The international PISA assessment has once again raised questions about the quality of school education: reading and math scores keep falling, and the decline is especially pronounced in a number of European countries. The explanations on offer include the aftermath of the pandemic, smartphones, artificial intelligence, and waning motivation to learn. But the problem may not lie only with the students: schools themselves are struggling to keep pace with a changing world. T-invariant spoke with teachers from several countries about what lies behind the test results and why successful education systems are trying to combine traditional academic training with new approaches.
Top news on scientists’ work and experiences during the war, along with videos and infographics — subscribe to the T-invariant Telegram channel to stay updated.
PISA is an international assessment of the educational achievement of 15-year-old students that the Organisation for Economic Co-operation and Development (OECD) has conducted every three years since 2000. It covers not only OECD member countries but other countries and economies as well. In 2025, 760,000 students from 91 countries and economies took the test. Russia has stayed out of the assessment since 2022; Russian students last took the test in 2018.
PISA results have been a cause for concern for several years now. Reading literacy held steady through 2015, before entering a statistically significant and accelerating downward trend starting in 2018.
Not Just Another School Test
Still, one might ask: what’s so alarming about these results? Back before smartphones, math teachers used to warn their students, “You won’t have a calculator handy in real life — learn your times tables!” But that generation has grown up, and now everyone really does have a calculator at hand. And along with it comes an obliging Claude that will not only tell you what two times two is, but also work out percentages, run the numbers through any formula, and, while it’s at it, tell you how much tile you need to lay a nice pattern in your kitchen. Do the results of yet another test really matter that much, if it may be measuring not so much students’ knowledge as their ability to handle a particular type of task?
But PISA isn’t just a math quiz, or even the Unified State Exam [Russia’s standardized national school-leaving exam, which also serves as the university entrance exam — T-invariant]. The assessment measures not so much knowledge as the ability to apply it. Its goal is to find out how well prepared teenagers are to tackle everyday problems. For example, can they figure out which car is cheaper to rent under given conditions, grasp the meaning and purpose of a text, and judge the reliability of information and the expertise of its author?
The test lasts two hours and includes tasks in science, mathematics, and reading. Results in each domain are scored separately. In 2025, special emphasis was placed on students’ ability to understand and evaluate scientific evidence and to use scientific knowledge to make decisions. Meanwhile, the share of math and reading tasks was somewhat reduced.
PISA’s methodology is designed to make the sample of students as representative as possible. Schools are selected for testing at random, taking into account their location, public or private status, and other characteristics. This keeps the sample from being skewed toward schools that offer either especially strong or especially weak education. When the data are processed, the results of students from different types of schools are assigned statistical weights so that the final figures reflect the situation in the country as a whole. The students themselves are also selected at random. Participants can decline to take the test, and in certain cases a school may exclude a student from the sample. However, the share of such exclusions is capped, and each one requires a justification — for example, insufficient proficiency in the language of the test.
The language barrier has become a particularly visible problem in Europe amid rising migration. Mikhail Gantman, a physics and chemistry teacher in Germany, notes that many immigrant children don’t manage to master the language of their new country well enough: “In recent years, the number of families in Germany where the children don’t speak German at home has grown dramatically. And the German school system turned out to be not quite ready for this. Ukrainian kids were supposed to study German for two years and then move into regular German classes. And I have to say, by no means all of them learned the language in two years. I’m not sure PISA sees or distinguishes things like that.”
Since only a limited share of students can be excluded from the sample, the test takers likely included children who were not native speakers of the test language.
Still, PISA aims to give a picture of how well prepared students are across an entire country. The tasks simulate real-life situations and test how well students can apply what they have learned. These skills matter not only for the students themselves but for society as a whole. For example, the ability to analyze texts critically is especially valuable in the age of social media, where people spend several hours a day and get a large share of their information. Based on that information, they make important decisions: they choose a country to live in, start a business or take a job, invest, and vote in elections.
That said, a teacher from the Chinese city of Chongqing who has worked in schools for many years and asked to remain anonymous for safety reasons doubts that PISA results reveal whether teenagers are ready for adult life: “PISA assesses only three subjects — reading, mathematics, and science — and the results hardly give a full picture of how prepared students are for further education and employment. The range of skills employers value is much broader and differs significantly from what PISA measures, so a high or low score is hardly a sound basis for predicting success or failure in further study and at work. There are already obvious mismatches within education itself: what is taught, what is assessed on tests, and what people will later need on the job don’t fully line up. So PISA scores can’t simply be equated with students’ actual skills.”
The Oxford researchers who prepared the national report on the test results for England also urge caution: “results should be interpreted cautiously, particularly when considering trends in performance over time and when making international comparisons. More confident conclusions can be drawn when making comparisons between groups of pupils within England, where we are not trying to generalise beyond our sample.” Yet one of PISA’s main purposes is precisely to compare how well prepared students are in different countries and to track how that changes over time.
Irina Demyanova, a language teacher who has worked in schools in Italy and Estonia, believes that PISA, like any test, has its “blind spots,” but it still helps bring real problems to light: “I think PISA results deserve close attention, but probably not as the absolute truth. Any test measures some things and inevitably leaves others out. So I probably wouldn’t draw conclusions from a few numbers about the quality of an entire education system or about how ‘smart’ kids have become. But if we see not a one-off dip but a fairly persistent decline, that’s a reason to think about what’s happening to children and to schools in general. Especially when it comes to the basics: being able to understand a text, put a thought into words, pick out the main point, solve a problem, and so on. These are exactly the skills people need regardless of whether they’ll have the most advanced AI at their fingertips ten years from now.”
Notably, even Dima Zicer — an educator and school principal in Estonia known for his skepticism toward standardized tests, grades, and other trappings of traditional schooling — urges people to take PISA results seriously. “In my view, this is pretty serious,” he says in an interview with Tatyana Felgenhauer. “These tests are simple; they aren’t designed for little whiz kids who are supposed to leave their peers in the dust in math. <…> And the problem is, it’s not so much that they flunked [the test] — it’s that there’s a decline,” Zicer adds.
Reading and Math: A Worrying Trend
The decline is indeed significant. The average reading score across OECD countries fell from 477 to 463 compared with the previous round. Until 2012, it held steady at 495–501 points, which means the overall drop has been about 40 points. In math, the OECD average once reached 503 points, but it has been falling since 2009 and stood at 463 in the latest round. Science scores fell the least: from 506 points in 2009 to 482 in 2025. That is roughly half the decline seen in the other two domains.
At the same time, results across the three domains can vary considerably even within a single country. The United States, for example, posts fairly modest math scores — right around the OECD average — but significantly outperforms that average in reading and science. Switzerland ranks in the top quarter in science and scores high in math, yet in reading it trails many countries that perform worse in the other domains. Ireland is another telling example: it placed seventh in reading, ahead even of Estonia, one of the leaders in the other categories. Ireland’s math and science results were more modest, though the country still stayed in the top quarter of the rankings.
The biggest surprise of PISA 2025 was arguably the performance of China and other Asian participants, which took the top spots in all three domains. Beijing, Shanghai, and the provinces of Jiangsu and Zhejiang, which take part in PISA as a single participant, lead by a wide margin in math and science. Their lead over Singapore, which came second in both categories, is 49 and 37 points, respectively. In reading, the Chinese regions ceded first place to Singapore. Japan and South Korea also consistently rank in the top seven across all domains, placing fourth and fifth, respectively, in reading.
Estonia’s results also came as a surprise: it became Europe’s top performer, overtaking Finland. The Finnish school system was long considered a model, but its students’ results have deteriorated markedly. From 2006 to 2025, Finland’s average score fell from 547 to 474 in reading, from 548 to 469 in math, and from 563 to 504 in science. In other words, Finland’s scores dropped by about 70 points on average. Estonia’s scores declined too, but far less than in many other countries. Estonian students performed best in 2012–2018. Compared with that period, by 2025 their scores had fallen by 14 points in science, 15 in math, and 24 in reading. Thanks to this slower decline, Estonia climbed in the international rankings.
So the new round of testing raises two main questions: why have scores been falling since the 2010s, and what is the secret of the top performers’ success?
The Pandemic, Smartphones, and Lost Motivation
The debate over the causes of the decline began as early as the release of the PISA 2022 results. The COVID-19 pandemic was named as one of the main factors. There were grounds for this: during lockdowns, students were forced to study remotely, which may have hurt the quality of education, especially for younger children. For example, according to figures cited in the study, math skills among preschoolers dropped by more than 29%, and early language development indicators by more than 34%.
Paolo Deho, a teacher who recently retired from an Italian public school, agrees that the pandemic had a major impact on education: “COVID really hit schools very hard. My colleagues confirm this, and so does my daughter, who occasionally serves as an examiner for university math exams. She notes that students are very poorly prepared and attributes this specifically to the COVID-19 pandemic — and, well, to artificial intelligence. As for Italian, over the past few years I’ve noticed a marked decline in students’ vocabulary. Kids spent too much time online, and it’s very hard to learn that way — and on top of that, they spent too little time interacting with each other, and that matters too.”
The teacher from China who asked to remain anonymous shares this view: “I think the pandemic is a very important factor behind the drop in PISA scores in some countries. It’s very hard for 15-year-olds to discipline themselves, and very hard to stay motivated when you have to study online from home, so it took a toll on their performance. That partly explains the decline.”
Elizabeth Gotowski, a math teacher in New York, points to other consequences of the pandemic as well: the opportunity to cheat during remote learning and the disruption of the familiar rhythm of school life. “There was less interaction with both teachers and classmates — all of that led to gaps in learning,” she notes.
However, the decline continued even after the pandemic ended. That is why many experts point to another sweeping change of recent decades — the spread of smartphones and social media. Now artificial intelligence has been added to the mix, and some specialists believe it could make matters even worse.
Many countries are debating bans on phones in schools, as well as raising the minimum age for opening social media accounts to 16. In the UK, for example, the government has issued guidance on restricting the use of mobile phones in schools. In many European countries, however, the specific rules are set by regional authorities or by the schools themselves. “Mobile phones were banned at my school,” says Paolo Deho. “The same was true at other schools, though not all of them — each school decides this for itself.”
If such restrictions become widespread, their impact on academic performance may show up in future PISA cycles. But there are doubts that bans alone will be enough. Irina Demyanova believes the causes are far more complex: “We probably shouldn’t reduce what’s happening to social media and artificial intelligence alone. I think the story here is much more complicated. Kids today live in a completely different information environment: an enormous amount of short-form content, constant switching of attention, the habit of getting information very quickly and having access to it all the time. In that kind of environment, reading a long text ‘slowly, thoughtfully, and with proper attention,’ working through a complex topic, and staying focused become objectively harder. AI amplifies a process that is already under way rather than being its sole cause. I would also be very cautious about the idea of simply banning social media and AI for kids. I think it’s far more important to teach students to understand when technology genuinely helps and when it starts doing your thinking for you.”
Mikhail Gantman is also skeptical of demonizing AI: “I really don’t want to write about what everyone else writes about: smartphones, short-attention-span ‘clip thinking.’ After all, my students can sometimes get quite decent results with the help of AI. That said, in the age of AI, when any fake can be produced in two clicks, genuine knowledge of your own becomes priceless. But this idea is getting through to students more slowly than I’d like, and not in every subject.”
China’s experience shows that strict restrictions on phone use in schools don’t necessarily solve learning problems. “In general, in China, from kindergarten through high school, the overwhelming majority of schools strictly control mobile phone use,” our source says. “But outside school, in my view, the use of social media and AI is causing clear problems. I think China’s phone policy is, if anything, too strict. Precisely because phones are strictly banned at school, many students go on a ‘compensatory entertainment’ spree at home, and the phone becomes almost exclusively a source of entertainment — one that eats up a lot of time. As for AI, primary and middle school students, in most cases, don’t use this tool to support their thinking either. Most often they use it to quickly find the answer to a problem or simply copy a ready-made result and get their homework out of the way. Students skip the process of thinking on their own, working out the solution, and absorbing the material, which in the long run undermines their own problem-solving skills.”
Elizabeth Gotowski, the New York math teacher, makes a similar point: “Social media certainly affects how much students read; they come home and scroll through TikTok and Instagram instead of reading, especially if no one is keeping an eye on it at home. And AI has made it possible to do basically nothing on your own: if a student has to write an essay on a book, they don’t even need to read the book — they can just ask AI for a summary, or even have it write the whole essay for them.”
At the same time, Gotowski advises taking into account not only external factors but also students’ individual differences, as well as the limitations of the test itself: “I think PISA offers a good snapshot of the situation, but the test can’t capture children’s real abilities. In my experience, kids grasp complex concepts reasonably well, but they differ enormously in their ability to sustain attention, in their intrinsic motivation, and in how much mentoring and support they need. Many kids struggle with focus and academic discipline, which hurts their ability to do well on tests — but that doesn’t mean they don’t understand the subject.”
Moreover, PISA results have no bearing on students’ further education: they don’t affect admissions, transfers to another school, or grades in any subject. Still, participants have to spend two hours solving problems with full concentration and no reward of any kind to look forward to. Mikhail Gantman notes that a willingness to put in effort for the sake of the problem-solving process itself is becoming increasingly rare among his students.
“I’m seeing fewer and fewer students who want to tackle hard problems. They like the feeling of success, that ‘I did it!’ moment, but they aren’t willing to pay for it with their own effort and time,” Gantman observes. “I see kids who play musical instruments: performing on stage matters to them, but according to my colleagues — the music teachers — very few are willing to sit and repeat a difficult passage thousands of times until it flows easily, confidently, and without mistakes… And that, I’d say, is my main challenge right now. But I don’t know whether it has to do with the times, the country, or the type of school where I work. What I really miss in students is a willingness to ‘suffer’ a little, to ‘sweat’ over a hard problem.”
According to Gantman, some immigrant children also don’t always treat school as a priority: “Children of immigrants from certain regions often have little sense of what school is and why it’s needed. This isn’t an attempt to insult or belittle them. Let me explain with an example. If a German student doesn’t do his homework, he knows that’s wrong. And he apologizes. Students from some other countries say, ‘I couldn’t do my homework, I had to play soccer with my friends.’ In other words, for a kid like that, school is a kind of annoying add-on to everything else he does.”
Irina Demyanova points to yet another factor that is less often mentioned among the reasons for falling PISA scores: students’ psychological well-being. “There’s one more point that I think is very important and that sometimes gets lost behind the test results. In these discussions, we talk about children as a set of competencies: reading, math, science. But these are real, living kids, not just test takers. Today we see anxiety, emotional burnout, depression, and motivation problems in a great many children. And I think this, too, can’t be left out of the equation when we discuss falling scores. It’s very hard to demand intense concentration, complex thinking, and strong motivation to learn from a child who is struggling psychologically. So I wouldn’t claim that children have become less capable of learning and understanding. It may be more accurate to say that the environment in which children grow up has changed, and schools haven’t yet fully learned how to work with this new environment.”
Dima Zicer also argues that the education system is failing to keep up with change: “The education system is simply not keeping pace with how the world is developing <…> The language of the world, of the environment we live in, is changing, while the language we use to talk to children isn’t changing at all.”
Why Some Countries Do Better Than Others
What sets apart the countries whose students score high on PISA? Are their education systems better at adapting to the modern world? Or does it come down to discipline, limits on social media use, and cultural factors?
Discussing the success of the UK, which placed in the top ten in all three PISA 2025 domains, science journalist Irina Yakutenko draws attention in a video on her channel to how the school curriculum is designed and how learning outcomes are assessed: “In England, the school curriculum is much more standardized: the state lays down fairly strictly what skills children should have and what they should be able to do by a certain age. And it doesn’t just lay this down — it tests it with exams. <…> For competencies to emerge, you first have to build a solid foundation.”
“England’s example shows us,” Yakutenko adds, “that countries that try to provide more solid academic knowledge come out ahead. This applies to Estonia, too. On the one hand, Estonia moved away from the Soviet system; on the other, it kept that system’s academic rigor.”
Dima Zicer partly shares this view: “The main approach [used to be] structured pressure — let’s call it ‘discipline.’ That tool has gone out the window, well, almost entirely. <…> In Estonia, education reform began some 20 years ago. Estonia adopted what were then the most up-to-date teaching methods from all over the world, but on the other hand, there’s the tradition of Soviet education. It’s awful that I’m saying this, but I’m saying it.”
Zicer stresses the importance of combining traditional, “vertical” education — with a fixed timetable, discipline, and clear boundaries — with modern approaches that offer more freedom of choice and the flexibility to vary the number of hours devoted to individual subjects. In his view, Finland at one point bet on soft skills, modular learning, and flexible scheduling at the expense of academic preparation (hard skills). Zicer sees this as one of the main reasons for the decline in the country’s PISA scores: “Self-discovery aside, people go to school to learn.”
Irina Demyanova also speaks of the need for discipline and a solid academic foundation: “It seems to me that in Italy, as in many countries, the problem isn’t so much the teaching methods themselves as how they’re put into practice. A lot of attention is now paid to inclusion, individualized instruction, and developing competencies, but sometimes there isn’t enough systematic work on basic skills, discipline, or expectations for results. In some cases, large classes, uneven levels of preparation among students, and excessive reliance on digital devices also take their toll.”
Comparing Italy and Estonia — the two countries where she has taught — Demyanova notes that Estonia’s success was driven by a combination of strong basic preparation, the high status of education, a consistent education policy, and the ability to adapt to the digital environment.
“For example, when it comes to classes on using all sorts of innovative tools, in 2025 Estonia launched the AI Leap program, aimed at bringing artificial intelligence into schools. The program is meant to give teachers, as well as students in grades 10–11, the chance to master AI tools and skills at the highest level. An app is being developed together with OpenAI that doesn’t hand out ready-made answers but instead helps students think for themselves by asking guiding questions. However, some specialists are already voicing concerns that excessive reliance on AI could weaken the role of the teacher, students’ academic independence, and live interaction between student and teacher. I suppose some time has to pass before it becomes clear whether such classes will do any good or quite the opposite.”
Still, Demyanova cautions against jumping to conclusions: “Here, too, I wouldn’t conclude that Estonia will now be the leader forever. In education, it’s generally risky to look for permanent winners: systems change, generations change, and what worked ten years ago may stop working tomorrow because of some new circumstances.”
But while the experts agree that Estonia’s success stems from a combination of modern approaches and a solid academic foundation, the picture is more complicated when it comes to China, whose regions top the rankings by a wide margin.
One thing that stands out in the PISA results is that China, unlike most other countries, is represented by several participants that are assessed separately. And this isn’t only about Taiwan, which is de facto governed independently of the People’s Republic of China. Topping the rankings in math and science is the entry B-S-J-Z (China) — Beijing, Shanghai, and the provinces of Jiangsu and Zhejiang. Macao, which ranks lower, is assessed separately. The rest of mainland China does not take part in the test.
Back when Russia still took part in PISA, its relatively modest rankings were often attributed to the uneven quality of school education. Moscow, St. Petersburg, and other large cities have more strong schools, while in smaller towns the opportunities for a good education are often limited. The teacher from Chongqing, who asked to remain anonymous, notes that similar disparities exist in China:
“In my view, the results of Beijing, Shanghai, Jiangsu, and Zhejiang can’t possibly reflect the overall level of Chinese students. These are the most economically developed regions of the country, with the richest educational resources and relatively progressive ideas about teaching. The gap in learning conditions and academic performance between these regions and smaller provincial cities and the countryside is very noticeable.
“First, there’s a gap in staffing: the requirements for getting a teaching job in Beijing, Shanghai, Jiangsu, and Zhejiang are extremely high — you usually need a degree from a top university — so the overall caliber of teachers there is very high, whereas elsewhere the hiring requirements are less demanding.
“Second, there are huge differences in family support: in developed regions, many families pour all their resources into helping their child with schoolwork and are willing to pay for all kinds of extracurricular activities. PISA emphasizes well-rounded abilities, so differences in how much families invest translate directly into test results.
“Third, the curricula have a different focus: in Beijing, Shanghai, Jiangsu, and Zhejiang, more room is given to ‘quality education’ (suzhi jiaoyu) — that is, to developing the very skills PISA pays particular attention to. If students across the whole country took PISA, the average score would most likely be lower.”
Then again, even if China’s nationwide average turned out lower, other Asian countries still rank high in PISA, outperforming many Western nations. One might assume this is due to discipline and the high status of teachers in Asian societies. But the teacher from Chongqing points to other factors as well: a tradition of exam preparation, the intensity of study, and families’ attitudes toward education. China places special emphasis on math and science (T-invariant wrote about Chinese scientists’ leadership in natural science research in the article “New ‘Undesirables’: Why the Pentagon Is Restricting Cooperation with Moscow State University and Higher School of Economics”).
“In my view, Chinese students’ high PISA scores come down above all to the fact that, on the whole, we’re very good at handling standardized exams. We have a long-standing exam tradition, and that mindset has survived to this day. Major exams such as the zhongkao (the high school entrance exam) and the gaokao (the university entrance exam) have relatively stable question formats, and we’re very good at preparing for exams.
“The same thing happens with PISA: teachers can study in detail exactly what is being tested and what the tasks are like, and drill precisely that; with a large amount of targeted practice, the skill becomes automatic.
“Besides that, there are differences between Eastern and Western approaches to teaching. Take math: in many countries, lessons focus on the teacher deriving formulas together with the students so they understand where the formulas come from, whereas our teaching tends to work differently: first the formula is given ready-made, and then the knowledge is consolidated through a large number of exercises. This way of thinking, geared toward generalization and systematization, gives students an advantage on standardized tests.
“Second, basic education in China generally rests on a solid foundation, and our curricula align very well with the three domains PISA tests. As for reading, the emphasis on practice doesn’t let up from primary through secondary school; students spend many hours developing their reading skills, so naturally their foundation is strong.
“Math is one of the top-priority subjects in our education system in general. The material at a given grade level is usually harder than in the West, and since society as a whole places great importance on math, it’s easier for students to shine on math tests.
“In addition, competition in education in China is intense, and students spend more hours a week studying. Finally, there are sociocultural factors. As we see it, education is a chance to change your fate and your station in life, and society as a whole places enormous value on it. Many families put a lot of effort into keeping track of their children’s academic performance, and thanks to the combined efforts of everyone involved, students find it easier to score well on tests like this one.”
How to Change Schools Without Breaking What Works
Although PISA results can’t fully show whether students are ready for adult life, they help identify problems in education and point to possible directions for change.
The experts T-invariant spoke with agree that education systems need to combine different approaches. On the one hand, they should maintain discipline, clear criteria for assessing knowledge, a solid academic foundation, and enough hours for mastering basic skills, above all reading and math. On the other, they should give students more choice of subjects and use modern teaching methods.
A separate challenge is teaching students to use social media and artificial intelligence in ways that help them learn rather than replace independent thinking.
“I think it’s important to teach students to understand when technology genuinely helps and when it starts doing your thinking for you,” says Irina Demyanova. “Schools could well have classes that are less about ChatGPT and more about working with information: how to check a source, how to tell fact from opinion, how to formulate a good question, how to check an AI’s answer, and, most importantly, how to think things through yourself first and only then use the tool.”
The teacher from China takes a similar stance. He opposes a total ban on phones in schools and believes it’s better to teach children from an early age to use digital devices mindfully.
“It’s better to help students develop a healthy attitude toward phones from an early age: not absolute freedom, but moderate control and sensible phone use. That would let children benefit from the educational opportunities of the digital world and keep up with the times, while the necessary limits would make up for their lack of self-control.”
In the teacher’s view, artificial intelligence can also be useful as long as it doesn’t replace students’ own work. What’s more, for some children it can become an additional learning tool:
“It’s the same with AI: what does harm isn’t the tool itself but using it instead of thinking for yourself. But if students use AI to discuss ideas and support their thinking process, it can actually help them learn. And it can be especially useful for children who tend to be shy and find it hard to ask the teacher or their classmates for help.”
Elizabeth Gotowski, on the other hand, sees no need to bring artificial intelligence into the learning process. In her view, developing critical thinking matters more than making learning more convenient and efficient.
“There’s no need to make education more efficient — what matters is that students learn to think critically, even if that’s less convenient or less efficient. I’m somewhat against the idea that AI should be used in the classroom at all costs.”
“I think we should go back to paper-based tools and have students write essays on paper in class,” she adds.
Changes in education are already under way, but they don’t always add up to a coherent reform. In Germany, for example, schools are trying to adapt to new conditions while holding on to many elements of the old system.
“When it comes to reforms, things in Germany are really complicated: 16 federal states, 16 curricula, 16 versions of the school-leaving exams,” says Mikhail Gantman. “Right now they’re trying to standardize the final exams. In some places, Gymnasium [the academic track of German secondary school — T-invariant] is being extended to 13th grade; in others, curricula are being changed. Until recently, every school was shouting about digitalization — now they’re starting to shout that digitalization has more downsides than upsides. In Germany, people still think homework is optional. All in all, there are changes, but I don’t see any real revolutions yet. What I see is an attempt to preserve the old system in a new world.”