 So I kind of want to start off with a big broad sort of maybe almost dumb sounding question But and starting with you Byron Is there something wrong with the labor market? Is there something fundamentally wrong with the labor market itself other than the fact that you know There are lots of people unemployed or is there something structurally off right now about it? Yeah I think there is something there. There are a few things wrong with it And I think you can tell not just by the people who are out of work But the wages are flat that many people in the workforce are stuck And that employers can't get all the skills that they need like if you take that you say how could all those three things be? True I think it does point to dysfunctions in the labor market And I think it's I think it's a three-part story The first part is familiar the other two I think are much less familiar the first part is You know the technology is changing work And that's sort of growing jobs at the high skilled end and some of the low end of the market But even the ones at the low end of the market require digital skills that not everybody has And so therefore you know people are out of work. That's the clinic classic skills gap story And there's a lot of truth to it, but it's a very incomplete story. The second thing is that our entire Labor market has not adjusted to some quite fundamental changes over the last 20 years in the way That employers hire and train and retain so in terms of hiring Employer US employers would hire massive numbers of people straight out of high school straight out of college We not only didn't have these specific technical skills and experiences. They didn't even honestly know how to work, right? They would be trained from you know job one and And as they trained them they also retained them so across the business cycle If you look at the first five or six recessions after World War two For every for the drop in demand Employers dealt with that only one-third by laying off workers and two-thirds. They kept them on their books They just took a lower, you know They had lower profit for a while, but they kept them because they had trained them And then and they hired back by the way as many of the ones they laid off as they could and that's how the cycle Work so it all worked together. So, you know roll the tape forward and in the US now You have employers Well in recessions the last two recessions a hundred percent of the decline in demand was taken out in terms of labor That changed very quickly in the scheme of things by the way that one-third two-thirds was always the same in the US And Germany in the UK in this last recession was a hundred percent taken out in labor in the US One-third still in the UK just as always and none in Germany They had gone the other way and to sort of work sharing of the like so it's not a necessary State of affairs based on technology or any other kind of objective measure It's a change in practices and then in terms of the hiring and training There is a desire now to hire Like looking for credentials looking for experience. I want someone with three to five years experience doing this exact job And with these degrees and by the way, I want them at this price So just to take an extreme example For administrative assistance 19% of administrative assistance in this country have a four-year bachelor's degree But 65% of new job postings new job listings for administrative assistance require Require a bachelor's degree in order to be considered, right? So these lead to a third a third dimension of the labor market, which is that you have a marked decline in mobility in voluntary job mobility The way most of wage growth has historically happened is people quit a job and get a better job Because they're up for it. They're ready. So think about the best administrative assistant Probably can't go get another job because they are credentialed out of that better job No matter how much they've been keeping up with their field and how good they are So you've got people stuck and you can see it in the statistics This is a much less quoted statistic than some others But there's a 20 been a 28% decline in voluntary job mobility in this country since the year 2000 We tell ourselves the old story of people are switching jobs very quickly And it's true that they are because they're being laid off But it's not true that they are because they're finding better jobs or that they're less loyal in fact, they're stuck and making transitions and on ramps and off ramps and anyone following a career path It's not the very narrow and traditional is at a really big disadvantage in this job market And so yeah, I think the labor market is broken in some pretty fundamental ways. Yeah, I ask a big question I'm gonna get a big answer So I kind of want to sum up some of that and then I want to ask basically same question to you But so again, you're saying there's sort of almost like you said three parts. I feel like it's almost four It's really again the classic argument that there's a skills gap There are people do not there aren't enough coders in the country There aren't enough people with kind of the mid-level technical skills for factory work at the same time fact those same factories are Laying people off like they never did before during recessions, which probably also are it's not willing to hold people Kind of cultivate talent through time And then there's this idea that they may want too much to begin with that They're asking for very specific skills in their job descriptions or whatnot And then finally this is all scared people from moving around the country from to job to job Is what it sounds almost like you're saying am I interpreting that right? Well, it's not that just that it's scared people and this fits with I think what Robert is doing It's also confused people so ask yourself how many people in the workforce There's 150 million adults in the workforce So while we need to improve high school and college like let's not forget the people already out there working How many of them could say if I work to do X in order to learn why I could get to job Z and like have a good picture of that in there before be able to unlock So in other words the skills gap is not it's just some natural thing It's it's the result of the fact that if you can't give people a path and a clear demand signal to say if I master this I can get that job. I can earn more. I can get to a better place Well, then the market's not going to work as markets to get driven off of demand signals And the demand signal in the labor market is very poor when you have such an erratic economy It kills the incentive to learn those skills in a way Yeah, Robert. I want to you know first off You know, we're talking about how technology can help deal with some of these sorts of problems and obviously, you know Glassdoor has done a ton to You know make the market a bit more transparent for one give people some sense of a career path And I don't know I assume most people have used it at some point to look up Basically what their their colleagues are making at work if if not you should it's a great website people can submit essentially their salaries their benefits and Do reviews of their offices to give you a sense of what it's like to work at these places and it's extremely valuable tool It's only getting bigger, but so I want to ask that same question is it from your vantage point Is there something wrong with the labor market and what are you guys doing to try to do your part to fix it? Yeah, let me start with what we're doing. I mean We are fundamentally addressing a lack of transparency in the labor market fundamentally the Hiring transaction that is finding a job and a company deciding to hire someone happens in a tremendous amount of opacity today There's very poor data on both sides of the equation I started Glassdoor in 2007 because literally I think where you go to work is probably one of the most important decisions You'll make in your life We joke the only thing more important is your spouse and kids and if there was a review website for that Most of us probably wouldn't have kids And it was crazy when we started Glassdoor that that Despite this being so important and figuring out like what is a fair salary for the work you do There is almost no good data to help you do that as a result of that poor cultural fits happen People get deep in the interview process before they realize that the job isn't going to pay what they need to pay Or it doesn't have some key benefit that they need it to have That's really wasteful for them It's really wasteful for companies and we believe it's a tremendous drag on the economy So we're about bringing transparency to that and helping people Use our service to find a job in a company. They love at a really macro scale. Here's what I think is happening Here's my here's my personal hypothesis The US I think is in the at the tail end of a of a large scale Transition in the economy from a labor economy to a talent economy Yeah, it used to be that you needed labor generally to produce stuff And by the way that labor was like it or not largely interchangeable This is why labor unions came to be because labor could not Really differentiate itself very much on a on a on an assembly line And so you needed the power of collectivism to protect labor As we've transitioned to a talent economy a couple things happen first off You can differentiate yourself and talent you can stand out And that has completely changed the way employers hire and employers think about their employer brand And they're all fighting over this a relatively modest group of people who can stand out in a talent economy Is the first thing? The you know the fundamental problem is that the labor market today Really was designed five decades ago in this in this labor environment It was not and it was not created in this talent environment And so we simply don't have the infrastructure our education infrastructure our hiring infrastructure to be able to support True development of talent the second big thing that's happened if you look at kind of the academia on this what I see is There's a consensus that it is routine They use this phrase routine jobs that is declining you some of you in the room Sure, I've heard this and it's true like when we when our economists the glass door look at this data What they see is a massive decline in jobs that it can be classified as routine Jobs that require the you know creativity or collaboration or our non routine They're the ones that are growing very very rapidly that is a skill set that we do not currently teach for in order We particularly foster in our very competitive workplaces So, you know what you're talking about we don't have this infrastructure set up to you know a create Perhaps create all the talent companies you know that front but also for employers to find it, you know What I'm wondering what are employers doing wrong are they also missing talent in some way? Are there are there ways are they doing things that they're there people out there with the skills? They needed that they're kind of maybe passing over them. And why is that happening? Yeah, I I personally think employers are missing talent and they're missing it on a rather huge scale and you know, there's There's there's talent in their skill. I think there's talent is very widely distributed, but you need to you know through a combination of Learning and sort of mentoring and experience. I mean both academic learning and then absolutely on the job learning and These sort of social networks and social capital all translate that into Mastery right within a given context and I think the opportunity to achieve that mastery is far from evenly distributed Yeah, so that's one thing, but even so you get people who you know, you talk about creativity and collaboration as these critical Sort of skills and I absolutely agree and those in particular Don't land on you just because you went to a four-year degree You know you have a four-year degree or whatever and so we've seen actually we're working Opportunity at work has started working first in information technology jobs where it is relatively easy to demonstrate Whether you can do something or not and when you create alternative hiring on ramps fine Turns out that there might be in a in a city hundreds or even thousands of people who actually can do that job today That would absolutely get screened out would never get an interview right if you want to know if someone can code You can look at their resume you can look at their transcript or you can look at their code But actually most employers 90% plus will screen you out right automatically before they actually see what you can do So we've seen that there's a lot of people, but there's something more Damaging to the overall economy, which is that if there's not a way for people to demonstrate their competence that they can do This job then there's no business model for taking people with the aptitude with the talent Right but dot not yet the skill and then actually helping them develop that skill So in other words, you don't get the pipeline effect, but even today employers are overlooking I would say just in the IT sector certainly tens of thousands of people who could do the jobs that they can't fill Do you think that's true? I mean since you're in that sector. I do to a degree. Here's what I definitely think is true We have very very poor data on the state of the labor market And that is a real problem for example one of my favorites is we know eight and a half million people are unemployed We have no idea what they actually did Yeah, because states have not released the data and it's not in any kind of a database And like if we actually knew what their occupations were like it would be great to map their skills And that's the holy grail and be able to figure out like what's training they need to get another skill Let's start with just what was their occupation? What did they do because then you can begin to direct training dollars and you can start to say Well, there's a whole bunch of people two counties over that had this skill Or had this occupation this factory shut down We actually and we need a bunch of drivers, you know two counties over or whatever else because a new warehouse open And there's a bunch of shippers and we can begin to connect those dots We don't currently have The the information systems that that bls uses and everyone else uses we're really designed 40 or 50 years ago The architecture that they use for Even what jobs to collect data on are antiquated. They're survey based. They're not real time And this isn't a world where real-time data is the norm Yeah, and so I think we're reaching crisis point on this because I don't think we can make good decisions Without good data when we have good data And we can figure out where and we can empower employers then to see that they're making these mistakes Because I think Byron's right about that. I just think that it's not clear at all Because our visibility is so poor. Well, so how What could government be doing differently? I mean, I imagine there have got to be ways they could Kind of partner with private sector Maybe, you know make more of the numbers they do have available But what more could they be collecting? What what is it? You know, I guess what concrete steps are there that you see that they could take the government could take I think um, there's Most of it falls on this concept of open data. I don't know that a whole lot more data needs to be collected There's a tremendous amount of data already collected via tax tax Forms via unemployment filings Most of it is locked in bureaucracies and no one wants to know like private information like literally like what is You know this person's tax what's where we just want to know like How many of this occupation are there in this county and how many in the state and like You know and we and the bureaucracy is sort of doesn't understand I think how this could be used in an aggregate anonymized way And so that I think is Getting the word out and trying to change that is could be powerful. Do you think I mean, is there Let's say that that data Was released. I mean, would there be a role for private companies to yeah, then that's the idea There's tons of companies. I mean, we would use the data, but we're not alone There's dozens of companies like glass door who have really good view into the jobs landscape Like we have all jobs available in the united states and it's up to the hour accurate for the most part We we crawl the web. We connect every applicant tracking system We connect with every major job board and so, you know And we're about to launch a map where you can enter a job and see on a map literally county by county where jobs are That's not the problem. The problem is where are the unemployed? What are they capable of doing and how could training as byron says connect them to those jobs? There's a few specific proposals out there. So the obama administration has a proposal Very inexpensive one five million dollars to modernize Onet that is the the department of labor sort of taxonomy for jobs and skills Which most of the job boards and the like are building their databases on top of that Taxonomy and it's very it's very we're not actually it's too poor Yeah, well you we've we've had to throw out on it and rebuild our own real-time one That that sort of extracts a taxonomy in real time absolutely and and and there's a couple of your other Other competitors that have done that but for the most part that that's the taxonomy that's being used and actually It's it's quite difficult. It's one of the reasons that h1b Sort of works so poorly because it's it's sort of they're in it in particular very broad category So it's very hard to distinguish whether someone's doing a job that many americans can do or whether it really is a scarce Skill because they're just too blunt too broad these categories. So there's that but there was there in the the workforce Innovation an opportunity act that was just passed There was there was an effort to get that sort of data sharing But that was rejected on the basis of fears of how the data might be misused. So that is definitely a concern out there But I would just like to point out though that While I absolutely agree that the government can do more From an open data standpoint and that would be very valuable and nonprofits for profits, etc Could use that data We do have to keep in mind that if you look at the money in the public workforce system Which is about let's say generously 20 billion dollars a year In the higher education system. It's sort of 10 times that much. Um, the Employers spend 25 times as much on training. Um, and they spend about 500 times as much on payroll So actually if you want to shape the market behavior It has very much to do with how employers hire who they screen out Who they screen in and I think there are there are ways and we are we are seeing ways that without completely upending the system You can create hiring on ramps that allow employers to hire based on competency based on readiness rather than based on Pedigree in history. Can you just describe an example of that maybe but what you know In the I I think I understand what you're saying the abstract but just a concrete example again of right So an example of a hiring on ramp is um a launch code which operates in st Louis and miami and is starting other places They find uh, they find people who actually can code we they often bring them in and meet ups like coder girl meet up in st Louis um and basically give them uh some coaching Pure network and assess their skills and also work with employers to assess What skills the employers really need because they go from the job description, which are usually outlandish Descriptions that very few of their existing employees could actually meet and like really work them down to like what is it that you really need? And then they make the match so they did about um 100 of these 150 of these in the past year They're on a big growth ramp and they're placing people with very high success rates into it apprenticeships and then in full-time jobs Or you have companies that have tried to do that and instead they've hired the people themselves and then competed based on projects Like catalyst it services, which actually Gets is a very profitable company that hires entirely on this basis about 40 of their people don't have college degrees They're not discriminating against people with college degrees But entirely based on capability competing at the relatively high end of the market And because they have because they get people to play at the top of their game They only have a 13 turn over a year in a field that has a 50 turn over a year So their economics are far better than those of their competitors because they've arbitrage This like this actually the sort of poor decision behavior by most employers Robert, uh, you know to me that sounds like it's sort of in keeping with um Silicon's about the silicon valley sense of itself is you know a very meritocratic place Why in your opinion aren't there more people kind of or more companies trying to use those sorts of on ramps already? Or are they in some in their own way? Is there you know, is there a reason why we haven't seen more of this evolve? um You know the tech community is very very meritocratic from the standpoint of most places I've ever worked doesn't don't really care where you went to school Right. They they they want to know basically what you were able to do and whether you were able to write the code Or not to your point. Um the like My experience is we really are In a dearth of technical skills Right. The education system is fundamentally broken here 25 states don't recognize computer science as a core skill in teaching The uc system Classifies it as a as a you know a elective the same as you know You know dance or something like this, right? And they won't let that they won't let kids take it as part of the math track We are not producing enough people that understand how to Work in a world that in which I really do believe agree with andres and the software is eating the world We need more people that understand how to do this. That is my industry It's not a valley thing. We see that in boston austin You know new york every office, you know sees this this kind of thing um I think the the code stuff is Is really cool. It needs to scale is is i'm excited about it because you know Everyone's starving starving for engineering talent Yeah, and that's right like we aren't producing enough people You know It's probably a million and a half jobs in the next few years that need this You're going to have probably 400 000 computer science degrees I agree with that teaching coding as a liberal art in schools I think uh It's not to say that you don't need this in our educational system because you do and by the way to make that happen You're going to need to teach teachers how to code because we don't it's much easier to teach a teacher A good teacher how to code than a good coder how to teach and we're going to hit big bottlenecks in teaching if we don't do that But is absolutely true though that we are That's the young people and we need that don't forget the adults 150 million of them in the labor market today And with this catalyst they're based in baltimore Interestingly in coding high end of coding baltimore metro baltimore is 26 african-american Their coders are 25 african-american, which is unheard of In the valley or boston or anyplace else and the reason is because they're hiring purely based on talent actually and capability rather than based on all sorts of other Implicit biases that we take as signals of quality, but that actually are Signals of our familiarity. So on that I wish we could keep talking about this all day But I think we've actually run out of time. So I want to thank both of you for A wonderful talk and