I spotted a wonderful article by Graham Witcomb in Intelligent Investor a couple of weeks ago about IDP and the language testing industry in general.

In it, Witcomb notes:

“In theory, there’s still choice. In practice, language testing operates as a cartel with tremendous pricing power. For their own convenience, governments typically pick just a few companies to run tests for immigration, so exam takers can’t shop around as they would for other goods and services.”

Sure, there may be a cartel, but stiff competition has eroded the profitability of individual testing firms in recent years:

“IDP would be a fantastic business were it not for one major weakness in its business model — its gatekeeper status isn’t earned, it’s ordained. Governments decide who its language gatekeepers will be, and they can take that privilege away in a heartbeat, or dilute its value by offering other companies the same deal.”

This is what I’ve been saying here for most of the past year.  Competition has finally arrived.  More competition is en route.  The old monopolies are dead or dying.  A lot of individuals whose livelihoods are affected by the ebb and flow of the IELTS monopoly might pile in to object, but it’s true.

Regarding Canada in particular, Witcomb notes:

“The new testing options will erode IDP’s market share. If the company’s SDS share slips from 100% to 70%, it would mean 80,000-110,000 fewer tests; at $300 per test, a 2-4% drop in total revenue would be the result and a slightly higher fall in net profit due to fixed costs. But with four new competitors, market share losses over the long term could be much higher than we’ve assumed here. We wouldn’t be surprised if IDP’s market share eventually settles below 50%.”

Witcomb astutely points out that lowered language requirements for Canada will also reduce the number of repeaters in the years ahead.

Despite all of this, Witcomb seems more bullish on IDP’s prospects than I am.  He suggests that cartel-like pricing will keep profits high despite market share losses.  He may be correct. But as I have noted here before, DET is coming, and they aren’t going to charge $300 per test.

It is only a matter of time.

Witcomb doesn’t touch on it, but as I have noted here before IDP’s long-term viability is somewhat dependent on its ability to create a “next-gen” IELTS which can compete with DET. The three-headed nature of the IELTS program may complicate that.

An interesting tidbit from the Australian Financial Review:

“A survey of 11,500 prospective, applied and current students by recruitment firm IDP Education found that there has been a sharp increase in the popularity of the US as news of more restrictive policies in Australia, as well as in Canada and in the UK, have started to influence decisions.

Jane Li, IDP Education’s area director for Australasia and Japan, said the US used to rank last on a preference list of four major student destinations but has now leapt to first. Canada has gone from first to last.”

What does this mean for the business of language testing? Well, you can expect DET’s volume to increase, as you don’t really need a TOEFL, IELTS or PTE score to study in the United States. As I’ve indicated here, I work closely with a decent number of students at very good universities in the USA and most of them applied using DET scores.

Perhaps it means a slight uptick in the use of TOEFL, as that test still has strong associations with schools in the USA.

\In any case, it certainly portends a decline in the use of the IELTS test. I think the importance of Canada to IDP’s business model cannot be overstated.

Below find a complaint from an IELTS test-taker that was shared on social media a few days ago.  Click to embiggen and get a clear view of the image.

This individual experienced a technical problem during his online (at home) IELTS. Their test was not validated and no refund was given. They then spent a large amount of money to take the test at a test center in a different country.  They report a total cost of 1400 Euro to get an IELTS score.

They also shared the complaint they sent to IDP CEO Tennealle O’Shannessy which touches on some of the recurring complaints of test-takers from the Global South.

A few things stand out:

  1. This whole ordeal could have been avoided had the test taker been given the benefit of the doubt and offered a free re-test. Remember that this sort of complaint serves as a big blinking billboard advertisement for tests that do provide the benefit of the doubt when something goes wrong. I’m not sure that the legacy test-makers realize that yet. 
  2. I should probably stop writing here because point number one is the biggest problem that legacy test-makers have when it comes to online testing. Unauthorized software running in the background during the test?  Canceled. No refund. Finished the test really quickly and still got high scores?  Canceled. No refund. Your internet stopped working? Canceled. No refund.  Mom accidentally opened the door to your room? Canceled. No refund.  Something weird happened and we can’t tell you exactly what because of privacy concerns?  Canceled. No refund. I understand that it’s harder for legacy test-makers to provide re-tests because their proctoring systems are more costly.  But it has been four years and clearly the loss of market share is costing them more than it would to pay people to proctor some re-tests.
  3. Don’t get me wrong.  The tests must be canceled when the above things happen.  But free re-tests would go a long way towards rehabilitating the image of these tests among the test-taking population. 
  4. The complaint made to IDP’s CEO is a bit trickier to parse.  But, again, I think it speaks volumes to why certain tests are growing in popularity and others are declining.  It’s increasingly obvious that certain legacy test-makers have seen profits from their home turf in Europe and North America evaporate over the past few years (decades?).  In response, they’ve pivoted to making as much money as they can from test-takers in the Global South. That’s valid, but they’ve shown a somewhat callous disregard for the circumstances from which people there are testing from.  I guess in the long run it doesn’t really matter because better managed firms will come and replace the legacy test-makers… but maybe the legacy players could do better and avoid that fate. 
  5. Don’t take my word for it.  Just look at the share price of IDP Education.  This is not a firm that people are optimistic about. I know, I know, they are also facing headwinds when it comes to placement fees… but IELTS is still, by far, their biggest revenue generator.

I posted the following on LinkedIn last week, and IELTS tutors piled in to my page to insist that it will never happen, and that people are proud to take the IELTS instead of some other test.  Look, I could be wrong.  But I must note that no TOEFL tutors came to say that.  TOEFL tutors already know what the Duolingo English Test has done to their business model.  And the smart TOEFL tutors are planning for when the NABP starts accepting the Duolingo Test.

A final observation re: market share.

In the relatively near future, the Duolingo English Test will likely be accepted by governments around the world for visas and study permits.  This change won’t happen in 2024 or 2025, but it will probably happen within the next five years. Certainly within the next decade. When this occurs, demand for the IELTS, PTE-A and TOEFL tests will decline precipitously.

Few people will opt for a $250 test when a $65 test is accepted by the same authorities and is just as respected (or more respected) by the universities they wish to apply to.

As I indicated earlier, I assume that work on the next-gen IELTS and TOEFL tests is already underway.  If it isn’t… now is the time to break the glass and get started on tests that can compete.  Remember that it takes quite a long time to develop a test – the DET was in development for more than six years before it went mainstream in 2020, and the TOEFL iBT was developed over about twelve years before it launched in 2005.

Now in open access from “Language Testing” is an excellent article about test score comparison tables by Ute Knoch and Jason Fan. It explores the concordance tables provided by the makers of the TOEFL, IELTS, PTE-A and C1 tests.

This is a wonderful area to explore in 2024 – most readers are probably aware that the topic of concordance tables has come up quite a lot in recent years as the English test market has become more competitive.

The article suggests that when it comes to creating score comparison tables, best practice principles are not always fulfilled. It notes:

“Our findings indicate that the information provided on the test provider websites about concordance tables is often vague or insufficient. Test users are not always provided with the research underpinning these concordance tables. When such research is provided, it tends not to fulfill the good practice principles and is usually presented in formats not easily accessible to non-specialist test users.”

And also:

“…the sample sizes are generally too small to provide robust score comparisons. Basic information is often not provided, such as concordance results for subsection scores (which are crucial for the requirements for Australian migration and other policy-makers), the number of observations at different score levels, and their standard errors. Test users are not usually informed about the potential limitations of using published concordance tables.”

But why? Well…

“At the moment, there is little motivation to invest in more robust concordance studies due to the absence of regulatory oversight and minimal demand for high-quality work from test users. It is also important to note that concordance tables are one site in which competition between test providers manifests, who may have a commercial interest in lowering their test scores to make it easier for applicants to achieve certain test score requirements.”

There is much more in the article. Do check it out.

Some anecdotal evidence of changes in language testing for university admissions.

Twice a year, first year international students (undergraduates) at Columbia University reach out to me for tutoring while they are preparing for Columbia’s placement test (the ALP Essay Test).  Basically, students whose first language is English take a test to determine if they’ll have to take language classes in addition to their normal courseload.  I help them get ready for that test.  I’ve worked with about twenty such students since 2022.

Of course, I take the opportunity to quiz them about the test scores they submitted when they applied to Columbia in the first place.

Here’s the thing. All but one of my students submitted Duolingo English Test scores when they applied to Columbia.  One student submitted IELTS scores.  None submitted TOEFL or PTE-A scores.

Pre-pandemic, I think, most of them would have submitted TOEFL scores.  A few would have submitted IELTS scores.  Things have changed a lot since then.  They may continue to change.

(PS:  None of them submitted SAT scores)

British Council’s 2022-23 annual report is now available. I believe it covers the twelve months ending March 2023.

British Council delivered 1.8 million IELTS tests in this period, a 12 percent increase from the year ending March 2022. Compare that to IDP’s most recent annual report, which mentioned just a 1% increase in the volume of IELTS tests delivered (and, notably, a 5% decrease in testing revenue in India).

Remember that BC doesn’t do IELTS in India nowadays (they ceded that market after selling their interests there to IDP in 2021). It looks like ROW is a growth market for IELTS, while India is somewhat flat.

Now in preprint from Daniel Isbell and Nicholas Coney is an article examining how English Language Proficiency tests are used at 146 research-intensive universities in the USA.  It examines which tests are used for admissions, what cut scores are used (and they compare across tests), and how subscores are used in admissions decisions.

The authors learned that the TOEFL iBT, IELTS, Duolingo, and PTE-A tests are most widely accepted, and in that order.

A few fun bits:

  • The TOEFL iBT is accepted for unconditional undergraduate admission at 135 schools, the IELTS at 133 schools, the Duolingo at 110 schools and the PTE-A at 61 schools.  I suppose this will be a priority of the folks at Pearson in the years ahead.  Though the test has (I think) moved into the #2 spot worldwide in terms of test taker volume, they still have plenty of room for growth in this area.
  • For unconditional graduate general admissions, the numbers are a bit different. The TOEFL is accepted at 117 schools, the IELTS at 116, the DET at 62, and the PTE-A at 54.  Obviously both the Pearson and Duo folks may wish to prioritize this area.
  • I was very pleased to see that the TOEFL CBT, which ceased to exist in 2006 is still accepted for unconditional admission to 11 undergraduate programs and 10 graduate general programs.  The TOEFL PBT, which was discontinued in 2017, is even more popular.  I suppose ETS ought to prioritize communications with score users in the years ahead.
  • As I have noted in my “score requirement tracker” posts, Duolingo cut scores have not always kept pace with revisions to their score concordance tables.

That’s just the tip of the iceberg. There is some really wonderful data here, so do check it out.

Note, of course, that the above figures may have changed since the time the data was gathered.

The folks behind the IELTS recently published a head-to-head comparison of the IELTS General and the new PTE Core, encouraging individuals on the journey to residence in Canada to opt for the former product.

The comparison is a very eloquent defense of their product, but it highlights some of the challenges that the so-called “legacy” test makers face when dealing with competition from newer tests.  Specifically, a lot of the purported benefits of the legacy tests may be considered somewhat antiquated by test-takers in 2024.

For instance, the article notes that the IELTS can be taken on paper if one prefers.  I’m really not sure that the paper option is a big selling point in 2024.  It later suggests that the IELTS is better because it doesn’t use any AI.  I’m not sure that is a big selling point in 2024, either as people really like AI nowadays.  There is also some stuff about the decades-long legacy of the IELTS, which test-takers probably don’t care about one bit.

The article concedes that the IELTS is 55 minutes longer than the PTE core, noting that the IELTS is  “a bit longer, but we promise, we’re worth it! – we test the skills you need to succeed so you can feel confident starting your new life in Canada.”  Maybe in the distant past people thought about the positive washback of their test prep, but I’m not sure they view tests through that particular lens nowadays.

On the other hand, there is some very valid stuff about how it can be distracting to speak into a room with many other test-takers present.  That really is something people worry a lot about.

Anyway.  Competition is very good for consumers.  I really do hope that work began on the next-gen IELTS and TOEFL tests at least a few years ago. I want them to appeal to young test-takers.  Despite my sometimes dismissive tone, I really don’t want those products to lose TOO MUCH market share to newer tests. That would be bad for consumers in the long run.

 

Wonderful new article (free to read) in “Educational Research and Evaluation” this week by Rachael Ruegg.  It examines the IELTS test and the in-house language test of a certain New Zealand university, and explores how well scores from those tests can predict academic success at the undergraduate level.  The article suggests that the predictive ability of reading, listening and speaking scores from both tests are equal.  But writing scores? Well, the author notes:  “IELTS writing scores demonstrated a non-significant negative effect on academic achievement, while EPT writing scores were strongly significantly predictive of academic achievement.”

Check out the article for a detailed comparison of the writing tasks on the IELTS and those on the in-house test.  The author suggests that the in-house test requires writing that is more challenging and requires more critical thinking.  They note:

“In the EPT writing test, both tasks require longer and more complex writing than the IELTS versions. This may contribute to their performance in university study, which usually involves producing written assignments that are significantly longer than the IELTS minimum length of 250 words (for the essay task) or 150 words (for the data writing task). Rather than specifying a word limit, students who sit the EPT writing tests are encouraged to write as much as they can within a 45-minute time frame for each essay, but with an emphasis on quality over quantity.”

The author suggests that the IELTS could be improved by increasing the length of its writing section:

“Suggested changes to the IELTS academic writing test include extending the time allowed and required length of both writing tasks, requiring greater complexity in written texts and demonstration of critical thinking skills in English, all of which are likely to provide more robust evidence of sufficient language proficiency for university study.”

That’s quite a suggestion in a world where English proficiency tests seem to be going in the opposite direction – embracing shorter and shorter writing tasks.

I’m reminded of how many international students at Columbia University are required to take a two-hour writing test before their studies begin (on top of the university’s TOEFL/IELTS/DET requirement). Someone ought to study the usefulness of that test.

IDP Education has released its fiscal results for the half-year ending December 31, 2023. The report indicates that the IELTS was taken 902,000 times in those six months, a drop of 12 percent compared to the same period last year. The report indicates that this drop was due to lower test volumes in India. I’ve written in this space about how IDP faces stiff competition from Pearson in that market. Interestingly, the report indicates that outside of India the number of tests administered actually increased by 17 percent.  Good lord, the drop in India must have been pretty big if the overall volume declined even though sittings in the rest of the world increased by a huge amount.

Investors will be happy to know that revenue from student placement is way, way up. That is due in part to a 56 percent increase in Canadian placements.

IDP’s share price jumped about 15% on the release of the report, but those gains were all wiped out within 48 hours.  Perhaps some of the early buyers weren’t aware of recent changes to the business of international education.

(Note that the above figure does not include IELTS tests taken via the British Council)

The Indian Express reports that new Canadian immigration rules will likely mean the end of the so-called “IELTS Wedding.”

When I summarized the changes a few days ago I didn’t even mention that spouses of most students will no longer be eligible for an open work permit in Canada. I guess that change is more meaningful than I earlier thought.

As regular readers know, a play about studying for the TOEFL won the Pulitzer Prize for drama in 2023. Perhaps we can expect a romantic comedy about the IELTS in the years ahead.