Ka’Ching! 2016 US Edtech Funding Totals $1 Billion

This is a repost of an article that appeared on EdSurge

Santa proved a little more parsimonious to U.S. edtech companies, which altogether raised an estimated $1.03 billion across 138 venture deals in 2016. Those tallies dipped from 2015, which saw 198 deals that totalled $1.45 billion. (Or, from a different perspective, U.S. edtech companies raised roughly 57 percent of what Snapchat did in its $1.8 billion Series F round.)

In this annual analysis, EdSurge counts all investments in technology companies whose primary purpose is to improve learning outcomes for all learners, regardless of age. This year startups that serve primarily the K-12 market raised $434 million; those targeting the postsecondary and corporate learning sector raised $593 million.

Since 2010, venture funding dollars for U.S. edtech startups have increased every consecutive year. It’s worth noting that even though 2016 marked the end of this trend, the dollar total still surpasses the years before 2015.

The downturn isn’t specific to the education industry but rather reflects a broader slowdown across all technology sectors, says Tory Patterson, managing partner at Owl Ventures. “There’s a broader shift in venture capital where there’s less exuberance companies that haven’t really nailed the business model,” he tells EdSurge.

Dealflow dips has also been felt in the health, real estate, construction and financial technology sectors. Across the globe, venture deals returned to 2014 levels, according to CB Insights. The market uncertainty has led some high-profile companies to hit pause on bigger plans. SoFi, which offers loans and other student services, pushed back plans for its initial public offering this year. Pluralsight, an online learning company that was also expected to IPO, is also on hold.

Venture-backed startups tend to swing between two spectrums, says Amit Patel, a partner at Owl Ventures. On one end are businesses “that grow aggressively but have no revenue associated. The other are those laser focused on business model and revenue. The mood is swinging towards the latter.”

Commitments to “impact” or “mission” aside, all investors—even in education—want to see returns. Often that means converting users into dollars.

“We’ve noticed VCs becoming more selective about their education investments, asking more questions about revenue growth and the leading indicators of product adoption, implementation timelines and ultimately usage,” says Jason Palmer, a general partner at New Markets Venture Partners. Unlike Instagrams and other “5-year consumer internet hits,” more investors, according to Palmer, now realize “it can take 10 or 15 years to build a sustainable education business.”

Breaking Down the Numbers

As in previous years, companies offering tools in the postsecondary and “other” categories out-raised other products. (“Other” includes a mix of products that help business professionals develop skills, are aimed at parents, or are not used in K-12 or higher-ed institutions.)

Expect this trend to continue, says Palmer, as investors come to “a greater recognition that higher education institutions adopt and implement more rapidly than K-12 [schools].” Tuition dollars may be one reason why they have adopted technologies such as student retention and predictive analytics platform. “Colleges and universities are facing financial pressures to keep students who contribute to their revenues. In K-12, you don’t have the same urgency of students as revenue drivers,” he suspects.

This year saw no mega-rounds for startups in the postsecondary sector—unlike 2015, which saw HotChalk, Udacity, Udemy, Coursera and Civitas Learning account for more than $520 million of funding. (Udemy did lead this pack in 2016 with a $60 million round.)

In fact, the biggest funding round of 2016 for a U.S.-based startup went to Age of Learning, which raised $150 million and accounts for 55 percent of the funding total for K-12 curriculum products. The Glendale, Calif.-based company is the developer of ABCmouse, a collection of online learning activities aimed for young children. First developed for the consumer and parent market, the tool is attempting to make headway into schools and classrooms.

Choosier Angels

Angel and seed level funding rounds, which signal investors’ interest in promising but unproven ideas, saw a small decline as well. The 66 deals at this stage are the lowest since 2011, although they totaled $62.5 million—roughly on par with 2014 levels.

Over the past five years, the average value of seed rounds has been increasing, from around $600K in the early years of this decade to roughly $1 million in 2015 and 2016. Discounting edtech accelerators, which typically invest $20K to $150K in startups, the 2016 seed round average actually surpasses $2 million. (We counted 28 such publicly disclosed seed rounds totaling $60.2 million)

Fewer but bigger seed deals are “a sign of maturation in the industry,” says Shauntel Poulson, a general partner at Reach Capital. Unlike previous years, where upstarts and ideas popped up the market, she believes the market is currently in a “stage of consolidation where leaders and proven ideas are emerging.”

Aspiring entrepreneurs ought to pay heed. What this means is that “the bar for seed rounds is getting higher,” Poulson adds. “Before it was about a promising idea and a great team. Now you need to show more traction and even some revenue.” Over the past few years investors have learned that “it’s best to focus on business model sooner rather than later.”

Palmer believes the days where startups could raise money before making some may be over. Expect to get grilled over “revenue growth, product adoption, implementation timelines and ultimately usage,” he says. To round out the questions, “VCs are also starting to ask about product efficacy.”

Looking Ahead

Unsurprisingly, investors held a cheery outlook for 2017, expecting funding totals to hold steady or even increase. More companies will be able to demonstrate sustainable revenue, predicts Owl Ventures’ Tory Patterson, and in turn woo investors’ appetite. “We think a lot of companies will be able to hit the $10 million revenue milestone.”

Emerging technologies such as artificial intelligence, augmented and virtual reality could drive further investments as their applications to help improve learning outcomes become clearer. Also expect to see Chinese investors paying closer attention, says Poulson. “There’s a big after-school market [in China] and an opportunity to leverage a lot of the content that’s being developed in the U.S.”

There’s also word on the street that several education-focused venture firms have re-upped their coffers with new funds to support proven, maturing startups. Stay tuned for more details.

Disclosure: Owl Ventures and Reach Capital are investors in EdSurge

NCLB is now Every Student Succeeds Act

The newest proposed version of the Elementary and Secondary Education Act—dubbed the Every Student Succeeds Act—is almost over the congressional finish line, with votes in both chambers of Congress imminent.

So how would accountability work under the ESSA, if approved? And how does it compare to No Child Left Behind Act, Classic Edition, and the Obama administration’s waivers?

Your cheat sheet here. Top-line stuff on accountability first, then some early reaction. Scroll down further if you want the nitty-gritty details on accountability.

And scroll down even further if you want more details on other aspects of the deal (an update of past Politics K-12 cheat sheets, including some new information on which programs made it into the agreement and which are on the chopping block, thanks to this helpful fact sheet from the Committee for Education Funding).

The top-line stuff: The ESSA is in many ways a U-turn from the current, much-maligned version of the ESEA law, the No Child Left Behind Act.

•States would still have to test students in reading and math in grades 3 through 8 and once in high school, and break out the data for whole schools, plus different “subgroups” of students (English-learners, students in special education, racial minorities, those in poverty.)

But beyond that, states get wide discretion in setting goals, figuring out just what to hold schools and districts accountable for, and deciding how to intervene in low-performing schools. And while tests still have to be a part of state accountability systems, states must incorporate other factors that get at students’ opportunity to learn, like school-climate and teacher engagement, or access to and success in advanced coursework.

And, in a big switch from the waivers, there would be no role for the feds whatsoever in teacher evaluation.

• States and districts will have to use locally-developed, evidence-based interventions though, in the bottom 5 percent of schools and in schools where less than two-thirds of students graduate. States must also flag for districts schools where subgroup students are chronically struggling. The School Improvement Grant program is gone, but there are resources in the bill states can use for turnarounds.

The deal goes further on accountability than either the House- or Senate-passed legislation. And, in a win for civil rights groups, it appears there are no more so-called supersubgroups. That’s a statistical technique in the waivers that allowed states to combine different categories of students for accountability purposes.

There are definitely some “guardrails” as one of the bill’s sponsors, Sen. Patty Murray, D-Wash., would say. (More on just what those are below.) But the education secretary’s authority is also very limited, especially when it comes to interfering with state decisionmaking on testing, standards, school turnarounds, and more.

So there’s some real ambiguity here. That will be something to watch going forward.

It’s still unclear just how the accountability or “guardrails” provisions of the bill vs. limits on secretarial authority dynamic will play out in regulation and implementation. But it’s possible lawyers and lobbyists may have walked away as big winners here. (Even Democratic and Republican aides see certain aspects of the bill differently.)

Put another way, there are definitely provisions in this deal that state and district leaders and civil rights advocates can cite to show that states and schools will have to continue to ensure equity. But, it will be hard for the U.S. Department of Education to implement those provisions with a very heavy hand, without at least the threat of lawsuits.

So what happens from here will be largely up to states. (More on the potential regulatory fights, and lawsuits, ahead in this story from Friday.)

“What can the secretary do and not do? I think that’s where the lawsuits will be,” said Chad Aldeman, an associate partner at Bellwether Education, who served in the U.S. Department of Education under President Barack Obama.

Early Reaction 

Civil rights groups say they’re waiting for real, live legislative language, not just a framework, before weighing in.

But, already, other accountability hawks are not happy campers.

“States are being given license to create systems that are significantly not based on student learning. That’s a problem,” said Sandy Kress, an original architect of the NCLB law. “This pretty much eliminates any kind of expectation for closing the achievement gap.” (Another take from Chad Aldeman at Bellwether Education Partner’s blog Ahead of the Heard.)

But some state chiefs say there’s no way that’s happening. After all, it didn’t under the NCLB waivers.

“I’m bothered when I hear people say that school chiefs won’t hold schools accountable,” said Brenda Cassellius, Minnesota’s education chief. “That’s not been evident with the waivers. … We’ve supported our schools and we’ve held them accountable. I hope America can see that.”

The nitty-gritty details on accountability, based on an analysis of a late-stage version of the framework:

Plans: States would still have to submit accountability plans to the education department. These new ESSA plans would start in the 2017-18 school year. And a state could get a hearing if the department turned down its plan.


  • No more expectation that states get all students to proficiency by the 2013-14 school year, as under NCLB Classic. (That ship has sailed, anyway). And no more menu of goals, largely cooked up by the department, as under the waivers.
  • Instead, states can pick their own goals, both a big long-term goal, and smaller, interim goals. These goals must address: proficiency on tests, English-language proficiency, and graduation rates.
  • Goals have to set an expectation that all groups that are furthest behind close gaps in achievement and graduation rates.

What kinds of schools will states have to focus on? 

  • States have to identify and intervene in the bottom 5 percent of performers, an idea borrowed from waivers. These schools have to be identified at least once every three years. (That’s something many states already do under waivers. And some, like Massachusetts, do it every single year.)
  • States have to identify and intervene in high schools where the graduation rate is 67 percent or less.
  • States, with districts, have to identify schools where subgroup students are struggling.

What do these accountability systems have to consider? The list of “indicators” is a little different for elementary and middle schools vs. high schools.

  • Systems for Elementary and Middle Schools:
  • States need to incorporate a jumble of five indicators into their accountability systems.
  • That includes three academic indicators: proficiency on state tests, English-language proficiency, plus some other academic factor that can be broken out by subgroup. (That could be growth on state tests, so that states would have a mix of both in their systems, as many already do under waivers.)
  • States also have to somehow figure in participation rates on state tests (schools with less than 95 percent participation are supposed to have that factored in, somehow.)
  • And, in a big new twist, states must add at least one, fifth indicator of a very different kind into the mix. Possibilities include: student engagement, educator engagement, access to and completion of advanced coursework, post-secondary readiness, school climate/safety, or whatever else the state thinks makes sense. Importantly, though, this indicator has to be disaggregated by subgroup. States are already experimenting with these kinds of indicators under the waivers, especially a cadre of districts in California (the CORE districts). Still, this is new territory when it comes to accountability.
  • Systems for high schools:
  • Basically the same set of indicators, except that graduation rates have to be part of the mix.
  • So to recap, that means for high schools: proficiency on tests, English-language proficiency, graduation rates, plus some other indicator that focuses a little more on whether students have the opportunity to learn, or are ready for post-secondary work. And also, test participation has to be incorporated in some way.

How much do each of these indicators have to count? It depends on who you ask. Everyone agrees that those academic indicators (tests, grad rates, English-language proficiency) have to weigh more, as a group, than that non-traditional indicator that gets at a students’ opportunity to learn (school climate, etc.)

From there, Democratic and Republicans aides have different takes.  A Republican aide said the academic stuff just has to be at least 51 percent of the system, and the other factor, or factors, can be up to 49 percent. A Democratic aide said the regulations might turn out differently, when all’s said and done. (In this aide’s view, the department could set a range for each individual indicator, ultimately giving the academic factors as a group significantly greater weight than the other factors.) More here. It’s also unclear whether the test participation indicator, which states can weigh however they want, will throw a monkey wrench into all of this. More here.

How do interventions work? 

  • For the bottom 5 percent of schools and for high schools with really high dropout rates:
  • Districts work with teachers and school staff to come up with an evidence-based plan.
  • States monitor the turnaround effort.
  • If schools continue to founder for years (no more than four) the state is supposed to step in with its own plan. That means states could take over the school if they wanted, or fire the principal, or turn the school into a charter, just like they do under NCLB waivers now. (But, importantly, unlike under waivers, there aren’t any musts—states get to decide what kind of action to take.)
  • Districts could also allow for public school choice out of seriously low-performing schools, but they have to give priority to the students who need it most.
  • For schools where subgroups students are struggling:
  • These schools  have to come up with an evidence-based plan to help the particular group of kids who are falling behind. For example, a school that’s having trouble with students in special education could decide to try out a new curriculum with evidence to back it up and hire a very experienced coach to help train teachers on it.
  • Districts monitor these plans. If the school continues to fall short, the district steps in. The district decides just when that kind of action is necessary, though; there’s no specified timeline in the deal.
  • Importantly, there’s also a provision in the deal calling for a “comprehensive improvement plan.” States and districts to take more-aggressive action in schools where subgroups are chronically underperforming, despite local interventions. Their performance has to look really bad though, as bad as the performance of students in the bottom 5 percent of schools over time.

What kind of resources are there for these interventions? The School Improvement Grant program, which is funded at around $500 million currently, has been consolidated into the bigger Title I pot, which helps districts educate students in poverty. But states would be able to set aside up to 7 percent of all their Title I funds for school turnarounds, up from 4 percent in current law. (That would give states virtually the same amount of resources for school improvement as they get now, through SIG.) However, the bulk of those dollars would be sent out to districts for “innovation”, which could include turnarounds.  It would be up to states whether to send that money out by formula, to everyone, or competitively, as they do now with SIG dollars. (More in this cheat sheet from AASA, the School Administrator’s Association, which has been updated on this issue.) Bottom line: There are resources in the bill for school turnarounds. But some of the money could also be used for other purposes, if that’s what districts and states want. 

What about the tests? The testing schedule would be the same as under NCLB. But in a twist, a handful of states could apply to try out local tests, with the permission of the U.S. Department of Education. And importantly, these local tests aren’t supposed to be used forever—the point is for districts to experiment with new forms of assessment (as New Hampshire is doing with performance tasks) that could eventually go statewide and be used by everyone. That way states don’t get stuck with the same old assessment for years on end.

What’s more, the framework allows for the use of local, nationally-recognized tests at the high school level, with state permission. So a district could, in theory, use the SAT or ACT as its high school test, instead of the traditional state exam.

Also, computer adaptive testing would be easier. More here.

What about that supersubgroup thing mentioned higher up? Supersubgroups are a statistical technique used in the waivers that call for states to combine different groups of students (say, students in special education, English-language learners, and minorities) for accountability purposes. By my reading of the bill, it would seem that’s a no-no. States now have to consider accountability for each subgroup separately. States liked the flexibility of supersubgroups. But former Rep. George Miller, D-Calif., and civil rights groups said they masked gaps. The deal appears to eliminate the use of supersubgroups.

What about the rest of the bill?
Scroll down for information on English-Language Learners, students in special education, school choice, teachers, and funding provisions.

English-Language Learners

Where does deal land when it comes to when newly arrived English-language learners must be tested? (Background on this issue here). States would have two choices.

  • Option A) Include English-language learners’ test scores after they have been in the country a year, just like under current law.
  • Option B) During the first year, test scores wouldn’t count towards a school’s rating, but ELLs would need to take both of the assessments, and publicly report the results. (That’s a switch from current law. Right now, they only need to take math in the first year). In the second year, the state would have to incorporate ELLs’ results for both reading and math, using some measure of growth. And in their third year in the country, the proficiency scores of newly arrived ELLs are treated just like any other students’. (Sound familiar? It’s very similar to the waiver Florida received.)

The compromise would shift accountability for English-language learners from Title III (the English-language acquistion section of the ESEA) to Title I (where everyone else’s accountability is). The idea is to make accountability for those students a priority.

Students in Special Education

The legislation mirrors a recent federal regulation when it comes to assessments for students in special education, saying, essentially, that only 1 percent of students overall can be given alternative tests. (That’s about 10 percent of students in special education.)


The bill largely sticks with the Senate language, which would allow states to create their own testing opt-out laws (as Oregon has). But it would maintain the federal requirement for 95 percent participation in tests. However, unlike under the NCLB law, in which schools with lower-than-95 percent participation rates were automatically seen as failures, local districts and states would get to decide what should happen in schools that miss targets. States would have to take low testing participation into consideration in their accountability systems. Just how to do that would be up to them.

For a deeper look at this particular topic, check out this blog post on opt-outs in the ESEA reauthorization deal.

On Programs

There’s more consolidation of federal education in the compromise than there was in the Senate bill.

  • The legislation creates a $1.6 billion block grant that consolidates a bunch of programs, including some involving physical education, Advanced Placement, school counseling, and education technology. (Some of these programs haven’t federal funding in years.)
  • Districts that get more than $30,000 will have to spend at least 20 percent of their funding on at least one activity that helps students become well-rounded, and another 20 percent on at least one activity that helps kids be safe and healthy. And part of the money could be spent on technology. (But no more than 15 percent can go to technology infastructure.)
  • Some programs would live on as separate line items, including the 21st Century Community schools program, which pays for after-school programs and has a lot support on both sides of the aisle in Congress.
  • Other survivors: Promise Neighborhoods, and a full-service community schools program. And there’s a standalone program for parent engagement. There are also reservations for Arts Education, gifted education, and Ready to Learn television.

Sen. Patty Murray, D-Wash. got the early-childhood investment she wanted—the bill enshrines an existing program “Preschool Development Grants” in law, and focuses it on program coordination, quality, and broadening access to early childhood education. But the program would be housed at the Department of Health and Human Services, not the Education Department as some Democrats had initially hoped. The Education Department would jointly administer the program, however. (The reason: HHS already has some early-education programs, like Head Start. Expanding the education department’s portfolio was a big no-no for conservatives.)

That new research and innovation program that some folks were describing as sort of a next-generation “Investing in Innovation” program made it into the bill. (Sens. Orrin Hatch, R-Utah, and Michael Bennet, D-Colo., are big fans, as is the administration.)

On School Choice

No Title I portability: That means that federal funds won’t be able to follow the child to the school of their choice.

But the bill does include a pilot project allowing districts to try out a weighted student funding formula, which would also essentially function as a backpack of funds for kids. The program would allow 50 districts to combine state, local, and federal funds into a single pot that could follow a child to the school of their choice. It is said to be a more workable alternative to Title I portability, which looked more dramatic on paper, but which few states would likely have taken advantage of because of its complexity, experts said. Importantly with this pilot, participation would be entirely up to district officials. And the language would give them a chance to better target funds to individual school needs.


The headline here is that states would no longer have to do teacher evaluation through student outcomes, as they did under waivers.And NCLB’s “highly qualified teacher” requirement would be officially a thing of the past.

There’s also language allowing for continued spending on the Teacher Incentive Fund—now called the Teacher and School Leader Innovation Program—which doles out grants to districts that want to try out performance pay and other teacher quality improvement measures. And there are resources for helping train teachers on literacy and STEM. Much more from Teacher Beat.

Funding and Other Issues

No changes to the Title I funding formula along the lines of what the Senate passed that would steer a greater share of the funds to districts with high concentrations of students in poverty. But there were some changes to the Title II formula (which funds teacher quality) that would be a boon to rural states.

The agreement would keep in place maintenance of effort, a wonky issue we wrote about recently, with some new flexibility added for states. (Quick tutorial: Maintenance of effort basically requires states to keep up their own spending at a particular level in order to tap federal funds.)

There was some chatter that the bill would also incorporate changes to the Family Educational Rights and Privacy Act. That’s not part of the agreement.

The framework would only “authorize” ESEA for four more years, as opposed to the typical five. That gives lawmakers a chance to revisit the policy under the next president, should they choose to do so. And its overall authorization levels are largely consistent with the most recent budget deal. 


For a full read of the act…. Every Student Succeeds Act 2015


Numbers Can Lie: What TIMSS and PISA Truly Tell Us, if Anything?

“America’s Woeful Public Schools: TIMSS Sheds Light on the Need for Systemic Reform”[1]

“Competitors Still Beat U.S. in Tests”[2]

“U.S. students continue to trail Asian students in math, reading, science”[3]

These are a few of the thousands of headlines generated by the release of the 2011 TIMSS and PIRLS results today. Although the results are hardly surprising or news worthy, judging from the headlines, we can expect another global wave of handwringing, soul searching, and calls for reform. But before we do, we should ask how meaningful these scores and rankings are.

“Numbers don’t lie,” many may say but what truth do they tell? Look at the following numbers:

Table 1: Scores and Attitudes of 8th Graders in TIMSS 2011

Country Math Scores Confidence (%) (4th Grade) Value Math (%)
Korea 613 03 (11) 14
Singapore 611 14 (21) 43
Chinese Taipei 609 07 (20) 13
Hong Kong 586 07 (24) 26
Japan 570 02 (09) 13
United States 509 24 (40) 51
England 507 16 (33) 48
Australia 505 17 (38) 46

These are the scores of 8th graders and percentage of them saying they are confident in math and value math. Top scoring Korea has only 3% of students feeling confident in their math and 14% valuing math, in contrast is Australia with much lower scores but significantly higher percentage of students feeling confident in math and valuing math. In fact, the top 5 East Asian countries in math scores have way fewer students reporting confidence in math and valuing math than the U.S., England, and Australia, all scored significantly lower.

It gives me a headache to understand these numbers: Do they mean that even if the Korean students do not think math is important, they study it anyway? and they have a very effective education that can make people who do not value math to be outstanding in it? Or since these are 8th graders, do they mean that after learning math for 8 years, the students feel the math they have been learning is not important in life? In the case of the United States, do they mean that American students value math but have poor math learning experiences that lead to low math achievement? Or could it be that their 8 years of math learning convinced them, at least a much larger proportion than in Korea, that math is important?

The same questions can be asked about confidence. Do the numbers mean that Korean students lack of confidence makes them study harder so they achieve better in math than their American or Australian counterparts? Or could they mean that the way math is taught in Korea made them lose confidence in math?

The data show that as students progress toward higher grades, they become less confident in their math learning. More fourth graders than eighth graders have confidence in math, for example.  Does this mean the more they learn, the less confident they become?

Or perhaps these numbers are not related at all. But the TIMSS report suggests that within countries students with higher scores are more likely to have a more positive attitude towards math, that is, a positive correlation. A negative correlation is found between countries and has been a pattern as Tom Loveless discovered in previous TIMSS. So somehow math scores, attitudes, and confidence are related. Perhaps whatever in an education system or culture that boosts math scores leads to less positive attitude and lower confidence at the same time. In this case, one needs to ask what is more important: scores, or confidence and positive attitude?

There can be other interpretations but whatever the interpretation is, these numbers show that results of TIMSS, or other international assessments such as the PISA, are a lot more complex than what the headlines attempt to suggest: Asians are great, America sucks, so do Australia and England. The TIMSS and PISA scores are perhaps worth much less than politicians and the media make of them, as the rest of this paper shows.

The Numbers Don’t Lie: A Long History of Bad Performance on International Tests

According to historical data, American education has always been bad and actually improving over the years. In the 1960s, when the First International Mathematics Study (FIMS) and the First International Science Study (FISS)[4] was conducted, U.S. students ranked bottom in virtually all categories:

11th out of 12 (8th grade -13 year old math)

12th out 12 (12th grade math for math students)

10th out 12 (12th grade math for non-math students)

7th out 19 (14 year-old science)

14th out of 19 (12th grade science)

In the 1980s, when the Second International Mathematics Study (SIMS) and Second International Science Study (SISS)[5] were conducted, U.S. students inched up a little bit, but not much:

10th out of 20 (8th grade-Arithmetic)

12th out of 20 (8th grade-Algebra)

16th out of 20 (8th grade-Geometry)

18th out of 20 (8th grade-Measurement)

8th out of 20 (8th grade-Statistics)

12th out of 15 (12th grade-Number Systems)

14th out of 15 (12th grade-Algebra)

12th out of 15 (12th grade-Geometry)

12th out of 15 (12th grade-Calculus)

14th out of 17 (14 year-old Science)

14th out of 14 (12th grade-Biology)

122h out of 14 (12th grade-Chemistry)

10th out of 14 (12th grade-Physics)

In the 1990s, in the Third International Mathematics and Science Study (TIMSS)[6], American test performance was not the best but again improved:

28th out 41 (but only 20 countries performed significantly better) (8th grade math)

17th out 41 (but only 9 countries performed significantly better) (8th grade science)

In 2003, in TIMSS[7] (now changed into Trends in International Mathematics and Science Study), U.S. students were not great, but again improved:

15th out of 45 (only 9 countries significantly better) (8th grade math)

9th out of 45 (only 7 countries significantly better) (8th grade science)

In 2007, U.S. improved again in TIMMS[8], although still not the top ranking country:

9th out of 47 (only 5 countries significant better) (8th grade math)

10th out of 47 (only 8 countries significantly better) (8th grade science)

Over the half century, American students performance in international math and science tests has improved from the bottom to above international average. The following figure shows the upward trend of American students’ performance in math. Because 8th grade seems to be the only group that has been tested every time since the 1960s, the graph only includes data for 8th grade math[9].


All the studies mentioned above have been coordinated by the International Association for the Evaluation of Educational Achievement (IEA). There is another international study, one that has gained more momentum and popularity than the ones organized by IEA. This is the Programme for International Student Assessment, better known as PISA, organized by the Organisation for Economic Co-operation and Development (OECD). PISA was first introduced in 2000 and tests 15 year olds in math, literacy, and science. It is conducted every three years. Because PISA is fairly new, so there is not a clear trend to show whether the U.S. is doing better or worse, but it is clear that U.S. students are not among the best[10]:

PISA Reading Literacy

15th out of 30 countries in 2000

17th out of 77 countries in 2009


24th out of 29 countries in 2003

31st out of 74 countries in 2009

PISA Sciences

21st out of 30 countries in 2003

23rd out of 74 countries in 2009

There are other studies and statistics, but this long list should be sufficient to prove that American students have been awful test takers for over half a century. Some has taken this mean American education has been awful in comparison to others. This interpretation has been common and backed up by media reports, scholarly books, and documentary films, for example:

1950s-1960s: Worse than the Soviet Union (1958, Life Magazine cover story Crisis in Education)[11]

1980s-1990s: Worse than Japan and others (A Nation at Risk[12], Learning Gap: Why Our Schools Are Failing And What We Can Learn From Japanese And Chinese Education[13])

2000s–: Worse than China and India (2 Million Minutes[14] (documentary film) Surpassing Shanghai)[15]

The Numbers Don’t Lie, but What Truth Do They Tell

Numbers can be used to tell stories of the past or the future. We can ask how we arrived at a certain number or what it means for the future. Asking about its past invites us to consider what we did or did not do to achieve a certain state indicated by the number. Asking about its future implications forces us to question if a certain number is desirable or meaningful. The latter must precede the former because unless the state measured by certain numbers has truly significant implications for a desirable future, the question about how we got there is practically a waste of time.

In the case of statistics from international educational assessments, the question about the future has rarely been explored. It has been assumed that these numbers indicate nations’ capacity to build a better future. And thus we must dive in urgently to learn about why others are getting better numbers than us. This assumption, however, may be wrong.

The Numbers’ Future

“Our future depends on the strength of our education system. But that system is crumbling,” reads a full-page ad in the New York Times. Dominating the ad is a graphic that shows “national security,” “jobs,” and the “economy” resting upon a cracking base of education. This ad is part of the “innovative, multitactical” Don’t Forget Ed campaign the College Board sponsored.

It is apparent America’s national security, jobs, and economy has been resting upon a base that has been crumbling and cracking for over half a century, according to the numbers. So one would logically expect the U.S. to have fallen through the cracks and hit rock bottom in national security, jobs, and economy by now. But facts seem to suggest otherwise:

The Soviet Union, America’s archrival in national security during the Cold War, which supposedly had better education than the U.S., disappeared and the U.S. remains the dominant military power in the world.

Japan, which was expected to take over the U.S. because of its superior education in the 1980s, has lost its #2 status in terms of size of economy. Its GDP is about 1/3 of America’s. Its per capita GDP is about $10,000 less than that in the U.S.

The U.S. is the 6th wealthiest country in the world in 2011 in terms of per capita GDP[16]. It is still the largest economy in the world.

The U.S. ranked 5th out of 142 countries in Global Competitiveness in 2012 and 4th in 2011[17].

The U.S. ranked 2nd out 82 countries in Global Creativity, behind only Sweden[18] in 2011.

The U.S. ranked 1st in the number of patents filled or granted by major international patent offices in 2008, with 14,399 filings, compared to 473 filings from China[19], which supposedly has a superior education[20].

Obviously America’s poor education told by the numbers has not ruined its national security and economy. These numbers have failed to tell the story of the future.

The Numbers’ Past

The past stories of numbers lie the lessons to be learned. The problem is that there are different ways to achieve the same number, although a set of factors have been identified to explain why American students perform worse than other countries or what made some other countries achieve better numbers. As a result, the most of the factors become debatable and debated myths, half-truths, or “duh!”

Time. American students spend less time studying. President Obama noted that on average U.S. students attend class about a month less than children in other advanced countries[21] in 2010. His Secretary of Education Arne Duncan said students in China and India attend school 25 to 30 percent longer than in the U.S.[22] A 1994 report of the National Education Commission on Time and Learning established by U.S. Congress observed “Students in other post-industrial democracies receive twice as much instruction in core academic areas during high school.”[23] However a study by the Center for Public Education says “students in China and India are not required to spend more time in school than most U.S. students.”[24]

Engagement and Commitment. American students, schools, parents, and governments don’t take school-based learning as seriously as other top performing countries. Not only students in other countries spend more time in school, “the formidable learning advantage Japanese and German schools provide to their students is complemented by equally impressive out-of-school learning,” noted the National Education Commission on Time and Learning in 1994[25]. “Compared with other societies, young people in Shanghai may be much more immersed in learning in the broadest sense of the term. The logical conclusion is that they learn more…” writes an OECD report explaining Shanghai’s outstanding PISA scores[26]. But the same report immediately notes “what they learn and how they learn are subjects of constant debate.”

Curriculum, Standards, Gateways, and Tests. The U.S. does not have a better (more focus, rigor, and coherence) common curriculum with high standards across the nation and an instructional system with clearly marked transition points. “…standards in the best-performing nations share the following three characteristics [focus, rigor, and coherence] that are not commonly found in U.S. standards,” says a report that calls for international benchmarking by the National Governors’ Association[27].  “Virtually all high-performing countries have a system of gateways marking the key transition points…At each of these major gateways, there is some form of external national assessment,” writes Marc Tucker in Surpassing Shanghai: An Agenda for American Education Built on the World’s Leading Systems (Tucker, 2011, p. 174). But Ontario, a top PISA performer does not, admits Tucker and schools in Finland, a much admired high performer on the PISA, “is a “standardized testing-free zone,”[28] writes Diane Ravitch.

Teachers and Teacher Education. American teachers are not as smart to begin with and are less well prepared than their counterparts in high performing countries. For example, while 100% of teachers in top performing countries –Singapore, Finland and South Korea — are recruited form the top third college graduates, only 23% are from the top third in the U.S., according to a study by the consulting firm McKinsey & Co[29].  Teachers in these top performing countries are also better trained, supported, and motivated before, during, and after taking the teaching job. This is one of the “duhs.”

Inequity and poverty. There is more social economic disparity among U.S. students and higher levels of poverty in the U.S. than other countries. “U.S. students in schools with 10% or less poverty are number one country in the world,” says a report of the National Association of Secondary School Principals[30]. The report establishes a direct connection between PISA performance and poverty and says the U.S. has the largest number of students living in poverty. But others disagree. “The U.S. looks about average compared with other wealthy nations on most measures of family background,” says the report from the National Governor’s Association, “Moreover, America’s most affluent15-year-olds ranked only 23rd in math and 17th in science on the 2006 PISA assessment when compared with affluent students in other industrialized nations.”[31]

There are of course other suggestions from access to natural resources[32] to cultural homogeneity and from sampling bias to parenting styles. Regardless, how each country achieved their international scores is not nearly as straightforward as the numbers themselves, making international learning a very difficult task.

The task becomes perhaps even more difficult, when the issues of economic, cultural, societal, and political contexts are considered. What’s more, learning from others may become not so desirable for the U.S. considering the fact that the test scores have not significantly affected America’s national security and economy. Moreover in the final analysis, since countries that have shown better numbers in tests have not performed necessarily better than the U.S., the U.S. education may have something to offer others.

The Numbers Don’t Lie, but Some Are Missing: Two Paradigms of Education

The fact the U.S. as a nation is still standing despite of its abysmal standing on international academic tests for over half a century begs two questions:

Is education as important to a nation’s national security and economy as important as believed?

If it is, are the numbers telling the truth about the quality of education in the U.S. and other nations?

If the answer to the first question is “no,” we need to disconnect the automatic association between test scores and education. In other words, the numbers don’t really measure education, at least not the entire picture of the education needed to produce citizens to build strong and prosperous economies.

In my latest book World Class Learners: Educating Creative and Entrepreneurial Students[33], I identified two paradigms of education: employee-oriented and entrepreneur-oriented.  The employee-oriented paradigm aims to transmit a prescribed set of content (the curriculum and standards) deemed to be useful for future life by external authorities, while the entrepreneur-oriented aims to cultivate individual talents and enhance individual strengths. The employee-oriented paradigm produces homogenous, compliant, and standardized workers for mass employment while the entrepreneurial-oriented education encourages individuality, diversity, and creativity.

Although in general, all mainstream education systems in the world currently follows the employee-oriented paradigm, some may not be as effectively and successfully as others. The international test scores may be an indicator of how successful and effective the employee-oriented education has been executed. In other words, these numbers are measures of how successful the prescribed content has been transmitted to all students. But the prescribed content does not have much to do with an already industrialized country such as the U.S., whose economy relies on innovation, creativity, and entrepreneurship. As a result, although American schools have not been as effective and successful in transmitting knowledge as the test scores indicate, they have somehow produced more creative entrepreneurs, who have kept the country’s economy going. Moreover, it is possible that on the way to produce those high test scores, other education systems may have discouraged the cultivation of the creative and entrepreneurial spirit and capacity.

Unfortunately there are few numbers that directly provide the same kind of comparison as TIMSS and PISA on measures of creativity and entrepreneurship, making it difficult to forcefully prove that American education indeed produce more creative and entrepreneurial talents. A piece of data I have found from the Global Entrepreneurship Monitor study suggests a significant negative relationship between PISA performance and indicators of entrepreneurship. The Global Entrepreneurship Monitor, or GEM, is an annual assessment of entrepreneurial activities, aspirations, and attitudes of individuals in more than 50 countries. Initiated in 1999, about the same time that PISA began, GEM has become the world’s largest entrepreneurship study. Thirty-nine countries that participated in the 2011 GEM also participated in the 2009 PISA, and 23 out of the 54 countries in GEM are considered “innovation-driven” economies, which means developed countries.

Comparing the two sets of data shows clearly countries that score high on PISA do not have levels of entrepreneurship that match their stellar scores. More importantly, it seems that countries with higher PISA scores have fewer people confident in their entrepreneurial capabilities. Out of the innovation-driven economies, Singapore, South Korea, Taiwan, and Japan are among the best PISA performers, but their scores on the measure of perceived capabilities or confidence in one’s ability to start a new business are the lowest. The correlation coefficients between scores on the 2009 PISA in math, reading, and science and 2011 GEM in “perceived entrepreneurial capability” in the 23 developed countries are all statistically significant[34].

Anecdotally, Vivek Wadhwa, president of Academics and Innovation at Singularity University, Fellow at Stanford Law School and Director of Research at Pratt School of Engineering at Duke University, wrote in Business Week in response to the latest PISA rankings:

The independence and social skills American children develop give them a huge advantage when they join the workforce. They learn to experiment, challenge norms, and take risks. They can think for themselves, and they can innovate. This is why America remains the world leader in innovation; why Chinese and Indians invest their life savings to send their children to expensive U.S. schools when they can. India and China are changing, and as the next generations of students become like American ones, they too are beginning to innovate. So far, their education systems have held them back.[35]

But there again are no numbers to prove these. However, other countries, particularly the high scoring Asian countries have all been reforming their education systems to be more like that in the U.S., as I have discussed in my book Catching Up or Leading the Way: American Education in the Age of Globalization[36].


I have put forth a lot of numbers of different sorts from a variety of sources. Taken together, these numbers suggest to me the following:

So far all international test scores measure the extent to which an education system effectively transmits prescribed content.

In this regard, the U.S. education system is a failure and has been one for a long time.

But the successful transmission of prescribed content contributes little to economies that require creative and entrepreneurial individual talents and in fact can damage the creative and entrepreneurial spirit. Thus high test scores of a nation can come at the cost of entrepreneurial and creative capacity.

While the U.S. has failed to produce homogenous, compliant, and standardized employees, it has preserved a certain level of creativity and entrepreneurship. In other words, while the U.S. is still pursuing an employee-oriented education model, it is much less successful in stifling creativity and suppressing entrepreneurship.

The U.S. success in creativity and entrepreneurship is merely an accidental by product of a less successful employee-oriented education, which is far from sufficient to meet the coming challenges brought about by globalization and technological changes. Thus in a sense, the U.S. education is in turmoil, inadequate, and obsolete, but it has to move toward more entrepreneur-oriented instead of more employee-oriented.

[1] http://dropoutnation.net/2012/12/11/americas-woeful-public-schools-timms-sheds-light-on-the-need-for-systemic-reform/

[2] http://online.wsj.com/article/SB10001424127887324339204578171753215198868.html

[3] http://www.washingtonpost.com/local/education/us-students-continue-to-trail-asian-students-in-math-reading-science/2012/12/10/4c95be68-40b9-11e2-ae43-cf491b837f7b_story.html

[4] Data source: U.S. National Center for Educational Statistics: http://nces.ed.gov/pubs92/92011.pdf

[5] Data source: U.S. National Center for Educational Statistics: http://nces.ed.gov/pubs92/92011.pdf

[6] Data source: U.S. National Center for Educational Statistics: http://nces.ed.gov/pubs99/1999081.pdf

[7] Data source: U.S. National Center for Educational Statistics: http://nces.ed.gov/timss/results03.asp

[8] http://nces.ed.gov/timss/results07.asp

[9] Since SIMS scores were reported in sub domains, I chose the lowest performance area for the U.S. students: Measurement.

[10] Data source: http://www.oecd.org/pisa/

[11] http://goo.gl/pAgnQ

[12] http://datacenter.spps.org/uploads/SOTW_A_Nation_at_Risk_1983.pdf

[13] http://books.google.com/books/about/Learning_Gap.html?id=HIfBn5W6LMcC

[14] http://www.2mminutes.com/

[15] http://www.amazon.com/Surpassing-Shanghai-American-Education-Leading/dp/1612501036

[16] Data source: International Monetary Fund: http://goo.gl/r7SFQ

[17] http://www3.weforum.org/docs/WEF_GCR_Report_2011-12.pdf

[18] Data source: http://www.theatlanticcities.com/jobs-and-economy/2011/10/global-creativity-index/229/

[19] Data Source: Chinese Innovation is a Paper Tiger http://online.wsj.com/article/SB10001424053111904800304576472034085730262.html?mod=googlenews_wsj

[20] Students from Shanghai China scored 1st on the PISA in all three subjects (math, reading, and sciences) in the last round of PISA released in 2010.

[21] http://today.msnbc.msn.com/id/39378576/ns/today-parenting/#.UEPqb2ie7sc

[22] http://www.centerforpubliceducation.org/Main-Menu/Organizing-a-school/Time-in-school-How-does-the-US-compare

[23] http://www2.ed.gov/pubs/PrisonersOfTime/Lessons.html

[24] http://www.centerforpubliceducation.org/Main-Menu/Organizing-a-school/Time-in-school-How-does-the-US-compare

[25] http://www2.ed.gov/pubs/PrisonersOfTime/Lessons.html

[26] http://www.oecd.org/countries/hongkongchina/46581016.pdf

[27] http://www.corestandards.org/assets/0812BENCHMARKING.pdf

[28] http://www.nybooks.com/articles/archives/2012/mar/08/schools-we-can-envy/?pagination=false

[29] http://mckinseyonsociety.com/closing-the-talent-gap/

[30] http://nasspblogs.org/principaldifference/2010/12/pisa_its_poverty_not_stupid_1.html

[31] http://www.corestandards.org/assets/0812BENCHMARKING.pdf

[32] http://www.oecd.org/education/preschoolandschool/programmeforinternationalstudentassessmentpisa/49881940.pdf

[33] http://zhaolearning.com/world-class-learners-my-new-book/

[34] http://zhaolearning.com/2012/08/16/doublethink-the-creativity-testing-conflict/

[35] http://www.businessweek.com/technology/content/jan2011/tc20110112_006501.htm

[36] http://zhaolearning.com/2009/11/14/3/

This is a repost from the blog of Dr. Yong Zhao.