Some popular tampering of data and graphics…

In Data Marketing We Trust. Of Beauty and (Self-)Deception.

Leveraging big data in HR tech has become increasingly popular in recent years thanks to its great potential. However, the current benefits often lag behind the marketing promises of vendors. Because of the highly unstructured nature of job-related data and the numerous factors that influence jobs, skills and labor market dynamics, finding good quality data and analyzing it reliably is still a huge challenge. In addition, many of us aren’t quite as data savvy as we’d like to think. What’s more, as human beings we’re highly susceptible to certain phenomena; that vulnerability is often taken advantage of in marketing, especially in an industry that tends to promise more than it can deliver. We all fall prey to the crudest tricks and manipulations time and again; these cheap ploys simply work too well. So, in this post, we want raise awareness by taking a closer look at two of these phenomena: social proof and misleading data visualizations, and how they can be misused. Hopefully, after reading this, we will all ask ourselves next time whether those enticing numbers and pretty graphics really are trustworthy enough to use or share without thinking. Or if we really should buy that product just because 7 out of 10 companies of some type have already made the same mistake…

In people we trust

Imagine you’re on vacation, and you want to eat out at a restaurant. There’s this one place you’d read great things about and you want to try it out. But when you get there, it’s almost empty, and the restaurant next door (that you’d never heard of) is almost full. Which restaurant will you choose? Most likely the one that’s bustling. Why? Social proof. A term coined in 1984 by Robert Cialdini in his book Influence, social proof is also known as informational social influence. It describes the principle that when people are uncertain, they will typically look to others for behavioral guidance. This is based on the instinctive assumption that the (social) group displays optimal behavior to achieve the best possible outcome. One consequence of this, which is used excessively in marketing, is that our minds use social proof as a shortcut equating popularity with quality, trustability and other positive attributes. In fact, the effect of this phenomenon can be so powerful that any other indicators may be completely disregarded—no matter how contradictory. Basically, our minds are wired to crave efficiency, so any perceived shortcut is subconsiously strongly favored over the alternative of strenuous research and carefully weighing pros and cons.

Social proof comes in many forms from experts, celebrities, friends, customers, followers, etc. The thing is, at the risk of stating the obvious, social proof is not proof. For neighboring restaurants, it’s more about luck than marketing: Whoever gets the first few customers of the evening, wins. Even if the food is terrible. In marketing, there are many ways to obtain or reframe social proof for optimal effect.

Let’s say you’re considering a certain well-established data provider for your HR tech, and on the homepage of the provider’s website you see that “67 of the Fortune 100 companies look to” that company for their data needs. Then you’re likely to believe it will work well for you too. And anyhoo, no one ever got fired for hiring IBM, right? (Just to be clear: We’re not actually talking about IBM here.) Throw in some trigger words like uncertainty, need, trust and secure, and it may well be virtually impossible to resist the bait. Even if the data provider’s sole argument is that you can trust their data… because a few dozen other companies did. But what does “look to” even mean? Elsewhere on their website you may find that the 67 companies also include one-time consulting gigs (and “consulting” may be used very loosely) as well as past clients, i.e., clients who stopped using their services. The facts have been reframed to boost social proof. However, by the time you come across this tidbit, your brain may well already be so biased toward the provider that it refuses to adequately process any information that goes against what your brain wants to think. Hello, confirmation bias.

 

 

Now, social proof may get this company’s foot, or even an entire leg, in the door. But they’re a data provider, so they still need to sell you on their data. How? According to the research, with visualizations – especially if they’re pretty.

In beauty we trust

Visualization is an essential tool for analyzing and communicating data. Well-designed visuals can help people form a greater understanding of often vast amounts of highly complex data. However, there is a dark side to data visualizations, particularly when they are aesthetically appealing: Studies show that beautiful design significantly increases trust in the presented data. Not only that, but it seems that our trust bias in favor of beautiful visualizations cannot be reduced by better education or even specific knowledge of misleading visualization techniques. On top of that, studies like this one have shown that misleading visualizations are indeed highly deceptive. Of course, these misleading charts are often created with no ill intent by people who simply do not realize what they are doing. However, there are also many individuals and organizations who exploit these techniques because they have an agenda. Be it for financial, political or any other gain. Thus, in the hope that – despite scientific evidence to the contrary – raising awareness will lower your susceptibility to deception, here are five of the most common ways to misrepresent data and mislead audiences:

1. Manipulating the baseline

Truncating the y-axis is an easy way to exaggerate changes and differences. Sometimes, the resulting chart is so obviously wrong, it’s just silly (or offensive):

 

Source: https://twitter.com/reina_sabah/status/1291509085855260672

 

More often, though, the manipulation is not quite so obvious to the untrained eye or less critical reader, particularly if combined with a sensational headline. For example, the bar chart below – produced by the Senate Budget Committee Republican staff – gives the impression that the number of federal welfare recipients in the US is spiraling out of control. It looks like the number more than quadrupled over the course of two years.

 

 

With the correct y-axis, shown below, one can see that the situation is far less dramatic.

 

 

In addition to the misleading visualization, the headline is misleading as well: The fine print at the bottom of the original chart states that the numbers include anyone living in a household in which at least one person received a program benefit. This includes individuals who did not themselves receive government benefits. On the other hand, the Survey of Income and Program Participation typically provides information on the number of households (not individuals), or on the number of individuals that either receive benefits from the program or are covered under someone else’s benefit (e.g. children). Since the source of the data is not stated clearly enough, it’s anyone’s guess as to what (or who) to trust here.

 

2. Manipulating the axes

This visualization tactic of changing the scale of a graph is often used to minimize or maximize a change. For instance, the graph below shows the average global unemployment rate from 1991 to 2021. The scale goes from 0 to 100, i.e. all possible percentages. This doesn’t make sense for the data. Instead, it serves to flatten the line and convey the idea that the unemployment rate hasn’t changed much in the past 3 decades.

 

Data source : https://data.worldbank.org/indicator/SL.UEM.TOTL.ZS

 

Here’s the graph again with a more meaningful y-axis, showing quite a different dynamic.

 

 

A more challenging example is the graph below from the World Inequity Report 2018, which illustrates the growth in global real income from 1980–2016.

 

 

The purpose of the graph is to show the distribution of growth across income brackets. It gives the impression that the high growth in the top income brackets is broadly distributed across the wealthier segment of the population. This is misleading. The top 1% of the population takes up only a quarter less space along the horizontal axis than the bottom 50% (or rather, the almost-bottom 40%: the horizontal axis starts at 10%). This is because the horizontal axis suddenly switches from a linear scale to a logarithmic scale: every 10% of the population is represented by the same distance along the axis. But once the graph reaches 99%, this changes abruptly to a logarithmic scale, in which smaller and smaller segments of the population take up equal width along the horizontal axis. At the far-right side of the graph, less than 0.001% of the population corresponds to a region the same size as used to represent 10% of the population across most of the graph.

The alternative graph, buried in the appendix of the report and shown below, is plotted on a linear scale throughout.

 

 

This graph shows the distribution across income brackets much more clearly: it indicates relatively modest growth in income across the right half of the graph, concluding in a sharp increase for a very tiny fraction of the represented population.

 

3. Cherry picking data

Cherry-picked data is probably the trickiest manipulation to spot, and particularly powerful. Shifting the data frame can create a very different story. For instance, if you wanted to create the impression of a debt crisis in the UK, you could post a graph like this which shows that UK public sector debt as a percentage of GDP has more than doubled since 2002. For an even more dramatic graph, you could post the total level of debt, which has increased even faster than debt to GDP.

 

 

However, if you place the growth in real debt to GDP in a broader context, say, by considering the data since 1900, the rise looks much more modest.

 

Data source: Bank of England – A millennium of macroeconomic data  and ONS public sector finances HF6X.

 

And here’s an example from a veritable master of cherry picking – with an apt response:

 

 

Cherry picking, as all these techniques, can of course be done unintentionally. This very often occurs when people are affected by confirmation bias. We touched on this in one of our previous posts.

 

4. Going against conventions

This chart is from a UN report on job polarization. The bars are set at 15 as a baseline, and the length of each bar represents the value of the corresponding data point minus 15. I have no idea why this visualization was created in this way; it simply defies all logic, making it very difficult to read and completely distorting the data visually.

 

 

Here’s what the chart looks like when adhering to conventions.

 

 

The following chart is not quite so confusing, but also goes against conventions. Proportional area charts are hard to read, especially ones with circles, and should be used sparingly. They are typically used to visualize the relative sizes of the data: each circle area is proportional to the value it represents. Going by the relative sizes of the circles in the chart below, the number of exact matches among all workers seems pleasingly high.

 

 

However, when you match the circle areas to the numbers, the finding the perfect (employee) match looks much more like the proverbial needle in a haystack.

 

 

And maybe this e-commerce shop should have opted for a more conventional visualization like the one a little further down…

 

 

5. Using the wrong visualization

The classic and most obvious example of an inadequate visualization is the pie chart for values that do not add up to 100 percent, for instance, the results of a poll where participants can choose multiple answers. Just because a pizza is round like a pie does not mean this is a good way to show your results.

 

 

However, there are much more subtle or complex examples of using the wrong visualizations. In midst the so-called “slide wars” between the EU and the UK, where both sides traded PowerPoint-style slides to influence post-Brexit trade talks, the European Commission was accused of publishing a misleading graph which overstated the UK’s reliance on trade with Europe.

The European side used a large, red blob to highlight how much of Britain’s trade goes through the EU. Statisticians then pointed out that the graphic wildly exaggerated the relative size of Britain’s blob compared to that of other countries and blocs.

 

Source: https://ec.europa.eu/info/sites/default/files/cwp-20200218-trade-geography_en_0.pdf

 

A European Commission spokesperson said: “The chart was generated with an Excel chart tool, based on data from Eurostat. The width of each bubble is proportionate to the total trade of each country.” Right there is one of the key problems: the convention is that the area of the bubbles is proportionate to the represented value – not the width. And this has been swept under the rug by using the ambiguous term size in the note on the chart. In addition, the size of the bubbles delivers no added value in terms of information because the two quantities share of total EU27 trade with 3rd countries and (absolute) total trade volume with EU27 are directly proportional. But the authors of this chart clearly have an agenda. This suspicion is further fueled by the fact that they have cherry picked the data: UK ranked #3 of EU trade partners, so they have conveniently excluded the top 2 partners US and China and instead included a seemingly random selection of trade partners from ranks 3 through 38. Another, albeit minor detail: the “approximate distance” is, well, very approximate.

Here’s the chart with the right bubble sizes and distances (and without the confusing colors).

 

 

However, one does wonder whether the approximate distances are truly necessary, or just complicate the chart. If you insist on geographical information, why not use a world map?

 

 

If you don’t need the geographics, the simplest visualization is often the best option.

 

 

But then again, it may not help your agenda. And it certainly won’t help you to “blind your readers with science” like the following example does very effectively.

This is an excerpt from a table in a McKinsey report on independent work. Each row, representing a country, has 12 numbers squeezed into 6 columns. It’s too much data in one table and what’s more, the data is packed into mini charts that complicate the visualization without adding any informative value. It may look clever or cute to some. But it’s not. It’s intensely busy and confusing.

 

 

In data we trust

But we can’t blame everything on the visualizations. When it comes to data itself, many people still hold the misconception that more is more. That’s why marketing like this works:

 

 

But here’s the thing. Contrary to popular belief, new data alone doesn’t necessarily equal new intelligence. Much of the gathered data lacks the quality needed to derive the promised insights. Especially unstructured data like resumes and online job postings. There’s whole slew of issues with this type of data, which you can read about in more detail in our series of posts here, here, and here. For now, we’ll take a brief look at two key concerns: This data is typically skewed towards certain industries, and overestimated in terms of usability. For instance, the World Bank Group partnered up with LinkedIn to investigate to what extent LinkedIn data can be used to inform policy. In their 2018 report, one can see that the data is highly biased toward the ICT industry and high-income countries.

 

 

Out of 17 ISIC industries, 12 had such low coverage that they had to be completely disregarded in the report, including the currently highly relevant industries human health and social work activities, transportation and storage, and accommodation and food service activities.

Skewed data is not the only issue. Rendering unstructured data machine-readable, for instance, by extracting and classifying skills, job titles and other relevant information, is extremely challenging and often faulty. Examples of this are abundant. For instance, the European project Skills-OVATE run by Cedefop and Eurostat provides publicly accessible data on skills and occupations extracted from online job postings across Europe. According to their 2021 data, knowledge of fine arts was requested in around 1 in 15 online job postings for refuse workers in Europe, and 1 in 25 ask refuse workers to create artistic designs or performances. Sounds like there is much more to being a refuse worker than most of us would assume..

 

 

(By the way, this system is set up and managed by the data provider in the marketing example above.)

You may brush this off as irrelevant. But once this data is in the system, it’s there to stay. And it leads to results like this:

 

 

It’s hard to imagine that 11% of job postings for actor/performers include food delivery as a skill. If you want a less exotic example:

 

 

The suggested skills may not be quite as absurd as knowledge of fine arts, but they are just as irrelevant to refuse collection.

It would be wise not to attempt to “understand the economy”, “describe the talent that businesses need”, or “the abilities that local people have” using data of this quality. Instead of vast amounts of mediocre data, look for a sufficient amount of representative smart data: high-quality data that’s been validated and contextualized with intelligent annotations – ideally by humans, but that’s a whole other topic.

A final word: None of these data visualization and marketing tactics that we’ve covered here are inherently evil, or even necessarily wrong for that matter. Data visualizations are all about conveying the data’s story and like any other story, people can take artistic liberties. Handled with care, data itself can have many interesting stories to tell. But don’t let yourself be fooled into believing a story the data is not actually telling you.

Four-day workweek? Forget about it, the countdown to the implosion of the labor market is running anyway

4-tage-woche

Zurich airport, 06:43 on a Friday morning in July, which has so far been marked by scorching heat. Baggage handler Mario V. is enjoying a short smoke break between LX 243 from Dubai and LX 1952 to Barcelona. Although “enjoying” seems to be a somewhat exaggerated choice of words these days. Since the start of the summer season, Mario’s working days – like those of most of his colleagues on the ground and in the air – have been marked by a lot of overtime, chaotic instructions from his employer and rude – or downright insulting – remarks from passengers who are now going on vacation in droves. All this stress is caused by the decision of several airlines to cancel hundreds of flights this year and to struggle through the remaining ones with a number of staff that has been drastically reduced since the pandemic (so much so that there is now a severe shortage). While scrolling through the news Mario reads an article and its comment section discussing the current shortage of skilled workers. Under calmer circumstances he might have laughed about what is written there but given the current situation it brings his blood to the boil even at this early hour: Some high earners, spoiled by home office and free coffee, pride themselves on demanding things like the 4-day workweek or the 6-hour working day. They claim this “especially for people in professions with hardly any rest for the head,” because in those “it is only the result that counts anyway” (and not the working time). ” Well, thanks a lot for your solidarity,” Mario thinks to himself. Do they actually realize that he has to reload their suitcases by hand and must not make a single mistake; that his work combines mental work and back-breaking work, and that his tasks demand the very performance of a Super Mario? The much-vaunted idea of “learning from errors” obviously does not seem to extend beyond some people’s own office corner. As soon as someone outside of their working environment has a mishap, tolerance falls by the wayside – this happened just recently when Skyguide’s air traffic control technology had failed. Nobody seems to be willing to lobby for workers like Mario here. Rather, these people like to portray his kind as an incompetent, lazy bum while on their way to Mykonos. Miserably, the baggage handler returns to the cargo area to prepare the loaded baggage cart for the airfield…

4-day workweek an impossibility? In any case, it shows a lack of solidarity

The fictitious example above illustrates in a simple way how many of the accumulating problems in the global labor market have now arrived on our doorstep. One wonders: Why has no one anticipated this disaster in the airline industry, so that Mario and many other workers do not have to resort once again to the ultimate (but unsustainable) leverage of striking? Or, even worse, to face the threat of an increase to 50 weekly working hours…

Exactly such an increase (to 42 h) has just been proposed by the German industry president, primarily to counteract the much cited, growing scarcity of skilled workers. The proposal was met with strong criticism and the proponent was accused of being stuck in the past. At the other end of the spectrum, experts (and wannabe experts) are talking their heads off about alternative, “contemporary” working time models such as the compression of a full-time job to a so-called four-day or 36-hour workweek. Euphorically, they rave about the few positive examples that actually exist, such as Microsoft Japan or the microcosmic island state (i.e. special case) Iceland. Anything to manifest your own wishful thinking… Leaving aside the injustice mentioned at the beginning of this article, namely that there are many employees in sectors for whom such demands are simply not feasible because their work is not based solely on productivity but also on on-call duty, compulsory attendance without any Monday-Friday/9-5 basis or flexitime. Think of the hospitality industry, nursing, retail, waste management or teaching. Or indeed aviation. Say, one chooses to ignore this lack of solidarity and how it contrasts with these already privileged employees’ cry for even more flexibility. Even so, the question arises as to how on earth this calculation is supposed to work out if we look at the current relevant labor market indicators.

Shortage of skilled workers and reduction in working hours: the math doesn’t add up

A side effect of the restriction to four working days per week, which is often mentioned by supporters, concerns the creation of additional jobs. In view of the current labor market situation, however, this would likely bring about even more problems instead of solving them – the term “contemporary work schedule” is therefore only partially apt. After all, it is by now a well-known fact that in Switzerland – as in several other countries – there is currently many times more work (in the form of jobs, i.e. hours to be worked by people) than people looking for work, and that this trend could intensify even further. We feel the effects of this during events such as the current one in the aviation industry: many workers do not (or no longer) want to put up with the poor pay and working conditions in certain sectors and choose or switch to other fields of activity under better terms. This happened in thousands, especially at the height of the pandemic, which created many additional unfilled positions.

All of this seems to be ignored when people praise alternative working time models from their cozy home office or nicely located, air-conditioned office room. The same goes for the fact that the unemployment rate and labor force participation are currently at all-time lows and highs respectively, meaning that most people who are willing and able to work already have a job. The same applies also to the circumstance that the baby boomer generation is about to retire without any expectation of an equal number of (let alone more) people coming into the workforce. In short, the countdown to the implosion of the labor market is running. Moreover, we will all feel its effects even more violently than in its current form at the check-in and baggage carousel on our way to vacation. With such a starting position, it is simply nonsensical to demand that everyone works one day less. Mathematically, this can be shown relatively simply using a rule of three:

 

janzztechnology_4-tage-woche-en

 

The crux of the matter is that there is not or will not be 20% more workforce available of what is already needed today (in certain industries, actually not even as much as is needed now, since there are already a lot of vacancies).

Are there any conceivable alternatives that would actively combat the shortage of skilled workers, but wouldn’t present us with a mathematical impossibility? One option would be to promote large-scale immigration for employment purposes; but not only for the highly qualified skilled workers, as certain politicians like to demand time and again. However, with respect to this idea it should be noted that recent evaluations indicate that in certain industries the recruitment opportunities for immigrants are exploited to the max, even if the free movement of persons is fully utilized. Thus, such a move would only partially counteract the current demand. This leaves us with the option of taking a completely fresh look at the labor market in its current fundamentals, with the aim of correcting certain malfunctions…

On his way home from the early shift at the airport, Mario thinks back to his moment of resentment in the morning; thanks to further flight cancellations it was not to be the only one of the day. It becomes clear to him once again that his industry and possibly many others are on the verge of a crash landing, yet no one seems to care in the long run. Neither “those at the top” of the airport and the federal government, nor the thousands of travelers who run in front of his baggage cart every day. But right now, Mario is so tired of non-stop standing, sorting and hauling that he’s postponing this headache to a later moment… Let’s grant Mario his well-deserved rest and let him roll out some more thoughts on topics like labor market planning and consumer responsibility in the next few posts (pun intended).

 

At JANZZ.technology, we collect a wide range of labor market information, including labor supply and demand, through a variety of projects. We do this not only in Switzerland, but also in collaboration with the Public Employment Services (PES) of countries around the world. Since 2010, this has enabled us to develop market-leading evidence-based solutions. Not only are our systems efficient, scalable and extremely powerful, they also rely on ontology-based semantic matching. Furthermore, all our tools provide unbiased results in line with the OECD principles for AI. We are committed to stimulating fact-based discussion and raising awareness in all areas related to labor markets and processes. To learn more about our services, please contact us at info@janzz.technology or via our contact form, or visit our PES product page.

4 unsung challenges of digital labor market solutions you must address

JANZZtechnology Unsung Challenges ILMS

A global economic slowdown, rapid digitalization, ageing populations, informal labor, skill mismatches, discouraged workers, structural shifts and increasingly dynamic career paths… In labor market governance and management, there is no shortage of complex challenges to overcome. To address these challenges, many governments have introduced public employment services (PES) and active labor market polices (ALMPs). There is no simple answer on how to do this, especially given the complexity and unique challenges of each labor market. However, one integral step towards more effective labor market management is instituting an advanced labor market information system. In a well-implemented system, accurate, relevant and timely information flows between all relevant stakeholders. This facilitates informed decisions and appropriate action—provided the information is reliable, valuable, and as complete and up to date as possible.

A key element of any such system are digital solutions to gather, validate, analyze, and disseminate information related to the labor market. Integrated labor market solutions (ILMS) combine these digital solutions into one system. This includes solutions that provide services such as job matching, career guidance, and other functions covered by the PES. Any complex system inevitably presents technical challenges, and this is also true for an ILMS. Apart from ensuring the core technologies and features work as they should, there are overarching challenges such as accessibility, compliance with data and privacy laws, connectivity and interoperability, to name just a few.

Fortunately, with good project management and sound requirement engineering, the larger challenges of the technical implementation typically reveal themselves early in the process. Therefore, they can be addressed before they jeopardize the success of the project. However, other, equally essential challenges are surprisingly often overlooked or ignored – with potentially fatal consequences for the project.

JANZZtechnology Unsung Challenges ILMS

1. Mindset

Large-scale changes like the implementation of an ILMS disrupt the everyday processes that PES employees and clients have grown accustomed to. Such changes fundamentally alter the culture and mindset of the organization and can be overwhelming for many involved. This often leads to disconcerted or disgruntled staff, jobseekers, employers, or other involved institutions. If the value of such a system is not effectively conveyed, there may be apprehension or even pushback. However, it will be highly detrimental to the project if these key stakeholders are unwilling or unable to endorse the new processes, or to collaborate and share information. Therefore, this challenge must be addressed. A multifaceted approach can accomplish this, including change management within the PES and other institutions, campaigning and building trust with the public, as well as critically reviewing and adapting policy and leadership.

Another key challenge for project owners and commissioning authorities is maintaining an open, yet critical mindset and the ability to take bold decisions. As the saying goes, good decisions come from experience. Experience comes from bad decisions. So, learn from (other people’s) mistakes and go ahead. When embarking on a complex project like this, there is often substantial pressure to do what everyone else does, i.e. simply purchase the same system or technologies that some other country used more or less successfully. But again, each labor market poses its own unique challenges. What works in one country or labor market may not work in another. Determining the right system thus requires pinpointing the key characteristics and challenges of the labor market, economy and culture at hand to obtain a clear idea of which problems the ILMS can and must solve. With this clarity, the project owner can review comparable projects critically, assessing similarities and differences in labor market, culture and political framework, as well as studying the pitfalls and mistakes made – and, if necessary, advocate a solution that is different to what others are pursuing.

JANZZtechnology Unsung Challenges ILMS

2. Community

An ILMS can only be successful if it is used. By all relevant stakeholders. One typical component of an ILMS is a job platform. When we implement these projects for public employment services, we very often see a strong focus on job platform features for jobseekers and PES staff. Of course, there are many aspects that must be considered for these user groups, including ease of use for a diverse audience, service delivery to users with limited digital access, accessibility for users with disabilities and so on. However, all these desirable features are of little use if the job platform is not populated with job postings. Therefore, it must first and foremost be attractive for employers. This requires infrastructure and staff to attract and assist employers looking to hire, as well as deploying an ILMS designed to make job posting and candidate sourcing as simple and effective as possible for businesses. It may also involve policy decisions, advocacy and campaigning.

In addition, an ILMS should also cater to users from educational institutions who leverage the data in the system to align curricula to market needs. In return, these institutions could feed data on their offering into the system to make it readily accessible to both job seekers and policymakers – again to deliver actionable intelligence and facilitate informed decisions. Similar considerations also apply to other components of an ILMS.

Another crucial challenge arises in federations where the federated units have a certain degree of autonomy. These units must agree on and adopt the ILMS – albeit with regional adaptations to cater for each unit’s individual labor market characteristics and needs. This way, the ILMS can help break through information silos and provide reliable and comparable data and services at both the federal and regional levels. For instance, to inform policy at both levels, and to improve job opportunities by removing regional barriers.

JANZZtechnology Unsung Challenges ILMS

3. Capacity

Sound capacity building is one of the aspects of ILMS projects that is most frequently overlooked or only partially and insufficiently executed, often leading to project failure. Recognizing the important role of capacity building is therefore crucial to ensure the ILMS reaches its full potential. Specific and targeted capacity initiatives need to be developed on multiple levels. At the individual level, this involves training PES staff, analysts and other involved institutional actors, as well as educating private and business users. At an institutional level, internal policies, processes and organizational structures should be reviewed and adapted in all key organizations. At a systemic level, the overall policy framework must be designed or reformed to facilitate effective and sustainable operations and interactions between individuals, organizations and the ILMS.

JANZZtechnology Unsung Challenges ILMS

4. Sustainability

Finally, a key goal of the project must be to sustain the ILMS over the long term: at least five years, preferably ten or more. Accomplishing this requires a comprehensive approach to project implementation covering several dimensions. For instance, funding and project ownership should be as independent as possible from changes in direct (political) leadership. The ILMS infrastructure and technologies should be maintained and regularly reviewed, updated and enhanced or adapted to changing requirements. In addition, focus on continued community participation must be maintained. This can be achieved through techniques and principles such as sustained supportive and transparent service provision and leadership, periodic reviews of user needs and satisfaction, campaigning, and other tools.

 

Overall, addressing these challenges is about creating an environment for a system where all stakeholders are both motivated to contribute and able to draw maximum benefit – thus strengthening the labor market and the economy and helping communities thrive.

If you’re looking to implement a successful ILMS in your country, get in touch and benefit from JANZZ.technology’s in-depth expertise. We are happy to assist and advise you right from the start.

Or visit our website and discover the advanced solutions we have created for public employment services and watch the explainer video for our integrated labor market solution JANZZilms!.

Would you like to read this article in indonesian?

Fear of the machine, rage against the machine? Why we are so afraid of AI in recruiting (and what could be done about it)

A new study from Germany shows that the use of artificial intelligence (AI) in job application processes is widely rejected and generally stirs up negative emotions in potential applicants. There were also numerous objections raised by the respondents. Depending on the context, fears of programmed bias or the negligent handling of personal data may well be justified. In principle, however, it would be well possible to dispel many of these concerns if employers and software providers made more efforts to ensure transparency and explainability in the use of AI in recruiting. In addition, it would be useful to have more comparative studies on the respective performance of humans and machines in order to curb the obvious resentment towards AI in the HR sector. Since the use of artificial intelligence in the recruitment process will be unstoppable, we are going to break down the massive black box around HR systems and clarify which requirements have to be fulfilled for a more successful deployment of algorithms & co.

Fear of the unknown

In total, around 65% of the study participants associate negative things with the idea of AI in recruiting. Since this represents a clear majority, it is especially interesting to consider the underlying reasons for this result. About two thirds of all respondents show no trust in decisions made within a hiring process using AI. The biggest weakness cited is that an automated process is impersonal. At the same time, however, only a small minority (6.3%) realized any contact with AI in recruiting in the past – although this is already a reality in many places. [1]

 

Facts and figures directly taken from the study

janzztechnology_rage-against-the-machine_graph-1-en

All of these findings point to one central problem: As long as there is no transparency regarding the topic of artificial intelligence in HR, it is impossible to convince job seekers and employees of the benefits of such processes. The poor image of AI in the recruiting process is therefore primarily due to a fear of the unknown, which in turn manifests itself in two ways. Respondents do not seem to know exactly how AI-based decisions are made. Neither do they realize how these decisions will affect the actual application process and their chances of being hired.

The X factor in AI

It is up to both software companies and employers to clarify this so that candidates are more aware of where and how artificial intelligence is used in HR. There are still many recruitment tools on the market whose machine-learning-based results cannot be adequately explained, replicated, or corrected by the developers when needed. As a result, such black-box processes also deny applicants a genuine option of consent for the collection, processing, storage, and deletion of personal data, which can lead to serious legal problems. Thus, in many places, both the requirements of the EU’s General Data Protection Regulation (GDPR) and the Organization for Economic Cooperation and Development’s (OECD) AI principle of transparency are not being met.

The answer to this problem is provided by so-called explainable artificial intelligence, XAI for short. In recent years, it has established itself as a proven approach to break open the black box around systems based on deep learning and artificial neural networks. At JANZZ.technology, we have been working with such explainable models for quite some time and, thanks to their combination with ontology-based semantic matching, we deliver numerous powerful solutions for all HR and labor market management processes.  It is of great importance to us that we make our services easy to understand and provide customers with the necessary knowledge about the mechanisms and processes behind our technologies. Our matching tool JANZZsme!, for example, does not simply deliver a rather meaningless matching score between a candidate’s profile and the job posting. Rather, it dissects all criteria into sub-aspects such as skills, language skills or experience, which all have their own, visible score and explain the results in a comprehensible way for both applicants and employers.

A large number of respondents expressed their desire for a personal contact person for queries during the hiring process. As we can see, this demand can be met to a large extent by means of explainable technologies and transparent information about them on the part of HR departments. In response to the finding that AI-based processes in the application process are slandered as being impersonal, it should be noted above all that today, final decisions still lie with a recruiter and this will also remain the case in the years to come. According to our expertise and many years of experience, there still is no fully automated hiring process anywhere that completely excludes human intervention from the process. It is therefore understandable, but unfounded, to fear that you as an individual will be reduced to nothing more than a string of ones and zeros during the application process. Likewise, your soft skills profile will not be completely disregarded.

Human versus machine: A comparison

In fact, we should ask why there is such a desire for human influence in the recruitment process in the first place when, paradoxically, half of the respondents say they fear the embedding of human biases in AI programming. [1] Moreover, the few meaningful comparative studies available on the performance of humans and machines by no means indicate that the former make better decisions in application processes. Another advantage of XAI in the area of recruiting is therefore that with it we get a better picture of the actual performance of automated processes and can quantitatively compare these results with those of manual processes.

Allow us to outline a short example from one of our own use cases. The assignment was to conduct a comparative POC for an international organization to find the most promising candidates for their highly coveted junior positions. For comparison, the selection was also made by the experienced HR managers who usually perform this process “manually” every year for a period of several weeks. Key parameters to be considered when comparing the results included the avoidance of bias, achieving the highest possible efficiency and, of course, finding the most suitable candidates.

The result is likely to surprise a majority of participants from the study described at the beginning: Firstly, when using our APIs, matching tools and parsers the processing time was reduced to a fraction of that of the manual process (3 days vs. 12-14 weeks). Interestingly, a quick process was the third-most frequently mentioned criterion for a positive candidate experience in the study. [1] Secondly, there was no bias at all in JANZZ’s XAI-based decision-making, while the HR managers’ choices showed massive biases in the variables origin, gender, and language – not surprising, given the myriad forms of (unconscious) bias that shape the manual hiring process. To be sure, our replicable process based on binding criteria meant that the objectively best candidates were selected and those did not always automatically meet specific diversity and inclusion expectations. But even such requirements are scalable if desired and can then be applied in a rule-based and consistent manner, provided that this decision is also communicated transparently to applicants. Surprisingly, in the study only 14% noticed one of the main benefits of XAI-based matching. Namely that it makes it easier and more reliable to find a job that actually matches your skills, competences and education. [1] For this purpose, our technology is based on multilingual semantic matching. This approach provides a flexible solution to the problems posed by an increasingly heterogeneous composition of knowledge, terminology and information in CVs and job advertisements, in turn making the matching process a whole lot more efficient.

 

Facts and figures from the JANZZ POC

janzztechnology_rage-against-the-machine_graph-2-en

Overall, our comparison clearly shows results in favor of AI. In order to reduce the existing fears of many potential employees, it would undoubtedly be valuable to have evaluations such as ours conducted on a broader and more regular basis. Still, the following conclusion can already be drawn from what we know by now: In terms of bias in the application process, AI enriched with deep learning and an underlying knowledge-rich system definitely does not perform worse than humans (see also link in last section). On the contrary, it even brings potential advantages such as an expedited process and, above all, objectivity. Moreover, the use of AI in HR is already gaining ground at an unstoppable pace. That’s not surprising, we’d say. Or does anyone see an alternative to cope with the increasing participation rate and movement in the labor market? Due to all these facts, at JANZZ.technology we adhere to the principle “No artificial intelligence without human intelligence”. XAI-based systems provide indispensable help in the tedious and costly pre-processes of manual recruitment and allow human recruiters to focus on the essential; finding the top candidates.

 

The capabilities of AI for HR go far beyond the hiring process, as it can also be used in a company’s strategic workforce planning, for example. If you would like to learn more about our broad range of services or get information about what JANZZ.technology can do for your specific needs, please contact us at info@janzz.technology or via contact form, or visit our product page for an overview of all our solutions. We also invite you to listen to our podcast in which we talk about interesting, related topics. In the current episode, for instance, we discuss the distinction between systems that are “knowledge-lean” and those that are “knowledge-rich” – a crucial difference, if you ask us!

 

[1] IU International University of Applied Sciences. 2022. AI in Recruiting: Emotions, Views, Expectations. The Impact of Artificial Intelligence on the Candidate Experience. URL: https://www.iu.de/en/research/studies/ki-in-recruiting-study/

Strengthening the economy through advanced labor market information systems

janzztechnology_lmis

Today’s changing world places many complex challenges to labor market governance and management: the slowdown of the global economy, the structural shifts and evolving skill demands connected to widespread digitalization, as well as increasingly dynamic career paths with more frequent job switching, geographical mobility and flexibility, and multiple transitioning between education/training and employment.

Advanced labor market information systems are key to improving labor market efficiency

To address these challenges, many governments have established active labor market polices (ALMPs) and public employment services (PES) to help workers find jobs and firms fill vacancies. However, due to the complexity and individual set of challenges in any given labor market, there is no simple answer as to how public employment services should be set up and organized. But a well-thought-out information strategy and infrastructure is certainly critical to the success of any PES. If nothing else, the most recent disruptions have shown that effective ALMPs and PES require agile and flexible frameworks to successfully adapt to rapid and at times dramatic shifts in their labor markets. But even the most agile of frameworks is only useful if it includes a system to identify labor market issues as they arise.

Identifying such issues relies critically on the availability and quality of data, information and analysis. Therefore, establishing an advanced labor market information system (LMIS) is an integral step towards more efficient and targeted employment and labor policies by delivering accurate, relevant and timely information to inform design, implementation, monitoring and evaluation of policies. According to the World Bank, advanced LMIS encompass institutional arrangements between key stakeholders (e.g. policy makers and the education system), collaborative partnerships with private sector actors and advanced technology solutions to gather, validate, analyze, and distribute information related to the labor market that is relevant, reliable, useful, and as comprehensive and up to date as possible.

Combining traditional labor market information with real-time data

Traditionally, labor market information (LMI) was primarily gathered from censuses, surveys, case studies, and administrative data. However, this traditional LMI has a disadvantage that is increasingly cumbersome: lag time. In an ever-faster changing world this carries risks such as policies being outdated before they can be implemented, rendering them ineffective if not obsolete. Therefore, an effective LMIS should also incorporate real-time (big) data from additional sources such as online job portals and networking sites. This type of data is not only much more up to date, it also typically contains more detailed information including job activities and requirements regarding education and skills. However, real-time LMI based on online job advertising data also has significant shortcomings: Apart from the challenges of duplicates and inconsistent levels of detail, it tends to be incomplete. Not all jobs are posted online, in particular, this type of data rarely captures the informal sector and is also often biased toward certain industries or occupations. In addition, the data may be distorted by ghost vacancies posted by non-hiring companies that want to cast a broad net for talent. Accordingly, real-time LMI is a complement to, rather than a substitute for traditional LMI.

Empowerment through interoperability

In addition to supporting policy makers and researchers, a strong LMIS should also provide additional services such as job matching, career and skills guidance and government support services through a government-managed online platform with interconnecting subsystems tailored to the different users. In this way, the LMIS strengthens the functioning of the labor market by helping all stakeholders in the labor market including workers, students, firms, and practitioners to make informed choices on a variety of topics such as job search and hiring strategies, curriculum design, career planning and training investments, and more.

International examples of modern LMIS

Worldwide, several countries offer examples of advanced LMIS incorporating LMI from traditional and big data sources and where the information feeds both into and from multiple interconnecting public interfaces to provide comprehensive, verified LMI for research and policymaking as well as job-matching, career guidance and skills development services. These sophisticated services include state-of-the-art tools and technologies such as AI/ML and big data analysis.

For instance, in Korea, information in the LMIS is used by the Korea Employment Information Service (KEIS) to monitor and evaluate public policies and generate analyses and forecasting for stakeholders such as job seekers, employers, researchers, and policy makers. Data is drawn from national statistics, surveys related to employment and skills, and databases from various interconnected KEIS networks, including HRD-net, a job-training platform, and Work-net. Originally established in 1998 as a publicly managed job-search portal by Korea’s Ministry of Employment and Labor, Work-net now provides comprehensive employment information and support services, including job matching and information on occupational outlooks, working conditions, and skills demand, as well as feeding user-generated data back to KEIS. With the progress of technology, it has added mobile services (2010), big data services (2018), chatbot services (2019) and AI-based job matching services (2020). [1]

The Norwegian LMIS also comprises interconnected subsystems that combine services for labor market supply and demand with data for decision makers and policy makers. The Norwegian Labor and Welfare Administration’s (NAV) online platform for job search and matching services, Arbeidsplassen.nav.no, has been using AI technology since 2019. It contains job advertisements both posted directly on the platform by employers and imported from external, privately managed job portals, as well as a CV database of job seekers, providing a comprehensive overview of the labor market. The system also has access to extensive information on the Norwegian education landscape to enhance the accuracy of job matching and career planning services. This modern digital platform provides automated and highly user-friendly services, and continuously self improves thanks to sophisticated machine learning algorithms in the backend. During the first wave of the pandemic, the system proved scalable by a factor of 8–10 within just a few days to deal with the surge in registrations caused by the dramatic disruptions in the labor market.

The technology behind the semantic search and matching engine and the underlying ontology of Arbeidsplassen.nav.no is provided by JANZZ.technology. JANZZ has been collaborating with several public employment services across the globe to assist their LMIS development. Our services range from state-of-the-art AI-based solutions to gather real-world labor market data and transform it into smart labor market intelligence – including job and resume parsing and automated classification and contextualization of job and skills data – over intuitive and powerful analysis and dashboarding tools that generate actionable insights including skill or workforce gap analyses, training and career guidance or semantic job matching, to designing entire system architectures from scratch. Visit our website and discover the advanced solutions we have created for public employment services or watch the explainer video for our integrated labor market solution JANZZilms!.

 

[1] https://openknowledge.worldbank.org/handle/10986/35378

Is Vietnam the next Singapore?

JANZZ.technology Viet Nam

Vietnam hopes to achieve high-income status by 2045. The country’s vibrancy is evident by investments in innovation and technology adoption that spur an innovation-driven private sector to build resilient businesses. Vietnam had a GDP per capita of $500 (today’s dollars) in 1985 which was one of the lowest in the world, and by 2021 it had already created a couple billionaire entrepreneurs.[1]

Vietnam’s performance is impressive as it was one of the poorest countries globally that achieved lower middle-income status in under a generation and became a dynamic East Asian economy. Its’ success can be credited to connecting to global value chains and offering favorable conditions to investors—much as it continues to do today according to the Prime Minister of Vietnam.[2] GDP per capita rose three-fold to about $2,800 and poverty drastically declined to less than 2 percent between 2002-2020. The Economist points out “it has been one of the five fastest-growing countries in the world over the past 30 years” ahead of Malaysia, Thailand, the Philippines.[3]

Achieving high-income status is an ambitious goal for a frontier market that already knows much about steady growth and global supply chains. Yet it will require 7% growth per year to achieve. Vietnam knows how to sell its goods abroad; trade exceeds 200% of GDP. Additionally, foreign direct investment (FDI) has been much higher than in China or South Korea for the past thirty years. Global companies were attracted by Vietnam’s cheap wages and stable exchange rate fueling a boom economy. But this export trade is mostly driven by foreign companies and not domestic ones.

With COVID-19, Vietnam had early success limiting the virus and GDP growth remained positive, albeit the lowest in three decades at 2.9%.[4] Yet the Delta variant upended the Vietnamese economy with factories shutting down disrupting supply chains for global companies like Nike, LG Electronics, and Samsung. In the end, the country’s growth outlook performed lower than the world average of 6% between 2% and 2.5%. Nevertheless, it was deep linkages to global manufacturing that sustained Vietnam’s economy in the pandemic.

How does Vietnam achieve high-income status? Answer: Better jobs.

As global uncertainty looms, Vietnam is thinking ahead about its’ future jobs landscape. The country knows it’s overly dependent on FDI and domestic firms underperform. Meanwhile, it’s difficult to remain competitive with increasing wages and ever-changing value chains. So, what can Vietnam do?

For a start, there are limits to what foreign firms can do to drive Vietnam’s development.

Vietnam’s economic success is attributed to its’ 50+ million jobs in recent decades. A big push in services and manufacturing reduced poverty in a country where 3 in 4 Vietnamese work in either family farming, household enterprises (unincorporated, non-farm businesses), or uncontracted labor. Economic growth happened because labor productivity increased alongside wages.[5] Yet Vietnam needs to further develop its services sector improving the quality of jobs if it is to achieve high-income status.

There is a strong government push to foster a Vietnamese chaebol system comparable to South Korea’s. Chaebols are the large conglomerates that helped develop South Korea’s new industries, markets, and export production making it one of the Four Asian Tigers. Vietnam already has the Vingroup with operations across education, health, real estate, and tourism. Developing a system of “national champions” may be the way to offset the widening gap between foreign owned firms and domestic ones, which have more barriers to access capital.

Vietnamese firms can also benefit from the growing Asian consumer class. There is a large consumer market waiting to be untapped in the region, especially if Vietnam expands its knowledge intensive services and modernizes its agro-business sector. Perhaps by creating jobs away from more traditional sectors, it can play a role in developing small and medium enterprises that better integrate into the larger economy and enhances supply chain connectivity.

Of course, this is not to say that Vietnam should forget traditional sectors completely. They represent most jobs in the country, about 30 million. Jobs in farming should diversify agricultural output into higher value-added crops and local value chains. And household enterprises must increase the quality of goods and services to remain competitive regionally and globally.

Human capital investments will be key in fostering an agile workforce ready to embrace tomorrow’s jobs. The Vietnamese labor force should build 21st century skills with adequate education and training. Future industries in Vietnam will require new skills sets, ways of working, and business models to export and expand. Automation may also displace jobs and enable others to become more efficient and productive.

It is evident that trade and consumption is already changing and impacting Vietnam. Much like Singapore, it can remain business friendly and competitive by focusing on public-private collaboration, innovation and digital transformation, and connecting qualified workers to the right jobs.

Here at JANZZ.technology we are ready to assist Vietnam towards its 2045 development goals.

 

[1] The Economist. November 27th, 2021 Edition. Vietnam has produced a new class of billionaire entrepreneurs. https://www.economist.com/business/2021/11/27/vietnam-has-produced-a-new-class-of-billionaire-entrepreneurs
[2] World Economic Forum. October 29th, 2021. Prime Minister of Viet Nam Speaks with Global CEOs on Strategic Priorities in Post-Pandemic Era. https://www.weforum.org/press/2021/10/prime-minister-of-viet-nam-speaks-with-global-ceos-on-strategic-priorities-in-post-pandemic-era/
[3] The Economist. September 4th, 2021 Edition. The economy that COVID-19 could not stop. https://www.economist.com/finance-and-economics/2021/08/30/the-economy-that-covid-19-could-not-stop
[4] International Monetary Fund. March 2021. IMF Country Focus: Vietnam: Successfully Navigating the Pandemic. Washington, DC. https://www.imf.org/en/News/Articles/2021/03/09/na031021-vietnam-successfully-navigating-the-pandemic
[5] The World Bank. Vietnam’s Future Jobs: Leveraging Mega-Trends for Greater Prosperity (Vol. 2): Overview (English). Washington, D.C.: World Bank Group. https://documents.worldbank.org/en/publication/documents-reports/documentdetail/670201533917679996/overview

When it comes to the use of AI in HR, it is past high time


There are numerous constitutional articles, laws, ordinances and regulations according to which companies must conduct their daily activities. And the number of these legal foundations is constantly increasing. A relatively new piece of legislation in Europe is the EU’s General Data Protection Regulation, or GDPR for short, which was adopted in 2016. The aim of this transnational regulation is to standardize the collection, processing, storage and deletion of personal data by private and public actors.  » Read more about: When it comes to the use of AI in HR, it is past high time  »

The Great Resignation or just a Great Misperception?

There have been many “great” waves in economy, especially in the US: the Great Depression (1929–1933), the Great Inflation (1965–1982), the Great Moderation (mid 1980s–2007), the Great Recession (late 2007–2009), and now we have a new one: the Great Resignation. But while the previous “great” events were undoubtedly real and had far-reaching impact on the economy and the labor market, this time round, there is reasonable doubt as to whether this wave of quits really is so great. Despite the huge buzz this catchy term has generated in the media – which is often based on shaky data promoted by data providers whose main interest is self-marketing.

The Great Resignation, a term allegedly coined by Anthony Klotz of Texas A&M University, was originally a prediction. Back in May 2021, Klotz anticipated a rise in quits based on pent-up resignations that hadn’t happened since early 2020 due to the significant job uncertainty brought on by the pandemic. He claimed that these numbers would be multiplied by “pandemic-related epiphanies” about family time, remote work, life and death and so on. Now that the quits numbers really have gone up, his prediction seems to have been turned into a prophecy, with the widespread consensus that it is happening for all the reasons Klotz stated. But there are good reasons to take a more critical view on this thesis, the most pertinent being the glaring lack of reliable evidence to support it.

The lesser resignation

In October 2020, 4.2m workers in the US quit their jobs, which is almost 3% of total employed workers. Viewed as an isolated number, this quits rate has indeed risen significantly over the last 12 months. However, the job openings rate has also increased dramatically to 7% of total employment plus new openings. This 60% increase from pre-pandemic levels far exceeds the numbers of the past two decades. Accordingly, the hires rate increased to a level last seen 20 years ago. If we put the quits into this context, say, by considering the ratios of quits to job openings, the situation looks far less dramatic.

 

The great misperception

 

It certainly makes sense to view these values together, since quits are strongly correlated with job openings: For the available monthly data from 2000 through 2021, the correlation coefficient is 0.82 – higher than for quits with unemployment (-0.77) or hires (0.75). And in relation to job openings, the current quit rate is by no means an outlier. Both the current rates of job openings and of quits are higher than usual, straying away from the typical cluster as shown in the figure below. And yet, their relationship still follows the same pattern as before the pandemic, when these changes could not have been attributed to, say, a desire for life changes newly discovered in lockdown. It thus seems more likely that quits have risen primarily for the more mundane reason that an unusually high number of job opportunities are opening up to workers, i.e., quits are simply at the level we would expect them to be, given the number of job openings. As Josh Bersin put it: Right now, there are just too many jobs and not enough people. And so we are likely seeing a great job hop as opposed to a great resignation. But it is not necessarily about doing something new.

 

The great misperception

 

The lesser wages

Looking at the figures in more detail, we see that of the four (sub)industries most affected by quits, three belong to the by far lowest paying industries: Retail Trade and Leisure and Hospitality. The latter was also the group that experienced exorbitantly high layoff rates at the beginning of the pandemic, and many front-line workers in retail were forced to work under precarious conditions with little to no monetary reward.

 

The great misperception

The great misperception

 

These low-wage workers are hardly quitting to indulge “pandemic-related epiphanies” or “craft careers”. Instead, they have been struggling to make a living since long before the pandemic. And now, with their industries among those with the highest net new job creation in 2021, these workers have a window of opportunity. If you have a job with no security, no appreciation and a salary that is barely enough to survive on, why not quit it for a job that pays $1 an hour more? Especially in a wage band that has experienced close to no growth in the past two decades.

 

The great misperception

 

The lesser reasons

Despite these numbers, much of the news coverage and reports have focused on burnout and remote work as the main drivers of the Great Resignation, claiming that white-collar professionals are shifting their career paths and leaving their jobs for companies that offer work arrangements that better suit their newly found values and preferences. If at all based on data, the most often cited source is LinkedIn. But LinkedIn data is extremely biased towards white-collar professionals. If we take a look at the much more representative data from the Job Openings and Labor Turnover Surveys, we see that in the largest white-collar industry, professional and business services, quits have risen at less than half the rate for the leisure and hospitality industry – despite an above-average job openings rate. In finance, real estate or information, which includes software, internet and publishing companies, quits are not rising much at all. In other words, employees more likely to be working remotely and thus with an increased risk of (self-reported) burnout are in fact less likely to quit.

Another supposed factor is health concerns related to COVID-19. However, many of the workers with these concerns had already left the workforce in 2020. And although labor force participation rates are still below pre-pandemic levels, they have been increasing steadily since April 2020 for most groups – except for older Americans (unsurprisingly, as the pandemic poses a much higher health risk to older people). But instead of calling them ‘resignations’, these quits could also simply be called retirements.

 

The great misperception

The great misperception

 

The lesser regions

It is also worth noting that, while there is a lot of talk about the Great Resignation being a global phenomenon, it is in fact just talk. At first glance, the situation looks similar in the UK: a high quits rate coupled with a high level of vacancies and fast wage growth. And again, taking a closer look at the figures, resignations as a proportion of job moves are simply back to pre-pandemic levels, with low-skill/low-wage workers driving the surge in quits – presumably for similar reasons as in the US: painfully low wages, bad working conditions and disloyal employers. Rather than quitting for something new (as in rethinking careers), a higher proportion of workers are moving to new jobs in the same industry. Moreover, while there was a surge in wage growth in Q2 of 2021, this figure rapidly decreased back to the level of Q1 over the second half of 2021.

 

The great misperception

The great misperception

The great misperception

The great misperception

The great misperception

 

And the UK is the only other country with any kind of similarity. In Canada, the number of quits is still far below pre-pandemic levels – especially the number of people who left their job because they are dissatisfied. In Japan, the percentage of people who quit their job for this reason has remained comparable to the level in 2015.

In the EU, there is no comparable quits measure. But there is also no other measure that suggests a big rise in resignations. Wage growth across the EU is the lowest ever recorded and the labor force participation rate has returned to pre-pandemic levels, with rates in individual countries like Spain and France even exceeding those levels. According to the latest labor force survey conducted in Spain, the number of voluntary resignations in 2021 is still lower than before the pandemic.

 

The great misperception

The great misperception

 

In Latin America and the Caribbean, the national economies are still reeling from the effects of the pandemic in terms of GDPs bottoming out and unemployment rates going through the roof. In most of these countries, the labor market is still far from recovery. The quality of available jobs has decreased, and the number of weekly hours of paid work is still significantly less than before the pandemic. According to the latest ILO report on Covid and the world of work, this is in fact true generally for low-income and lower-middle-income countries.

 

The great misperception

 

Unsurprisingly, the number of quits has not risen in these countries either. Overall, there is simply no evidence of a big rise in resignations anywhere in the world apart from the US and the UK. Let alone for the reasons promoted by the media.

The great hype

Basically, if you take the time to look at the hard data, there is neither reason to panic, nor to celebrate this not-so-great resignation. For a truly great resignation in the sense conveyed by the media and self-marketing data providers, the majority of workers would need to be in the comfortable position of actually having a choice regarding their lifestyle and consumption needs, and the job that fits those needs. Currently, we are still very far from any such scenario.

What has happened in the US and UK is that low-wage workers who previously had to take whatever work they could get, now have some agency. After a long period of employers exploiting the fact that they had workers on tap and thus hiring on whatever terms they deemed fit, the tables have turned. Jobs paying minimum wage may have to change their pay and benefits to become more attractive. The question is how long it will last. And if it does last, will pay rises and perks be enough? Or will it require more profound changes and rethinking our attitudes to low-wage jobs as a society? More on this in our upcoming whitepaper.

‘So clever I don’t understand a word of what I am saying’ – AI’s potential for handling text-based data is far from unlimited

janzztechnology

The often-expressed fear that AI robots are on the verge of infiltrating and taking control of every aspect of our lives is admittedly understandable, given the capabilities of AI already proclaimed today: writing guest articles for newspapers, answering basic customer service queries, diagnosing medical conditions, solving long-standing scientific problems in biology and much more – if you believe the sources [1],[2]. But are these purported successes really evidence of unlimited potential? Will AI systems really be able to solve any task given enough time and data? You may not want to hear this, but the answer is a resounding NOPE. At least not if researchers and developers stick to the knowledge-lean approaches they are currently so fixated on.

To start with, artificial intelligence is not even remotely comparable to human intelligence. Any AI system is a sophisticated ‘bag of tricks’ designed to lull humans into believing it has some kind of understanding of the task at hand. So, to develop AI-based technologies that are intelligent in any sense of the word there is no way around feeding these technologies with extensive human knowledge – which involves a substantial amount of human labor and will do for the foreseeable future. As a consequence – and again, you may not want to hear this – relying on solutions for HR and labor market management based solely on deep learning (DL) or other statistical/machine-learning (ML) approaches is a bad investment. And it is not simply a waste of money, it is also an ethical issue: Especially in the field of HR, the reliability of data and evaluations is crucial as it can deeply affect human lives. For instance, all too often perfectly suitable candidates are screened out by AI-based systems like ATS just because their resume does not contain the exact keywords specified in the filter or is associated with false contexts. Which a human recruiter would have realized, had they seen the resume themselves. This is just one of many examples of how real people can be affected by underperforming AI technology.

Artificial – as in fake

While researchers define AI systems as ones that perceive their environment and take actions that maximize their chance of achieving their goals, the popular perception of AI is that it aims to approach human cognition. Intelligence is typically defined as the ability to learn, understand, and form judgments or opinions based on reason or to deal with new or trying situations. However, this requires a key cognitive ability: storing and using commonsense knowledge, which we humans develop through a combination of learning and experience – and that so far, AI systems simply do not have and won’t achieve in the foreseeable future. These limitations are most obvious in natural language processing (NLP) and natural language understanding (NLU) techniques based on ML because commonsense knowledge is absolutely essential when it comes to understanding natural language. As an example, consider the following statement:

Charlie drove the bus into a tree.

Nowhere in this sentence does it explicitly state that Charlie is a human being, was in the bus, or that this is uncommon behavior. And yet our commonsense knowledge allows us to draw these and many other conclusions from this simple sentence without much effort. This ability, coined ‘linguistic competence’ by linguist Noam Chomsky, distinguishes computer systems trained in NLP and NLU fundamentally from human cognition. While we humans acquire this linguistic competence at an early age and can use it to discern the meaning of arbitrary linguistic expressions, knowledge-lean AI models will never be able to do so to the same extent because they work on a purely quantitative basis: their ‘intelligence’ is based on statistical approximations and (occasionally pointless) memorization of text-based data. ML systems can, at times, sidestep the problem of understanding and give the impression that they are behaving intelligently – provided they are fed enough data and the task is sufficiently narrowed down. But they will never actually understand the meaning of words; they simply lack the connection between form (language) and content (relation to the real world) [1].

This is precisely why even the most advanced AI models still struggle with these types of statements: because they contain so much implicit, often important information and causalities. For example, GPT-3, a state-of-the-art AI-based language model (which wrote the newspaper article cited at the beginning), was unable to correctly answer the simple question of whether a toaster or a pencil was heavier [1]. This is somewhat reminiscent of a quote from Oscar Wilde’s The Remarkable Rocket: “I am so clever that sometimes I don’t understand a single word of what I am saying”…

A major reason for this problem is that commonsense knowledge comprises an unconceivable number of facts about how the world works. We humans have internalized these facts through lived experience and can use them in expressing and understanding language without ever having to encode this staggering amount of knowledge into a written form. And precisely because this tacit knowledge is not captured systematically, AI systems have no access to it – or at least knowledge-lean AI systems don’t, i.e., systems based purely on statistical/ML approaches. So these systems are faced with unsurmountable challenges when tasked with understanding language. Because it is ‘unexpected’.

Another simple example: In a statistical analysis of words related to the English word pen, an ML system may spit out the words Chirac and Jospin, because these names are often mentioned together with the French politician Marie Le Pen, who of course has nothing to do with writing tools. It gets even more complicated when the same expression takes on different meanings depending on the context – think writing pen versus sheep pen. Systems based purely on ML often have great difficulty in discerning the nuances of such everyday language because they do not store the meanings of a word; connections are just based on cooccurrence. So, in the knowledge-lean world, there is still a long way ahead to reliable NLU.

No AI without HI

Having been around since the 1950s, AI has cycled through phases of hype and disillusionment many times. And right now, at least in the subfield of NLU, we are cycling back into the ‘trough of disillusionment’, as Gartner has so aptly coined it. Nevertheless, many are still clinging on to the great promises, blithely publishing, touting and investing in knowledge-lean AI technologies. But relying completely on ML-based algorithms for any application that requires language understanding is nothing but an expensive mistake. As we already explained, it is a huge leap from automated processing of textual data (NLP) to meaningful human-like understanding (NLU) of this information by machines. Thus, many automation plans will remain an illusion. It is high time to switch to a strategy that can succeed in these challenging tasks by effectively creating artificial intelligence through human intelligence.

In our area of expertise here at JANZZ, where we (re)structure and match job-related data, we understand that many automated tasks in big data require a significant amount of human labor and intelligence. Our job and resume parsing tool JANZZparser! has relied on NLP and NLU since the beginning – but always combined with human input: Our data analysts and ontology curators carefully and continuously train and adapt the language-specific deep learning models. NLP tasks are trained using our in-house, hand-selected corpus of gold standard training data. Parsed information is standardized and contextualized using our hand-curated ontology JANZZon!, which is the most comprehensive multilingual knowledge representation for job-related data worldwide. This machine-readable knowledge base contains millions of concepts such as occupations, skills, specializations, educations and experiences that are manually linked by our domain-specialized experts according to their relations with each other. JANZZon! integrates both data-driven knowledge from real job postings and resumes as well as expert information from international taxonomies such as ESCO or O*Net. This is the only way to ensure that our technologies can develop the kind of language understanding that actually deserves the name artificial intelligence. Generic phrases such as flexibility are given the relevant context, be it in terms of time management, thinking, or other aspects. As a result, false matches such as Research and Ontology Management Specialist with occupations like those in the figure below, due to overlap in wording but not in content, are excluded from matching results in our knowledge-based systems. The unique combination of technology and human intelligence in machine-readable form can achieve highly accurate, reliable and cross-linguistic/cross-cultural results when processing job-related data. Errors like the one in the pen example simply do not occur because each word is conceptually linked to the correct and relevant meanings and points of association.

 

Throwing good money after bad

The fact that we are on the right track with our hybrid, knowledge-based method of combining human intelligence with state-of-the-art ML/DL methods is not only confirmed by our own experiences and successful cooperation with businesses and public employment services (PES) across the globe, but also widely recognized by – non-commercial – NLU researchers. The outlined problems around the missing cognitive component in knowledge-lean AI systems will not be resolved in the next 50 years. As soon as human language is involved, there will always be countless cases where a 3-year-old child can make the correct semantic connection while a machine-learned tool either fails or does so only with absurdly high effort. Although knowledge-based systems like ours provide reliable and explainable analysis of language, they fell from grace because researchers and developers perceived the manual effort of knowledge engineering as a bottleneck. And the search for other ways to deal with language processing led to the knowledge-lean paradigm. Nowadays, supported by the immense speed and storage capacity of computers, most have come to rely on applying generic machine learning algorithms to ever-growing datasets for very narrow tasks. Since this paradigm shift, many developers and consumers have invested a lot of time and money in these systems. Being so heavily invested financially, they are simply not prepared to admit that this approach cannot produce the results they are looking for, despite the growing evidence against them.

However, the hybrid, knowledge-based approach of combining ML-based features with human-generated semantic representations can significantly improve the performance of systems that depend on language understanding. In our case, by adopting this approach, our technology can avoid the pitfalls of knowledge-lean systems based on uncontrolled AI processes, simple keyword matching and meaningless extractions of intrinsically context-poor and quickly outdated taxonomies. Instead, our matching and analytics solutions can draw on the smart data generated by our ontology. This context-based, constantly updated knowledge representation can be used in a variety of ways for intelligent classification, search, matching and parsing operations, as well as countless other processes in the area of job-related data. Especially in HR analytics, our solutions achieve above-average results that far exceed the performance of comparable offerings on the market. Thanks to these insights, employers are able to make well-informed decisions in talent management and strategic workforce planning based on smart, reliable data.

Do the right thing and do it right

Finally, there are the ethical concerns of applying AI to textual data. There are numerous examples that illustrate what happens when the use of machine learning systems goes awry. In 2016, for example, a software manufacturing giant’s chatbot caused a public controversy because, after an unsolicited, brief training session by Internet trolls, it almost immediately started spouting sexist and racist insults instead of demonstrating the company’s NLP technology in an entertaining and interactive way as planned. The challenge of developing AI that shares and reliably acts in accordance with humanity’s moral values is an extremely complex (and possibly unsolvable) task. However, given the trend toward entrusting machine learning systems with real-world responsibilities, this is an urgent and serious matter. In industries such as law enforcement, credit or HR, the inadequate use of AI and ML is all the more delicate. Talent and labor market management, for instance, directly affects the lives of real people. Therefore, every decision must be justifiable in detail; faulty, biased or any kind of black-box automation with a direct impact on essential decisions in these matters must be weeded out. This stance is also taken by the European Commission in their whitepaper on AI and the associated future regulations, especially in the HR sector. As a matter of fact, almost all of the highly praised AI systems for recruiting and talent management on the market, mainly originating from the USA, would be banned under these planned regulations. JANZZ.technology’s approach is currently the only one that will be compatible with these planned regulatory adjustments. And this has a great deal to do with our knowledge representation and how it allows us to produce not just AI technology that comes very close to understanding language, but in fact explainable AI. So ultimately, the way forward is to appreciate that – in the words of NLU researcher McShanethere is no bottleneck, there is simply work that needs to be done.

At JANZZ.technology, we have done this work for you, with experts from diverse backgrounds in terms of language, experience, education and culture. Their pooled knowledge is incorporated into our ontology JANZZon! and made readable and processable for both machines and humans. Together, our experts have created and continuously curate the best possible and most comprehensive representation of the ever-growing heterogeneity of job-related knowledge in the field of human resources and labor market administration. Enabling multilingual, modular and bias-free solutions for all HR processes – and bringing you a step closer to truly intelligent HR and labor market management solutions. If you would like to learn more about our expertise and our products, or benefit from advice tailored to your organization’s individual situation, please contact us at info@janzz.technology or via contact form, visit our product page for PES and listen in on our new podcast series.

 

[1] Toews, Rob. 2021. What Artificial Intelligence Still Can’t Do. URL: https://www.forbes.com/sites/robtoews/2021/06/01/what-artificial-intelligence-still-cant-do/amp/
[2] GPT-3 (Guardian). 2020. A robot wrote this entire article. Are you scared yet, human? URL: https://www.theguardian.com/commentisfree/2020/sep/08/robot-wrote-this-article-gpt-3

Teleworking, teletravail, teletrabajo… Who is working remotely?

In Washington D.C., metro ridership is only 30 percent of the 2019 ridership. The hustle and bustle of the city has not returned as employers are uncertain on when and how to reopen offices due to the Delta variant and at present Omicron. A Capital Covid survey conducted by the Greater Washington Partnership revealed that less than half of employees were expected to be back in the office on an average workday this fall. The slow return to work in the Washington region highlights new trends in work-from-home and hybrid arrangements becoming the business norm.

Across the world, employers and workers alike are coming to terms with more flexible working arrangements. In 2020, employers were not ready for their entire workforce to work remotely. Prior to the pandemic, about 17 percent of American’s worked remotely 5 days a week. Today, 45 percent of full-time employees in the United States were partly or fully remote in September 2021 per Gallup’s monthly employment trends update. About two-thirds of white-collar workers remain working from home (41 percent) and/or with a hybrid option (26 percent).[1]

In Europe, highly skilled professionals were more likely to be working from home (WFH) pre-pandemic than other workers. Approximately, 5 percent of EU nationals worked from home before Covid-19 while now that figure stands at 12.3 percent who do “home office” as it is called in Europe per Statista data. These figures vary depending on where in the European Union workers find themselves.

Home-based work is nearly non-existent in Eastern European countries such as Bulgaria and Romania with less than 3 percent working remotely. In comparison, one in four Finnish workers do home office (25 percent) followed by Luxembourg and Ireland with about 20 percent teleworking. In countries such as France, Germany, Spain, and Portugal between 10 to 15 percent partake in WFH.

Unsurprisingly, the prevalence of remote-based work also varied by industry and profession pre-pandemic. Knowledge workers or those in ICT-intensive sectors in the Netherlands and Sweden (about 60 percent) did some form of telework, while less than 30 percent did so in Italy, Austria, and Germany.[2]

Yet, this is not an option for workers in professions that require face-to-face interactions such as healthcare, hospitality, retail, and education. The gap between those who are WFH and in-person appears to create societal cleavages making society more unequal – as is currently seen in public debates in Switzerland.

On the employer-side, remote work brings new challenges to companies that rely on knowledge and creativity to spark new ideas and drive innovation. Workers miss out on face-to-face contact or “water cooler” chats that foster collaboration and help employees share information in ways that are limited or siloed by Slack channels, chat rooms, and email. Many executives also believe WFH cannot replace personal interactions that foster company culture. Productivity gains may also suffer in the long-term as collaboration declines amongst workers.

But so far, the Economist Intelligence Unit finds divergent views on workplace productivity. Nearly 39 percent of executives believe WFH has increased productivity while 33 percent find it has declined.

Globally, the study finds that company size and nature of the business impact productivity more than geographic location. Larger firms have more resources and digital tools to allow business continuity remotely – so, perhaps smaller firms without ICT uptake witnessed a productivity decline during the pandemic.

The uptake of remote working was accelerated by the pandemic, yet it remains more pronounced in the United States than Europe – even as EU countries encourage home-based work a few days per week due to the nascent Omicron variant. Overall, American workers report being happier with the more flexible WFH lifestyle and improved well-being, coupled with lessened commute times. Gallup’s State of the Global Workplace reveals that 91 percent of U.S. workers who work at least partially remote hope to continue splitting time between the office and work. Hybrid work is favored by 70 percent of workers partially on WFH and almost half of those fully on-site with jobs that can be feasibly performed from home. Only 6 percent of fully remote workers stated wanting to return to the office full time.[3]

Reevaluating work and the hybrid paradox

It has been nearly two years since the world heard about Covid-19. In that time, organizations and employees have been nimble to embrace all the surrounding complexities and disruptions to work life. The pandemic upended individual’s relationship with work and made many rethink not only how they work but also when and where.

While the United States witnessed the “Great Resignation”, worldwide about 40 percent of workers considered leaving their current job in 2021. Microsoft’s Work Trend Index points towards a new social contract between organizations and employees. Successful organizations are those likely to appease to individual’s different work styles. Globally a “hybrid paradox” appears to be gaining momentum with workers – people want to work from anywhere yet crave in-person connection.

At JANZZ.technology we strive to connect job seekers with jobs and businesses with talent powered by cognitive computing to find the best fit in labor market solutions.

 

[1] Saad and Wigert. October 2021. Remote work resisting and trending permanent. Gallup News. https://news.gallup.com/poll/355907/remote-work-persisting-trending-permanent.aspx
[2] European Commission.  2020. Telework in the EU before and after the COVID-19: Where we were, where we head to. Science for Policy Briefs. Brussels.  https://ec.europa.eu/jrc/sites/default/files/jrc120945_policy_brief_-_covid_and_telework_final.pdf
[3] State of the Global Workplace: 2021 Report. Gallup. https://www.gallup.com/workplace/349484/state-of-the-global-workplace.aspx?thank-you-report-form=1