Thursday, March 24, 2011

Models Can Be Superficial in Politics, too

From the NYT

March 24, 2011, 12:55 am

Last week, Harry Enten, a blogger who studies statistics and elections, published a model that suggested that Republicans were likely to retain control of the House in 2012.

This is not, in and of itself, a remarkable conclusion. Incumbency is normally a fairly powerful advantage in Congressional elections. Republicans have a reasonably large majority in the House now, so the safe money is on their being able to keep it.

What was more daring, however, was the confidence that Mr. Enten was willing to assign to his projection. He stated that there was a 95 percent likelihood that Republicans would win between 228 and 248 seats (218 are needed for a majority). “Unless a historic event occurs,” Mr. Enten wrote, “Republicans will still be in control of the House of Representatives after the 2012 election.”

Just how historic an event are we talking about? If one takes his model literally, it implies that there is less than a 1-in-100,000 chance that the Democrats will regain a majority.

Mr. Enten’s model got picked up in several blogs that I read, and I’ve received other requests for comment about it. How can he be so confident when, among other things, we don’t know the identity of the Republican candidate for president or how strongly President Obama will perform against his opponent, the effects of redistricting, the state of the economy, the number of retirements in each party and a whole host of other things?

Actually, he shouldn’t be so confident. Mr. Enten normally does outstanding work and will probably have my job one day. But sometimes statistical models are as skin-deep as models on the runway, and this one is such an example.

The issue with this model, and some others like it, is what’s known in the statistical business as overfitting. This occurs when the number of variables is large relative to the sample size: in this case, the full version of Mr. Enten’s model contains six variables, but is used to explain only 15 cases (Congressional elections in presidential years since 1952).

A general rule of thumb is that you should have no more than one variable for every 10 or 15 cases in your data set. So a model to explain what happened in 15 elections should ideally contain no more than one or two inputs. By a strict interpretation, in fact, not only should a model like this one not contain more than one or two input variables, but the statistician should not even consider more than one or two variables as candidates for the model, since otherwise he can cherry-pick the ones that happen to fit the data the best (a related problem known as data dredging).

If you ignore these principles, you may wind up with a model that fits the noise in the data rather than the signal. Mr. Enten’s model, for instance, contains a variable for cases in which there has been an “unprovoked, hostile deployment of American armed forces in foreign conflict” that “has resulted in at least 1 fatality during the past term,” but it applies only when the war was started by a president of the same party as the one that currently occupies the White House, or when the current president is of a different party but has perpetuated the war for more than one term, conditional on the fact that the Congress and the White House are controlled by different parties.

Those of you who find this definition confusing have the right idea. It is so narrow that it applies to only two cases, 1976 and 2008, that the model would otherwise do a fairly poor job of explaining. While there is no doubt that wars can have some impact on voters’ assessment of the Congress, if one is willing to apply literally a half-dozen different qualifications to determine which wars he deems relevant, he could presumably come up with any number of other riffs on the definition that would apply to different years instead. (For example, if the elections of 1968 and 2004 were the ones the model had a rough time explaining, he could invent another version of the definition of “war” that would apply solely to those years.)

The problem with an overfit model is that, because it is so fussy about handling past cases, it tends to do a poor job of predicting future ones. Imagine that I was a petty criminal of some kind, and that I deputized you to come up with a way to help me pick combination locks. I also gave you three locks to experiment upon.

What I’d really be looking for would be some set of principles on how one picks locks: perhaps a certain type of paper clip works especially well, or a disproportionate number of combinations contain numbers like ‘7’ and ‘13’. Instead, after studying the issue for a few days, you report back to me that you’ve found the perfect solution. If the lock is blue, use the combination 45-12-26. If it’s red, use 33-9-16. And if it’s black, use 22-10-41. That would certainly be a very reliable way to pick these three particular locks, but it wouldn’t tell me anything about how to pick locks in general. This is essentially the same thing that happens when one produces an overfit statistical model.

Sometimes, the person creating the model will not discover this until it’s too late, especially if the variables in the model otherwise seem reasonable. In this case, however, it is fairly easy to demonstrate the model’s limitations.

Mr. Enten’s model is built on data from 1952 through 2008. He’s using it to forecast the election outcome in 2012; we don’t know yet how that prediction is going to turn out.

What we can do, however, is see how the model would have done on a case outside of its sample: 1948. There’s not an obvious reason to include 1952 in the data but not 1948; both are in the postwar period.

If we use the model to make a “retrodiction” for 1948, however, it fares very poorly:

The model would have predicted that Republicans, who then held the majority in the House, would have maintained it by winning 241 seats. Instead, they lost 75 seats to Harry Truman’s Democrats and wound up with just 171. That’s a 70-seat error, for a model that supposedly had a 95 percent confidence interval of plus or minus 10 seats. The odds against that, if the model had been specified correctly, were roughly one tredecillion (a 1 followed by 42 zeroes) to one.

Perhaps 1948 was special in some way. But if 1948 was special, 2012 could be, too, and the model could be just as inaccurate.

One alternative is to use a model with the same structure as Mr. Enten’s, but to include 1948 in the data that we use to build it. If we do that, the mean projection isn’t terribly different — Republicans are projected to win 230 seats rather than 238 — but the margin of error is much higher. Instead of having a 95 confidence interval of plus or minus 10 seats, we’d instead have one of plus or minus roughly 43 seats, which means that anywhere from 187 seats to 273 would be a reasonable guess. That model implies that Democrats have about a 38 percent chance of regaining control of the House, which is in line with what betting markets think.

There are some other problematic elements in Mr. Enten’s model. For instance, whether we define Libya as an “unprovoked, hostile deployment of American armed forces in foreign conflict” — I have no idea whether it qualifies — makes a difference of 23 seats in the answer the model comes up with.

The fundamental error, however, is in assuming that a model that is optimized to explain the past will do an optimal job of predicting the future. These are two very different things. When data sets are extremely large, the distinction may not be so important; in fact, people are probably too reluctant to indulge some complexity in their models in such cases. But small data sets, like one containing just 15 elections, are much less forgiving, and almost always require extremely simple specifications that are well-grounded in political theory.

It’s easy to forget about these distinctions when you’re trying to squeeze every last bit of juice out of a model (I’ve certainly made these mistakes myself on occasion). Several Congressional forecasting models, for instance, are based on using the Gallup generic ballot poll, and the Gallup generic ballot poll in particular. Why use the Gallup poll alone instead of an average of polls? We know that, over the long run, polling averages are almost certainly better than any one poll taken alone. But the Gallup poll had happened to be a little closer to the mark in the small sample of recent midterm elections, and that made it a popular among those who were trying to optimize the fit on past data.

Those models paid the price last year; whereas the polling average did a good job of predicting the size of the Republican wave, Gallup massively overestimated it by forecasting a 15-point Republican win, about double the actual margin. As a result, one popular model that relied on Gallup data and that billed itself as being able to predict these outcomes called for almost exactly a 77-seat Republican gain in the House, well above the actual total of 63.

This is also a problem to a greater or lesser degree with the many other models that are used to forecast Congressional and presidential elections based on fundamental variables like economic growth, wars, incumbency, the results of the previous election and so forth. Within the category of economic variables alone, there are at least five or six plausible metrics to consider (unemployment, G.D.P., personal income, inflation, consumer confidence), and each of these can be specified in any number of different ways (for instance, the absolute level of unemployment, or the change in unemployment as compared to some prior period).

When you have so many candidate variables and so few elections (just 16 presidential years since World War II), the potential for overfitting is enormous. And the consequences of overfitting are much greater than people realize. In Mr. Enten’s case, when we added just one more data point, his model’s error increased not by 10 or 20 percent, but by more than 400 percent.

That doesn’t mean that models like these are completely hopeless. For instance, a model to predict presidential elections called “Bread and Peace”, which uses personal income and war casualties as its only two inputs, seems to be sensibly designed. But when you come across a model that seems implausible or overconfident, you absolutely ought to be skeptical of it: there often isn’t much there when you look under the hood.

Sunday, March 13, 2011

California's Education system is failing

From the SF Chronicle:

The facts are hard.

A generation ago, California had what was considered the best education system on the planet.

Today, our daughters and sons attend one of the worst-performing education systems in the industrialized world.

We are failing on the rock-bottom basics. California students' ability to read is ranked 49th in the country by the U.S. Department of Education. Our kids' ability to do math is ranked 47th and we are second to worst in science. Compared globally, the situation darkens further. Of the top 35 nations, the United States is ranked 29th in science and 35th in math. Your neighborhood school might be good by California standards, but that is a very low bar indeed. Our education crisis is a human tragedy and a looming economic disaster.

The Bay Area Council resolutely refuses to accept this crisis as our state's fate. Let's get past the political gridlock and get down to the real business of dramatically improving California schools. We know, as every honest study has shown, that it will take a combination of real dollars and major changes in the way we deliver education.

Several years ago, the Bay Area Council successfully fought to pass legislation creating a fully functioning statewide education data system. That system is now emerging and will be tied to a widely praised common national curriculum. We were inspired by Florida, which implemented the nation's best data system, and now has far superior education outcomes. Florida's student population is quite similar to California's.

For students, the data system should be used throughout the year to help their teachers diagnose needs, guide instruction and monitor performance. It can guide program progress and help principals identify teachers who are getting results, and whose efforts and approaches can be modeled. Finally, considered with other factors, test scores must be a part of meaningful teacher evaluations to ensure our kids are getting the best teachers possible, not just the ones that have accumulated the most seniority.

What happens in a classroom is complex and the fix has multiple answers, but demonizing data throws out a critical flashlight we need to get out of this wilderness.

Read more

Thursday, March 10, 2011

The 25 best Financial Blogs

From the TIME

· Business Insider

· Grasping Reality with a Sharp Beak

· Econbrowser

· Rortybomb

· Dealbreaker

· Paul Kedrosky

· The Wealth Report

· WalletPop

· Naked Capitalism

· Real Time Economics

· Megan McArdle

· DealBook

· Street Sweep

· Free Exchange

· Economix

· The Big Picture

· Zero Hedge

· Planet Money

· Ezra Klein

· The Consumerist

· Freakonomics

· Calculated Risk

· Marginal Revolution

· Felix Salmon

· The Conscience of a Liberal

Times Higher Education World Reputation Rankings

from the THES

10 March 2011

A university's brand - crucial in helping to attract students, staff and funding - is built on esteem. Times Higher Education's first World Reputation Rankings reveal how academics view the strength of institutions' teaching and research, while John Morgan explores brand values, virtues and vices

In an increasingly competitive higher education marketplace, branding has become big business for universities.

Institutions know that, in a sense, the degrees they confer are worth only as much as their brand. In nations where tuition fees are established, students "buy" a brand that will appeal to the right businesses when it is time to find a job; their choice of university will become part of their own "brand identity". To attract the right calibre of academics, a university relies on its brand. And when those same academics submit a proposal for research funding or a paper to a leading journal, the brand of their institution may play a role in how their research is judged. The university's brand becomes part of their own brand as an academic.

The notion of a university as a brand is one that many in higher education are comfortable with. It induces a wave of nausea in others, who warn that by focusing on branding, universities promote a view of higher education as a commodity rather than as a good in itself.

But if a university is a brand, a key factor determining its strength is reputation in teaching and research (brand and reputation are distinct but related). And the views of academics on university reputation are crucial, for they give an insight into which institutions are best placed to attract top talent, and also influence the views of students and parents.

The results of the first Times Higher Education World Reputation Rankings (see related files, right) shed light on this increasingly important measure. The reputation ranking is drawn from a survey of more than 13,000 experienced academics worldwide, carried out by polling company Ipsos for our rankings data provider, Thomson Reuters. The data informed the current Times Higher Education World University Rankings 2010-11, but are now published in isolation for the first time, revealing clear discrepancies between some institutions' reputations and their overall ranking.

"It is great to be able to see reputation in 'pure' form and to compare the outcome with the outcome by other measures," says Simon Marginson, professor of higher education at the University of Melbourne. "Reputation is itself an identifiable market - one that matters and has material effects. It is created by a number of accumulating factors: the weight of activity in hundreds of different institutional sites, relations and transactions; conscious promotional campaigns, major events that result in news reporting, memories of past activities; and word of mouth effects."

One notable surprise is the strong performance of Japanese institutions, with the University of Tokyo, Kyoto University and Osaka University all performing better in the World Reputation Rankings than they did in the World University Rankings.

Japan has five institutions in the reputation top 100, making it the best performer behind the US and the UK and producing a better score than major higher education nations such as Canada and Australia.

The US remains dominant in reputation, taking 45 of the places in the top 100, but again, there were some surprises.

The California Institute of Technology, rated second in the World University Rankings, is 10th on reputation. Marginson suggests that the institution's "specialist science research profile is tailor-made for the THE (main World University Rankings) and some other rankings", but that it is "not as well known as the more comprehensive Ivy Leaguers".

The UK universities generally do better in the reputation rankings compared with their overall rankings, with the London School of Economics rated 37th on reputation but 86th overall.

China's top-rated institution is Tsinghua University, in 35th place.

India, despite having no representative in the top 200 of the World University Rankings, can boast of one in the reputation table, the Indian Institute of Science (in the 91-100 band).

Of the four Australian universities in the reputation top 100, the leading two - the University of Melbourne and Australian National University - place lower than they did in the world rankings. This shift may betray a perception that "top Australians are not stellar quality", Marginson suggests, which "might be partly due to the fact that Australia markets 'Brand Australia' in the international education market as if all institutions are equivalent".

Crucially, there appears to be only a small number of globally recognised "super brands". Those completing the survey were asked to highlight what they believed to be the strongest universities in their specific fields. After the 20 most frequently cited universities, there is a rapid drop in the number of mentions given to institutions. Indeed, there were only very narrow differences between the scores of those below the rank of 50 (which is why they are ranked in groups of 10).

There is also a high correlation between the scores respondents gave to institutions for teaching and research. Nevertheless, in general, US and Japanese universities seem to have better reputations in research than in teaching.

It seems logical that larger institutions would fare better, because they are more likely to gain attention as they generate greater numbers of papers and offer more scope for academic exchanges and collaborations. However, Simon Pratt, project manager for institutional research at Thomson Reuters, says the data "do not show a strong correlation between institution size and reputation".

But he adds: "A size-related aspect does emerge when comparing differences between the results in the Reputation Rankings and the results for the World University Rankings. For example, those that do better in the Reputation Rankings than in the World University Rankings are often quite large universities such as the University of Tokyo, Kyoto University, the University of Massachusetts and the University of Wisconsin."

So size, and the greater likelihood of international exposure, may help to explain Japan's strong performance (see box, below).

What are the practical implications of all this? "Reputation in the academic world is a necessary condition for the success of a university," says Bernd Huber, president of the Ludwig Maximilian University of Munich. "To be well known as a place with high academic quality builds the foundation for hiring and recruiting excellent young researchers and professors."

Of course, reputation and brand are not the same thing. But Pat Freeland-Small, chief marketing officer at the University of Melbourne, says the former feeds the latter.

The world elite of universities, he says, do not need to advertise, "but in a way they are advertising. They are constantly communicating the quality of what they do through their people and what they naturally put out...It is their people, their quality of research - notions that come through the academic community - that advertise their international profile.

"It will be the quality, the capital, that universities are perceived to have that will start to drive...the development of global brands."

Freeland-Small joined Melbourne from the Foster's Group, where he was business development director. He oversaw the university's branding as it introduced the "Melbourne model", a radical restructuring of the undergraduate curriculum designed to align it with structures in the US and Europe. He says he initially had to work just to make people aware of the value of the university brand.

"There hasn't been any sort of formalised evaluation of university brands because universities are not listed on the stock exchange."

So what is Melbourne's "brand"? And is it true to the reality or just a marketing man's construction? Freeland-Small says it is about being "guided by the public-spirited aspects of being a university within the city of Melbourne and state of Victoria, but with world excellence in mind".

Phrases such as "heart and soul", "truthful" and "authentic" recur in his description.

"The academic community can be very cynical," he says. "It was important to keep things real. The proposition was basically one of the university striving for excellence in research and learning" - a branding message that was "the sort of proposition that the academic community bought into anyway".

As that implies, not all branding derives from and is directed by marketing teams. An interesting example of what might be termed "soft" branding - based in representations of a university's research or scholarship rather than overt marketing - is the broadcasting of the lectures on justice by Michael Sandel of Harvard University, the top institution in our reputation rankings.

After being shown by a Boston-based public TV station and online, the lectures by Sandel, the Anne T. and Robert M. Bass professor of government, were picked up by the BBC in the UK and by a public broadcaster in Hong Kong.

Viewers see Sandel deliver his lectures in Harvard's impressive Sanders Theatre: all dark-wood panelling, pews and balconies. Sandel himself, with his receding hair and thin lips, is said to be the physical inspiration for Montgomery Burns in The Simpsons. But unlike that character, Sandel comes across as likeable and witty. He smiles, he jokes, he engages with his students' ideas and takes the trouble to ask their names. He makes the work of Jeremy Bentham and Immanuel Kant accessible without dumbing down.

But the camera spends as much time on the listening Harvard students as on Sandel. They are a racially diverse group, they make lively contributions and laugh too; they appear intelligent but "normal". After seeing all that, what bright school pupil wouldn't want to go to Harvard? The broadcast allows promising students to imagine fitting in there.

Christine Heenan, vice-president for public affairs and communications at Harvard, says that for many viewers, this was their first "actual" image of the university, as opposed to the fictional representations in films such as Love Story or The Social Network. The latter gives an unflattering portrayal of some Harvard undergraduates.

"In addition to being stimulating and thought-provoking," she says of Sandel's lectures, "one of the other values is that it is demystifying: this is what the college looks like, this is how the professor sounds."

The university is using digital media in other ways, too, to spread its message to new audiences. A web simulcast panel discussion about the 10th anniversary of the completion of the mapping of human genome, chaired by Harvard president Drew Gilpin Faust, generated intense interest. And 8,000 live viewers followed a recent panel discussion among Harvard Kennedy School scholars about Egypt's future.

Then there is Harvard's Daily Gazette, emailed automatically to all faculty and any other subscribers. Readers now share stories about Harvard research and scholarship with colleagues and family around the world.

"We make it as easy as possible for that information to be shared, to be uploaded to Facebook or to other social networking sites, and to extend these stories as far and as wide as possible," says Heenan.

Her job is about "reinforcing positive associations" and seeking to "combat the false dichotomies" of Harvard's brand. "We were viewed as traditional. That doesn't mean we're not open to new ideas, new ways of doing things. We were viewed as old - we are - but that doesn't mean we're not innovative."

Would Heenan agree with the argument that the best university branding involves the honest representation of research and scholarship? "I think this is the future of any kind of effective branding," she says. "It is more show me - not tell me - and let me draw my own conclusions."

Brand and reputation are not static. They require care and management - especially when events threaten to tarnish an institution.

University College London found itself facing a potential reputational nightmare after Umar Farouk Abdulmutallab was accused of trying to blow up an aircraft with a bomb concealed in his underwear on Christmas day 2009.

The Nigerian had been a mechanical engineering undergraduate at UCL and president of its Islamic society. The media asked whether he had been radicalised at UCL, and whether the college had been too lax in its oversight of the Islamic society's activities.

Mark Sudbury, UCL's director of communications, says the institution's immediate message centred on establishing a "credible starting position" by demonstrating the work the college had already done in challenging extremism. But it also stressed its "commitment to freedom of speech".

"Ultimately, we did not suffer major reputational damage because people understood the position we were coming from," Sudbury says. "We did some research with key audiences a few months after, and a fairly strong message came back that it hasn't damaged us."

On more orthodox branding territory, UCL has undergone a major shift in creating what Sudbury describes as "a coherent corporate identity that could be used across the organisation".

UCL is now branded with the phrase "London's Global University".

The college already had "a collection of quite strong brands in their own right", Sudbury says. But many outsiders were unaware that world-famous names such as the Slade School of Art and the Bartlett School of Architecture had any connection with UCL.

There has been "a big impact on our reputation globally since (people) know these sub-parts are part of the UCL family", Sudbury says.

To further refine and clarify its brand, the institution also gave a great deal of thought to the name University College London, which "can be quite complex for some audiences", Sudbury says. "There are issues about how that fits in with the University of London. Our focus has been very much on developing the 'UCL' brand. We want to be known as UCL."

Of course, a university's branding is a significant consideration not just for an institution and its success, but also for its students and potential students.

"A diploma from a university is one of the most personal brand choices you can make from an individual point of view," says Doug de Villiers, chief executive officer of brand consultancy Interbrand in Africa. "We put a massive amount of value into it."

Businesses' perception of university brands also adds value, points out de Villiers, who is also adjunct faculty at the University of Pretoria's Strathmore Business School.

"If you look at (consultancies) like McKinsey or Deloitte, these guys have a specific profile of graduates that they attract. They look to places like Stanford or the Massachusetts Institute of Technology, which adds value to those academic institutions' brands."

As the number of people with a degree increases, a premium brand degree becomes all the more valuable to distinguish individuals. "Given the choice of going to a public university or getting an MBA at Harvard, you would choose Harvard - because of the reputation and because of the brand," de Villiers says.

While de Villiers admits that it can be "tough" to talk to academics about branding, he says "you will see academics becoming more comfortable with the concept of brand and brand value. It also attracts faculty: it has implications for their publishing and the value of their personal brand."

He continues: "Universities are understanding that they need to consider the entire stretch here. The ability to attract students, the ability to engage in expansion with other universities globally - this all comes from branding."

For some, that could be a dispiriting vision of higher education - a calling reduced to a commercial service.

Roger Brown, editor of Higher Education and the Market (2010), warns that the increasing interest in branding paves the way for universities to spend more money on marketing and amenities to attract students, rather than on teaching and research.

"The more you put higher education on a market basis, the more resources are diverted into this sort of activity," he says.

Brown, professor of higher education policy at Liverpool Hope University, sees reputation surveys as part of the problem. "Unfortunately there is very little basis for any of the reputational surveys...Very few people know much about other institutions."

Because the substance of what happens at universities occurs at departmental level, out of sight of most observers, reputation surveys "tell you nothing about quality", Brown contends. "They reinforce the great push for status that is the curse of our times and which (demands for) higher graduate contributions (to the cost of their education) will continue to increase."

More generally, the focus on reputation and brand "gets us even further from what ought to be the purpose of higher education", he says. "As a student, you should go to the university and (do the) course that is most going to develop you individually and make you think."

But if it is true that students are increasingly buying a brand when they choose a degree course, and increasingly seeking a premium brand, then some important trends emerge for universities.

Premium brands can charge premium prices. In countries such as England, where the government already wants to create a market in tuition fees, brand and reputation will be central to universities' thinking on how to set fees and how to pitch to students.

Our survey suggests that there are only a handful of globally recognised brands in higher education. As universities in the West and in Asia increasingly look beyond their regional and national environments - where they may have strong profiles - and set up overseas campuses and partnerships, they need to establish themselves quickly with foreign students and academics who may not be familiar with them.

More universities will want to join the ranks of those with a global reputation, whether that is a reputation as an elite institution or one with proven vocational and professional qualifications. In a crowded, competitive marketplace, you need something that makes you stand out instantly: a brand.

Eastern eminence: Japanese stand out in the eyes of the world


Japan's reputation results will be encouraging for a nation traditionally regarded as an underachiever in world higher education.

Roger Goodman, Nissan professor of modern Japanese studies at the University of Oxford, says the disappointing performance of Japanese universities in rankings has led to "academic panic and a lot of soul-searching".

He notes that Japan has a population of 127 million, "the world's third-largest economy and the second-largest investment (if one adds public and private contributions) in its higher education system. Its research output is commensurate with its size; it has a 150-year history of investment in higher education at the national and private level; it has one of the highest rates in higher education participation in the world - 75 per cent of 18-year-olds go on to some form of tertiary education in Japan; and Japanese schoolchildren come pretty close to top in all global comparative league tables."

In that context, the country's international rankings performance is regarded as "a pretty paltry result", Goodman says. Its top universities in the Times Higher Education World University Rankings 2010-11 are Tokyo (26), Kyoto (57) and Tokyo Institute of Technology (joint 112).

So why does Tokyo believe it fared so much better in the reputation rankings, where it comes eighth? A university spokesman believes reputation "could capture many elements that other numerical measures may miss.

"For example, achievements of our staff in many disciplines, such as those in humanities and social sciences, are not accurately reflected in standard citation datasets as they are mainly published in Japanese or in journals not included in such datasets.

"Also, overall and long-term contribution to society may not be easily reflected in many numerical, flow-based measures."

Continental shift: Excellence is back on the agenda in Europe


Why does no continental European university appear to match the stellar reputations of leading US and UK institutions?

Bernd Huber, president of the Ludwig Maximilian University of Munich, one of the highest rated continental universities in the survey (48), thinks the explanation lies in the political and social trends of the post-Second World War era.

Universities in Germany, and continental Europe as a whole, were organised with a "very egalitarian approach", he says. "The idea that each university would produce roughly the same quality of research and teaching was predominant in Germany and (across) Europe. That basically had the effect that universities had quite a good average quality but no centres of excellence were evolving."

It was in the 1980s that change began with the waves of liberalisation and deregulation, Huber says. This started "a differentiation process within Europe", which he believes will ultimately allow the continent's best universities to close the gap on their Anglo-American competitors.

Huber says Ludwig Maximilian has made "considerable progress regarding our international reputation".

In a possible sign to the US and the UK that brand competition is becoming fiercer, Huber says the perception of German universities overseas has "completely changed" recently.

"People know much about German universities; they know a lot about Ludwig Maximilian. The key thing that has enhanced perceptions is the Excellence Initiative. It was an enormous success in terms of international visibility and reputation."

The Excellence Initiative, a joint federal and state government scheme, involves just under €2 billion (£1.7 billion) of funding in the initial rounds between 2006 and 2012 to create "internationally visible research beacons".

But in branding terms, what makes Germany's universities unique?

They certainly can boast of their age and history. "All this reflects academic distinction," Huber says. "This is a strong selling point, especially compared with the relatively young US universities. A more than 500-year tradition is something that is respected by most people."