect of this is that we put together what we call Literacy and Math Design Collaboratives, and those are curriculum tools aligned to the Common Core. People are using those and getting really fantastic results.

We’re also seeing innovative tools that derive from the Common Core. When we had many different state standards it wasn’t clear when you went online and you wanted to learn a piece of material how you would find the right information. Or if a teacher had a thing that explained a concept extremely well, where would that fit into the curriculum of different states. There was no taxonomy, no way to find it and put it in. Because we have a standard, the Common Core standard, it is now very clear. There is literally a place in that learning line that any innovative piece can be slotted in. In some ways this standard of what kids learn – not the curriculum, but what the knowledge is, and how that’s ordered, that’s a standard like an electrical plug. It means that people can innovate in how they bring small size pieces to help with that or large size pieces. So you go up to these new web sites, like Kahn Academy, you are going through the Common Core progression. So not only can students move from state to state, or teachers move from state to state, but all of the tools are now aligned and comparing how do they work to help a kid move down that progression. So when somebody says “yes, a state should be able to deviate,” yes, they should be able to deviate. If they want to change their electrical plug or their railroad width, it’s fine for them to do that. But they have to think about what are the tradeoffs? Are they designing something that is so much better that getting off of that entire innovation, which is now at critical mass because of the standard, is that a net benefit to them?

One of the areas these tools are driving towards that’s very exciting, and we have a specific breakout on it, is personalized learning. Particularly in math, this idea that you are always at the level where you feel challenged, and you’re using new formats, like kids learning together, using gaming software, small sessions, and the math day has always got a lot of variety, a lot of each of these different approaches to learning is used – that’s very exciting. We’re seeing just phenomenal math learning, not just ten percent improvement, but fifty percent improvement, or even more improvement, particularly for the kids who don’t think of themselves as very good at math. So that’s one of the things that the standards are enabling.

If we look at this whole movement, for driving – helping teachers move up that learning line faster, I am concerned about whether we’ll have a dramatic impact in the long run. The majority of teachers are still in systems that don’t help them improve that much. The quantity of feedback has gone up in many cases, but in most cases, the quality of the feedback isn’t good enough. And unless the right investments and seriousness is put into it, it’s not hard to see why. If you simply take a principal, and without a lot of training, and without freeing up a lot of their time, and ask him to go in and do these classroom reviews, they won’t be able to magically do that very well. It takes an investment, it takes selecting principals for this as a primary skill set that they have to have.

We have many systems today that are viewed negatively because they are mainly about hiring and firing. They are not a tool for this learning. If we don’t get that balance right, the whole evaluation system does not strengthen teaching, it actually inhibits it. So you get cases where teachers would prefer to have no feedback at all, which was the system a decade ago most of them worked in.

Every teacher has a right to ask about these evaluations, “Is it designed to help me get better?” When I hear about systems like the one in Denver, and I compare them with what most people have, you see a huge difference. And you see the student outcomes now matching with the investment that’s made in that.

So if we want to reduce inequity, we really believe that doing this well is critical. That’s going to require not only advancing the state of the art and understanding how these things work, but also spreading best practices. We need to see which districts and states are getting out ahead, and make sure others are open minded to see what’s working well there. The goal here is that over 80% of low income and minority students should graduate from high school ready for college. That seems pretty basic. It seems like an obvious thing. But where are we today? Only 25% of Hispanic students and 10% of African American students graduate from high school ready to go to college. So we have an unbelievably large gap that we need to fill. I’m confident that we can, by focusing on effective teaching and spreading those best practices. I am worried that these systems are fragile. In the past, even before our foundation worked on education, there were cases where evaluation systems would come in, but if they weren’t nurtured, if they weren’t constantly improved, they often would be diluted and eventually disappear. So we have a lot of excitement, we have teachers embracing these systems, but we also have other places where people are talking about reducing the amount of feedback, going back to the system where the teacher was really working pretty much in isolation, and that misses out on a huge opportunity.

So it hangs in the balance. Will best practices be adopted throughout the country, or will we retreat from these reforms? The future of our students really hangs in the balance. There’s no outside force, like schools going bankrupt like a business would, that automatically takes low-impact evaluation systems and turns them into high impact systems. It’s really up to all of us who are passionate about this work to make sure that happens.

We have a lot of great philanthropists who are helping push this forward. We have incredible charter schools that have shown examples of what can be done with best practices and have shown amazing student gains. A lot of the charter stuff has been advanced by federal and state policies that have gotten stronger over the last eight years. The philanthropic sector has really stepped up to play a role, as someone said not a central role, but a complementary role, to help accelerate those activities.

We believe very much in charters. They’ve been a model for learning. They are a great source of best practices. But even so, only five percent of the kids are in those schools. And although that number has a chance to grow, the vast majority of kids are going to be in the normal district schools, so that’s why we think it’s important that there’s a focus there. There’s a great number of philanthropists who have joined us in working in those areas. Broad, Bloomberg, Hyde, Kaiser, Shusterman, are among the family foundations that are involved in this work. And we hope that even more of them come in and bring their own ideas and approaches to this, because it really does help drive the experimentation, which will be the key to improving these systems.

I hope that teachers will demand systems that help them improve. In the long run that’s absolutely the only thing that will sustain these systems. Nobody speaks with more authority about what’s needed. We need to show them what’s possible, but then they’re the ones who will insist on the high quality feedback and make sure that those are kept in place. For politicians, this idea of setting high academic standards, being realistic about what the measurement is today, and how that compares up to what the student deserves – that’s critical. Also making sure that the funding and the priority for these feedback and improvement systems is done well. A governor or state chief can look at the teacher support system and see how they compare.

Three things to look at. Is it balanced? Is it using the right measures? Is it being embraced? Does it come with the data that teachers trust in terms of how that feedback is working? And finally, does it have the resources to really drive improvement? Have you taken all the elements of evaluation and professional development and evaluation and coupled those together so that not only is it helping teachers improve, but they really see that, and they become the biggest backers of the system.

This is where we’re focused. We’re very committed to this work. Over the next decade we hope to see incredible progress in this. It’s a difficult task. We won’t reach this goal overnight. But we do have powerful new tools to help us, and we have many inspiring examples to guide us. If we stay focused on this goal, we can help all teachers rise to the top together, and change the lives of millions of students. Thank you.