BELZER, Dr Michael, Private capacity [by video link]

Committee met at 08:36

CHAIR ( Senator Sterle ): I declare open this public hearing of the Senate Rural and Regional Affairs and Transport References Committee. The committee is hearing evidence for its inquiry into the road transport industry. I welcome everyone here today. This is a public hearing and a Hansard transcript of the proceedings is being made.

Before the committee starts taking evidence I remind all witnesses that in giving evidence they are protected by parliamentary privilege. It is unlawful for anyone to threaten or disadvantage a witness on account of evidence given to a committee. Such action may be treated by the Senate as a contempt. It is also a contempt to give false or misleading evidence to a committee.

The committee prefers all evidence to be given in public, but, under the Senate's resolutions, witnesses have the right to request to be heard in private session. It is important that witnesses give the committee notice if they intend to ask to give evidence in camera. If a witness objects to answering a question, the witness should state the ground upon which the objection is taken. The committee will determine whether it will insist on an answer, having regard to the ground that is claimed. If the committee determines to insist on an answer, a witness may request the answer be given in camera. Such a request may of course also be made at any other time. If you agree to provide an answer to a question taken on notice during the hearing, please note that answers are due by close of business Monday 10 May 2021.

Finally, on behalf of the committee, I thank all those who have made submissions and sent representatives here today. It does give me great pleasure to welcome my very dear friend Dr Michael Belzer, Professor of Economics at Wayne State University, Detroit, appearing via videoconference. I have had the privilege of hearing Dr Belzer on a number of occasions through Washington and Geneva. That's why I'm rapt you could join us today, Dr Belzer. Dr Belzer, I now give you an opportunity to make an opening statement. Your opening statement might be quite lengthy compared to normal, but it's imperative that we get all of your evidence on the record. With the time left we will go to a few questions. We do have a full book today. I'm very keen to hear from you. Dr Belzer, over to you.

Dr Belzer : Senator Sterle, it's great to see you again. I'd rather be in Australia and seeing you in person, but this is what we have to do in a pinch, so here we are. My presentation is on economic incentives and market pressure and why markets actually require safe rates regulation. I'm going to walk through these—or maybe I'm going to run through them fairly fast so I can leave more time at the end; we've lost the first 10 minutes here. I've got about 35 minutes to present. I'll go through it quickly and then get the responses. Is that okay?

CHAIR: Yes.

Dr Belzer : Alright. I will go to the first active slide. I will dispense with saying, 'Slide, slide, slide,' but, if anybody is confused, let me know. I'm looking at the slides here. What we have here in the United States is truck fatalities on the rise again. They had dropped at the time of the financial crisis, in 2008, and the great recession that followed, but they've rebounded along with the economy, so we're really back in the same place we have been for some time, with the number of fatalities somewhere in the neighbourhood of 5,000 per year in truck crashes. Are we getting bang for our buck? We lose millions of dollars with these crashes and fatalities, but crashes and fatalities still go up. In my view, we have to recognise that large trucks operate in a market and are subject to the market's complex system, so we have to consider the market, and of course macro-economics is a part of the market where labour is such a big factor, as it is in trucking.

There is another slide here that has got links in it, and you will be able to get access to these links. I can make the electronic version of this available if you like. This is about the human cost of Amazon's free shipping. Amazon is a case in point because you can see the influence of industrial organisation as well as institutions and how the institutions influence safety and health. Amazon passes the cost of fast, free shipping, which is what they sell, to the public. A basic principle in economics is that any cost not captured in price is external to the market. This presents a problem in this particular case, and we'll see over the course of this presentation that this is a regular problem in a research agenda that I've been working on now for the last 25 years. Commerce call these external costs externalities. An externality is not an efficient use of resource. So we have an efficiency problem and an equity problem on our hands. Society subsidises external costs like that—safety and health—in the form of death, injury and property damage on the highway.

Low rates is a big part of this problem. Those result in low compensation. The low compensation results, of course, in vulnerable workers, and the vulnerable workers result in a vulnerable public. That's where all of this is a major public policy issue. The cargo owner or client, as you call them in Australia, which is Amazon in this case, controls part of time on demand work. Amazon subcontracts its delivery business. Subcontracting deflects the liability, it drives down rates and it takes advantage of an unregulated chain of responsibility, which is our problem in the United States.

What costs are externalised? Fatigue and fatigue related illnesses, injuries and crashes are externalised. Stress related illnesses, injuries and crashes are externalised. Costs associated with fatigue and stress are transferred to victims, including the highway users and workers. Economic costs include, in addition to these risks, a damaged market, unpaid taxes, underinsured vehicles, underinsured drivers, disregard for the public, driver shortages and driver turnover.

Competition drives carriers to the lowest price. Transport is a commodity. A unit of transport is a unit of transport, and that's how it gets priced. It took me a long time to realise that and to accept it. Lowest prices drive carriers to the lowest costs. The lowest costs drive freight rates down. The lowest costs then squeeze the drivers, because freight rates are low. What do we get? Unqualified and dangerous drivers, dangerous workplace pressures, dangerous and unhealthy hours of work and outsourcing that reinforces cost-cutting.

The stresses associated with work as a commercial motor vehicle, or CMV, driver put drivers at significant OHS baseline risk. These are also stressors associated with just-in-time logistics, which is our new buzzword, or has been for the last three decades or so. This requires a pressure for rapid and scheduled delivery and pressure to cut costs. You know the joke: 'I want that yesterday, and I want it for free.'

US truckers work really long hours. This may seem like they're not long enough. They're not as long as they are for Australian truckies. I'm not really sure what the data are in Australia. We did a survey back in 1997 showing that the median—that's half above and half below—non-union driver worked 65 hours per week. Fifty-five per cent of the drivers are not paid at all for their non-driving labour, and 70 per cent of them are not paid for waiting or other on-the-job work time.

Another survey was done in 2010 by NIOSH, the National Institute for Occupational Safety and Health, which showed that the median employee driver, almost all of whom by now are non-union, works 60 hours per week. Employee drivers, according to the NIOSH survey, average 63 hours per week, and 22 per cent of all driver time is unpaid, which is 10½ hours per week. That's quite a lot. Twenty-seven per cent of an employee driver's work week is unpaid labour. The kicker is that the NIOSH survey showed that 20 per cent of all drivers exceed 75 hours per week. You can see a picture from the raw data right here. The black vertical bars on the right are the hours of work in excess of the legal limit of 60 hours per week. The average is 61.5, but the median is 60. This is with 1,254 long-haul drivers, carefully sampled to be representative of the United States. On the bottom axis you have hours worked per week, and on the vertical axis you have the number of drivers total. It shows you just how many drivers are working over hours.

Why so many hours? The bottom line here is that, in the US, the answer rests in the conflicting definitions of work. I'm not positive, but I have some idea that this is more universal than we might think. This isn't just in the US. The Department of Labor, DOL, is responsible for enforcing a law called the Fair Labor Standards Act, FLSA. This is an 85-year-old law that defined what work was—or was supposed to be—for American workers. Under the FLSA, all time during which employees work for an employer, including their waiting time, is payable, unless the employer frees the worker from work for a very specific time and the employee knows in advance that work time's starts or end, unless the worker has practical freedom to leave the place of work to go about his or her personal activity—watch the stress on practical freedom—or unless the worker is not engaged in work for which he or she was hired, including being available for a call.

According to the Department of Labor, all time is payable if the worker is away from the employer's place of work. All time is payable. Wait time is part of the employee's regular work. In any other industry we would know what this meant, but in trucking it seems we have this idea that there are special rules. The Department of Transportation—the Federal Motor Carrier Safety Administration—has a different definition of work. It has taken me a while to realise this and the significance of it. Waiting time is non-work time, according to FMSCA, if: 'The driver is relieved of all duty and responsibility for the care and custody of the vehicle, its accessories, and any cargo or passengers it may be carrying,' and, 'During the stop, and for the duration of the stop, the driver must be at liberty to pursue activities of his/her own choosing.' Under that is a lot of exceptions. Companies may interrupt the driver's free time and his sleep time, so drivers have to be available. Off-duty time during a shift may have an indeterminate start or end time, nobody really knows. The reality is that off-duty drivers lack practical freedom and are tied to their job.

As an aside, Glenn, you're a former truckie and I'm a former truckie. I've been assigned and chained to that truck before; I know what it's like. When I was a teamster, I got paid. Now the problem is that they're not getting paid. FMCSA regulations now allow carriers to order drivers to log this non-driving time to DOL, what we'd define as worktime, as off-duty. This explains why drivers log unpaid worktime as off-duty. The results end up being inconsistent with the policy goal to limit hours of work, particularly in a safety related job like trucking. These definitions are inconsistent. They're inconsistent because time is money. That shouldn't be a surprise to us. We all know it; we all say that phrase. But I'm here to show you the data behind that phrase.

Economic theory predicts that workers will trade income for leisure as their earnings increase. Why do truck drivers work such long hours? Stanley Sedo and I wrote a paper that was published in the University of New South Wales's Economic and Labour Relations Review in 2018. It tests the target earnings hypothesis. Workers seek their target earnings to pay their bills. This is not all that complicated. Drivers work to reach those earnings targets and when they reach those targets, they'll start reducing their work time. They don't want to work anymore hours because they've made enough money to pay their bills. We expect to find that reduced work time reduces crashes, injury and illness. What data did we use? First, we used the UMTIP driver survey data from 1997-98. I'd love to have more recent data. We have some recent data, but it's not adequate for the job, so this is the best one that exists. According to the survey, employed drivers worked an average of 64.5 hours per week, and drivers earned an average of 45c per mile in today's dollars, which is five per cent less than what they earn today. There's been a five per cent increase in these dollars in the last 25 years, approximately.

Here's the part that you've got to put your seatbelt on for, because I'm going to give you some econometrics, and all you've got to do is follow along with me, and don't get hung up by it. This is a two-stage least squares regression. The goal here is to do a complicated thing, and I'm sorry to put it like this, but it's going to give a result which is really interesting and it's going to give you a lot to think about. We first estimate the wage rate based on the information that we have on the drivers—that is in table 1. The rates of I mean the mileage rate for the individual driver. So we're estimating the individual driver's mileage rate. The Xs are the characteristics of the driver and the job that determine the rate. The betas are the parameters that we're going to estimate. We're going to plot that. We're going to calculate that first. We're going to estimate the mileage rate for all drivers. It's a highly significant model. You'll see the variables that you'd think would be significant, like experience, tenure, union membership, size of the company, miles. Those are all very significant predictors of what the mileage rate is going to be. We're going to take the mileage rate that we're going to predict in this equation, and then we're going to the second stage. We're going to plug it in here and use this 'fitted wage', as we call it, to estimate the hours of work at each wage rate. It's going to produce a pretty picture when we're done, so bear with me. Again, the hours are the weekly hours of the individual driver. The W is the fitted wage of the individual driver from the wage estimation equation. Now we're using that calculated wage rate. The Zs are characteristics of the driver and job, just like we had in the last one. These are the extra variables that we can plug in because we have them.

That gives us this model—this outcome in table 2. This is the weekly hours of work equation. Bear with me, there's one more step. We plug the estimated wage rate from the first equation into this hours of work equation. That gives us a fitted rate, the square of the fitted rate—that's going to turn out to be really important as I'll show you in a minute—age and age square are classic control variables that you'd expect, and then there's the percent of non-driving time that the driver has: how much time does a driver have that's not driving; and a separate variable, the unpaid time, because remember more than half of all drivers are not paid anything; they're paid zero in the US for their non-driving labour. This allows us to predict the hours the drivers work based on their wage rates, and what we get is this curve. This is called the backward bending labour supply curve. It's an unusual thing to find because you don't normally in our world find people working such crazy long hours as truckies work. So we estimate that drivers would work 60 hours at just less than 60 cents per mile in 2017 US dollars.

These numbers that are on the graphic are the original numbers from 20 years ago. I'm updating that according to the inflation rate so you can realistically understand what that is. Higher-paid drivers will reduce their hours of work, leading to greater safety and health. If you start at the bottom where the bottom hook of that loop is, the fish hook, you can see that around 66 hours per week, which, again, is six hours more than the legal limit, drivers are getting paid 25 cents a mile in the pay rates of that time, which comes out to around 40 cents per mile in today's rates. If you pay them more money, they will work more hours until they reach that inflection point, and that inflection point that corresponds to 70 hours, which is the second to last vertical hash mark, is the point at which drivers will start to choose to work fewer hours. This is, frankly, what we want them to do—to get paid while working fewer hours. Really, what's wrong with that? What they do is they choose to turn down work at that point. You see it junctions at 60 hours per week, on the left-hand side of this graphic. That's the point at which they're getting paid 60 cents per mile in 2017 dollars. Essentially, that suggests, in our terms, in terms of wage rates, this is a starting point for what a safe rate might be. The point estimate would be they're working 60 hours per week, and working 60 hours per week is still a lot of work to any civilian who's not a truck driver, but it's a lot better than drivers working 70, 75, 90 hours per week.

The second research paper recently done is 'Safe rates and unpaid labor'. We're looking here at non-driving pay. Remember, I talked about this idea that they have unpaid non-driving labour time. My former PhD student Takahiko Kudo and I published a paper in The Economic and Labour Relations Review—everything good seems to go into Australian journals! The workers trade labour for leisure. This is the basic premise in economic theory. This is a theoretical model. It's really simple though. This is a utility function, and drivers utility for their money and where they're going to choose to work fewer hours in trade and take some time off rather than working so hard. That 'I*' is at least a theoretical point at which that takes place. The 'I' is that regime change: the point the regime changes is the level of income at which the marginal utility of more income decreases acutely as the income goes over, and that's what drives workers to turn down work and work fewer hours, which, again, is what we need them to want to do of their own accord if we want to make the highways safer. So 'I*' might be called the 'time is money point'. I came up with that idea as I was reviewing this for the presentation. I like the 'time is money point'.

So, what do we do here? For the statistical model here, we're trying to predict the hourly work hours, and we have two variables that turn out to be significant. One is other drivers pay for the non-driving labour, and, the second one is the mileage rate. The mileage rate is estimated what they actually make. So we have some regression results, and I'm just showing you the top line fixture here. What you see is the non-driving pay is highlighted in red, so it's easy to see. It's highly significant at the 1/1,000th level, and it makes a really big difference. If drivers are not paid in any model we use, if drivers are not paid for their non-driving labour time, it's going to predict that they work more hours. It's exactly what we think it would be.

So, what does this tell us? Paying drivers for all their work time, which includes all non-driving work time, predicts the drivers will work fewer hours. If drivers are paid for their non-driving work, the trucking companies, cargo owners or clients for which they haul this freight will avoid delaying the drivers, because time is money. They'll get them back on the road. They don't want to pay for that time. Right now that time is free. It doesn't matter how much of that time we waste, because it's a truckie, we don't really care. This makes them care, because they have to pay for it. This explains why paying for all driver time is essential.

I want to go to another study that was done together with Dr Kudo—this is all very recent stuff—and published in Safety Science. We again used NIOSH long haul driver survey data. We used this efficiency wage theory, which says that better wages attract better workers and more stable workers and it gives them incentive to be safe, which is a good thing. This theory predicts that higher paid drivers will be safer. It also uses the same theory of the labour-leisure trade-off. We use the economic theory that is standard to make these predictions. We're not doing anything fancy here. This is just straightforward labour economics. We want to predict the expected number of moving violations in the past 12 months. Here's our regression model, and the only things that are significant are—here is the kicker again—only mileage pay rate and employer provided health insurance predict moving violations. The more your pay, and whether you're actually getting health insurance from your employer, there will be fewer moving violations. That's our safety measure. What does it tell us? Higher paid drivers are safer. I feel a little relentless, but I'm just going to be relentless, because that's what I do. You already knew that. Higher compensation is associated with a lower probability of having moving violations. Employer paid health insurance is also associated with a lower probability of having violations. We want them to have fewer violations, just like we want them to work fewer hours, because we want them to be safer.

I'm going to switch back to the granddaddy of this project, to a big paper that we did. All of this is cited and linked in the last three pages of this presentation, so you can always go back to the links. The paper we did in 2002—I know it's been cited repeatedly in Australia—I'm referencing it, and specifically going forward to a paper by Rodriguez, Targa and myself in 2006 that filed a particular one of those studies. In that study of JB Hunt, a national motor carrier—the second largest truckload carrier in the country—we had 96,000 driver observation months for one year before and after a major wage raise. A major wage raise is 39.1 per cent. That turned out to be a lot of money. They advertised to hire new drivers that would be good, and they got them. So we implemented something that's even more complex called the Cox Regression or survival analysis to predict the probability of a driver crash month to month.

So, bear with me, that's the overall model. If you want to study the econometrics, it's all present in those links. Doing that allowed us to figure out what the probability was that drivers would have their crash based on their pay. We had that, and here's what happened: the crash risk declined between 50 and 75 per cent immediately after that pay raise. On the left side you see 'before pay raise'—an empirical crash risk different by different months between 2.5 and three, somewhere around there. 'After pay raise' starts at about 0.02 approximately and drops. It drops all the way down to one. So we have an empirical crash rate of 0.01, whereas we had an empirical crash risk in the neighbourhood of three, which is one-third the crash risk, once the pay raise was implemented. After mean, each 1,000 miles driven per month was associated with US$381 of greater earnings, which produced almost 25 per cent lower turnover. Why? The drivers were making their money and the company were enforcing the hours of service regulations so they could go home or they could at least stop working and take some rest. One per cent higher pay resulted in 1.33 per cent lower crash probability, and the biggest reduction in crash probability happened at the lower pay rates. We know that those low-paid people that I showed you in the backward bending labour supply curb were at high risk, shifted out of risk. That's what the company wanted, and they were willing to pay for it. It was a very impressive experiment on their part.

Our newer study was done by another former PhD student and me. We drove this to completion in the fall of 2019. This was part of his dissertation—and there's a citation of his dissertation; it's going to go out to review for a journal, hopefully by the end of the week. We are measuring the relationship between regulatory violations, which are defined by Federal Motor Carrier Safety Administration; pay or pay incentives, which is how we think about it in economic terms; and crashes. Now we have a real hard measure of crashes, because it comes directly from the Motor Carrier Management Information System, which records all crashes. All reportable crashes are there. So our starting point is to measure our crashes. Then we use average earnings data, which comes from the Bureau of Labor Statistics, and that's state by state. As a result we just use intrastate trucking companies. That's the point here.

The student came up with this ingenious idea. We were kind of stuck, because we didn't have data on pay for the drivers or at the carriers. So we had to improvise, and he said, 'What about the pay rate, on average, within the states and we'll just look at intrastate carriers?' There are 14,500-odd intrastate carriers, so we had a lot of carriers to look at. And we did, because we have good data. Crashes are a function of violations in MCMIS. We know that they're related to what are called the 'basics'. There are the regulatory categories—working time, unsafe driving, driver fitness, substance abuse and vehicle maintenance—and we add to those. That's what they measure already, and we've found that some of them are statistically significant but most of them are not. We plug in state-average hourly pay rates for drivers of heavy trucks as our proxy for an estimate of their hourly wages.

The thing I want to show you in our results is this line that says 'L-wage', down there. It's highlighted all the way across. That's a really big number. You can see that hours-of-service violations have an influence, but look at the size of the negative influence that wages have. Wages drive down crashes by a lot. HOS violations are significant predictors of crashes, but the average statewide wages for heavy-duty truck drivers have 15 times as great an influence on crashes as hours-of-service violations do. So if you really want to know where the crashes are going to be, find out how much these guys are paid.

What does 'working time' mean? Working time matters. An increase in one carrier-level hours-of-service violation adds 1.28 more crashes. Remember that work hours are inversely related to pay rates, so achieving fewer hours-of-service violations comes from working fewer illegal hours. It seems like an obvious thing, but I'm just pointing it out. Working time and pay are two halves of the same coin, because time is money. The effect of hourly wages is huge. A one per cent higher hourly wage at the mean corresponds to 3.16 per cent fewer crashes for a motor carrier. That's more than a three-to-one ratio. It's by far the biggest effect of the model and it far outweighs any other factors that the FMCSA has been collecting to predict crashes.

Finally, let's go to safety pay. There's another PhD student, Michael Faulkiner, working on the JB Hunt data and following up on the research that we did back in the day. Safe rates and the return on investment are connected. You know that companies want to know what the business case for safety is. What I'm here to say is that the business case for safety and the public policy case for safety are basically the same. Companies will make more money with higher paid experienced drivers, and we can show this.

Here are higher paid experienced drivers in the comparison of safety. That spike, up there at the top, shows inexperienced drivers at the lower rate of pay. You can see that the factor difference is really big, in the neighbourhood of 2½ times as high as the experienced drivers' crash risk. The probability of them having a crash is way higher. This will not surprise you: experienced truck drivers are safer. This gets better by about the first year, plus there are so many quits involved here. When truck drivers find out how rough this job is they drop out, so it's only the survivors that end up at the bottom, and they are better. Meanwhile, we've had a lot of people injured and a lot of damage and maybe people killed as a result of this low-paid policy, and the higher paid policy worked. But how does it work for the company?

Here's a measure right here on the graph. Here is the measure of driver productivity by experience. The experienced drivers are on the top line and the inexperienced drivers are on the bottom line. The experienced drivers can produce almost 10 per cent more productive revenue miles per month than the inexperienced drivers. That stays true for the entire period of observation that we have. This is a really big thing. It says they would be better off if they hired more experienced, better paid drivers. So higher paid drivers who are more experienced are more productive and stable. They get about 1,000 more paid miles per week. They have a significantly higher attention rate. That is a really big factor. The fleets that have lower turnover are much safer, because the workers on new jobs are more risky. The workers who've been on the job for a while and have some tenure are safer.

What's the return on investment? We measured the expected net present value. We estimate this. The ROI shows having high pay for experienced drivers pays off. This is smart for the company as well as being smart for public policy. Higher paid experienced drivers have a 285 per cent higher expected net present value than the lower paid experienced drivers. This is not what, in economics, we would call the zero sum game: 'Whatever I pay you is what I lose.' This is a joint games thing. This is what people in the bargaining look for, to find where both sides can win. This is a win-win. The expected net present value to the firm of the higher paid experienced drivers is almost $10,500 greater than the expected net present value of the lower paid inexperienced drivers, and this is stable over multiple years. This is a really stable result. It shows that safety pays and it pays to be safe.

I want to show you one more thing, that isn't my own research at all. I contributed by talking to these people when they asked me to. It's detention time. This is a study of detention time done by the Office of the Inspector General, at the US Department of Transportation. Detention is the extra time during which cargo owners hold up driver pickup or delivery. Detention time adds risk, and they were able to determine that. This is data they were able to get that I wasn't able to get. Detention time is defined as holding up a driver for more than two hours. Holding up a driver for any time is a problem if the driver is not paid. Holding up for more than two hours is a really big problem or a much bigger one.

Before 1980, before deregulation of the industry, 60 per cent of all truck drivers were teamsters. Collective bargaining contracts required payment for all work time, according to the FMCSA definition of work. I'm pointing back to the definitions of work that I talked about before, not the FMCSA definition of work that developed pretty much during the Bush administration. It was two administrations ago, during the 'oughties', I think they call them. After deregulation, unionisation in the US for trucking dropped somewhere around 90 per cent, and shippers still expected all this free time. So non-union trucking companies could not collect from the cargo owners and they stopped paying drivers.

Well, the drivers didn't get paid, so, as a result, the drivers couldn't collect on the time that they put in. One of the things that we've done to try to solve this problem is this electronic logbook. I think you guys have tachographs and maybe something similar to this. It's supposed to be that you can't rig it or cheat it, but I found out that it's systematically rigged and cheated, and I'll explain how. An ELB—electronic logbook—cannot determine what the driver's doing. It only knows whether the truck is moving or stopped, and it knows where this is happening, but that's all it really knows. FMCSA inspectors have to take the word of the driver and the carrier that they really are off duty. When you show up to a shipper or a consignee and they tell you, 'Wait a couple hours out in the lot and we'll get to you when we get to you,' you log off duty. You're supposedly off duty and free to go about your business, but you're waiting for them to call you. This is not off duty. This is trust without verification, as Ronald Reagan called it. Drivers log 'off duty' from what the FLSA defines as 'work time', because they don't get paid for it and because their bosses tell them to do it. They are required to do it.

The economic principle behind the intention is that people will consume an infinite amount of a free good, so we're back to time is money. So shippers and receivers, clients have no incentive to conserve free carrier delay time and labour delay time. They want that truck ready when they want it, and they really don't care about the driver for a trucking company sitting there, twiddle their thumbs waiting to move the freight. The ATA estimates that an average length of haul today, for a long-haul load, is around 550 miles. That's about one day's drive. So that means the average driver may load and unload once a day, but of course, if he gets held up, he gets demerged or he gets held up in detention, he won't get that next load off at all. He won't get that next load on, either. So, the drivers give away up to four hours of free time every day. That unpaid delay time values carrier and truck and motor carrier, and driver delay time, at zero. Again, consumers will consume an infinite amount of whatever is free.

The OIG detention times study has some powerful results for safety. The first 15 minutes of detention time, which is over two hours, increases the average expected crash rate by 6.2 per cent. That causes one additional crash per 1,000 power units; it causes more than 6,500 additional crashes per year. Crashes may increase just because the detention time increases. A really bold point here is every five-percentage-point increase in the proportion of stops that result in detention is associated with a 4.7 per cent increase in the expected crash rate. What FMCSA found in their 2014 study is that 10 per cent of all stops experienced two or more hours of detention time and that every one of those detention times, that 10 per cent, averaged 1.4 hours, which means that 10 per cent of all stops had total stop times averaging 3.4 hours for which the driver and the carrier won't get paid. This is not a good situation. It's not tenable for safety and it's not right.

Detention time costs money. Time is money. Detention is associated with between $1.1 billion and $1.3 billion lower annual earnings for for-hire CMB drivers in the truck load sector. That's between $1,281 and $1,534 per driver per year. I will say to anybody who will listen: if you have to write to your employer a cheque for between $1,281 and $1,534 per year as a tax for going to work, you might quit your job too. It's really not a big surprise. People act awfully surprised when truckies quit their jobs because they misunderstand what they do. Detention reduces motor carrier net income by between $251 million and $303 million per year. Unpaid labour time contributes to excessive driver labour time. Excessive labour time drives up crash risk. And so there are policy implications for this.

Higher pay rates and pay for all work time will reduce drivers' incentives to work illegal hours. Requiring pay for all labour time would reduce the incentive to log unpaid work time as off duty. So driver incentives will line up with our policy objectives, what we want to achieve; carriers could not whipsaw the drivers, cargo owners and clients could not whipsaw the carriers, and neither could race to the bottom for cheap labour; and it might make truck driving attractive again—sort of an echo of a current meme. Summing up, the low road costs the economy billions of dollars every year. It encourages an inefficient use of all resources. It's expensive in safety and health costs, and those expensive safety and health costs reduce gross domestic product, which is national income, which explains why economic forces explain safety and health outcomes.

I'm quickly going to walk through all of these—the exam happens at the end! These will be available to anybody who wants them. They are supplemental and reference resources that I've got. My bottom line, to sum up—glad to stay as long as you like to answer questions, but I recognise we're probably running short on time—is the economic approach to safety and health points the way to policy solutions for everybody's benefit. Everybody wins if we do this. Safe freights will save lives. It will allocate resources efficiently, and it will grow the economy. Thanks very much for being patient with me as I rattled through these things so fast. Back to you, Glenn Sterle.

CHAIR: Dr Belzer, thank you so much for that. I'll just give my colleagues Senator McDonald and Senator Hanson the chance to unmute so that I can go to them. Just quickly, based on my extensive work with the teamsters and with President Hoffa and talking to many, many truck drivers in the United States, I know our problems are very similar. This should come as no shock to you, Dr Belzer, as you do know that. There are a couple of things I will say here, because I do want to get to the top of the supply chain squeeze. In Australia, we have been a very productive industry—there is no doubt about that—but in Australia productivity means 'how much more of your crap can we cart for nothing.' During my last visit to the States I noticed that you've started road-training through a number of states, which we've been doing for many years, but it doesn't equate to the company getting paid any extra, and, to this day, I still struggle with why there's forward loading and back loading. I don't know if you have that rubbish, but we do. One thing you do have on us is that at least in the United States—thumbs up from Dr Belzer, you do have back loading—you have the common sense to have a decent sized cab. We aren't even mature enough to have that over here, because then we talk about increased axel loadings and we can't get enough free rubbish on the back. There you go, Dr Belzer. Senator McDonald or Senator Hanson, do either of you have questions?

Senator McDONALD: No questions, Senator Sterle.

CHAIR: Senator Hanson?

Senator HANSON: I was very interested in your presentation about the hours they work and what they get per mileage. Senator Sterle, you probably know more about it, how does Australia compare with America on the mileage and the hours they work? Are we on the same path? Where I'm going with this is that, if we don't address the inland rail, and that's what this is about, and about cabotage, then we're going to have more trucks on the road. You say about 5,000 truck accidents a year happen in the States. Are we headed down the same path, or are we there already?

CHAIR: I think, in fairness to Dr Belzer, yes, we do have similar arrangements. We have decent employers who do it properly and pay by the kilometre and then they top it up with hourly payments, but, unfortunately, we have a lot that pay kilometres and then they don't pay the actual kilometres worked. It's the same as what Dr Belzer said, that we have the same problem over here, and it has been evidenced in this committee. Many truck drivers are waiting at distribution centres, at clients' premises, where they do not—and these are their words—put those hours in their work diaries or logbooks, because then their hours to drive after they've loaded or unloaded are limited. It's exactly the same as the United States.

And in terms of the inland rail, Dr Belzer, in this nation there's a project to build a massive inland rail along the eastern states spine. The project is so flawed that it will probably not get used by the rail industry; hence our freight task is going to double by 2030. Unfair to you, but that's what's happening over here in Australia. It's probably no different to the United States.

Dr Belzer : We call it intermodal, so we have intermodal in the US. I would imagine that with intermodal, if you could get the rail lines to survive, you could run from Alice to Western Australia, to Perth. You could even run from some of the eastern cities to Perth. In the United States, anything shorter than 500 miles is not economically viable for intermodal. It costs too much money to take a container off a truck, put it on the train, take it off the train and put it on a truck. That's what our dynamic tends to be. Our countries are approximately the same size. I never thought that there was so much open space as I have found in the Australian outback, it does stretch out for some way, but other than that we're really pretty similar. We have good companies that pay by the mile and, as you say, top off with hourly payments. That's the way it was in teamster contracts before deregulation. I can't speak to the current teamster contracts—there are relatively few of them—but the idea is that you pay miles and hours. It's hours when you're stopped and mileage when you're driving. If we had that kind of system, most of our problems, I think, would go away. You'd have to pay people minimum wage once you were actually paying people, but they don't even have to pay minimum wage in the trucking industry for non-driving labour time, as long as your overall time, including all of your mileage, is still minimum wage. But you have no record of the number of hours worked, because truck drivers are not logging them. You get in a line at a grain terminal to pick up grain, and you're going to sit in that grain terminal and inch your way forward. You might spend eight to 10 hours in that line to get forward. You're going to mark that as off duty, and then you're going to go to work. You might well be overloaded. One of my colleagues in Australia told me that was a common problem with grain. I think that's true in our country as well. So this is not the solution. They need to be logging the drivers in when the they come in. They get in line—that's logged in and that's off duty—and let the company move the unit. If you're not paying the drivers enough money, you won't be able to keep drivers. So companies either pay drivers more money or go out of business. That's fine. Pay the people more money. Pay safe rates.

Senator HANSON: To touch on cabotage in America—I don't know how they term it, Senator Sterle—in the trucking industry, is that in competition with domestic freight travel?

Dr Belzer : I'll try to answer that, but tell me if I've misrepresented something. First of all, 'cabotage' is the French term, and it refers to picking up and delivering in different countries in the European Union. That's the first representation of that. I have a paper on that in a book that was just published, in 2020, in honour of Professor Michael Quinlan, and I'd be happy to refer you to that. It's described in some detail. In the United States, cabotage is not legal. But we're talking about cabotage between the US, Canada and Mexico—different countries. That means that a truck that comes with a load from Canada to the United States has to have a prearranged return load back to Canada or a prearranged ongoing load to Mexico and vice versa, but you cannot do work within the United States. So a Canadian trucker and a Mexican trucker cannot compete directly with a US trucker doing that work. Whether they do or not is another question. The data are not very good, so it's really hard to tell. Australia is one modest-sized continent—the same size as the United States is—but it's all one country. So that's why I don't quite understand your local definition of 'cabotage'—are you talking about state-to-state cabotage?

Senator HANSON: We have foreign shipping in Australia. They allow it to move the cargo around, which does affect our rail freight here. So they're taking the business away, which is putting strain on our own Australian businesses here.

Dr Belzer : Are these with foreign drivers operating their trucks? This is just like—

Senator HANSON: Foreign ships in Australia.

Dr Belzer : Every time I thought I'd seen the worst, you're going to show me something worse.

CHAIR: Dr Belzer—

Dr Belzer : This is exactly what we found when we studied Europe. We found that, since the European Union expanded, most of the work of western European truck drivers, many of whom are union members, is done by eastern European, eastern bloc, drivers, and some of those have actually laundered licences. They might be from Kazakhstan, the Philippines or some other place—Russia, Ukraine; places that aren't in the European Union—and they're working for about 10 per cent as much money as the western Europeans; there are terms for this in European Union language, but I mean the original European Union drivers. That sounds like the problem you have with drivers who are literally imported from poor countries at those rates to compete directly with the drivers in western Europe. When I realised that was happening, that was a shocker. That's why I think that paper and the research that we did are a big takeaway for me. We are actually a more conservative country. We don't allow that. But the European Union—

CHAIR: No, you've got the Jones Act.

Dr Belzer : What's that?

CHAIR: Senator Hanson was going particularly along applying coastal trade here, which is affecting our rates. Because rail is getting squeezed out, trucking is trying to compete with that, which they can't. So our—

Dr Belzer : Got it. That's cabotage around those ports.

CHAIR: Cabotage around the coast, yes.

Dr Belzer : Okay, I know what you mean.

CHAIR: You've got the Jones Act. I hope that helps, Senator Hanson.

Dr Belzer : I didn't realise that. Thank you very much.

Senator HANSON: Thank you very much, Chair. No further questions.

CHAIR: Dr Belzer, before I let you run away, my argument and a lot of the arguments here in Australia are that this is not an employer-versus-driver situation. A major proportion of the representatives of the road transport industry in Australia want to work together to push back against the push from the top of the supply chain. We don't see it as a boss-versus-employee thing, which I know you don't either. But I also want to raise something with you. No. 1, if companies are paid properly, then it goes to the drivers, bearing in mind we still have the ratbags out there that will pocket the extra money; that's everywhere. But you said that if the pay rates are better—and you've had 25 years of this, and I haven't; I've been on the road doing my stuff, and you've been studying all these projects and all that—it would help to go to easing the pain of the driver shortage. Here in Australia we have a shocking driver shortage, but we also have—and I know you might not have the answer to this—companies that do pay properly. A lot of small-family businesses treat their drivers like they're family, and they pay properly. Yet we still can't attract drivers. We've been banging our head on this wall for 25 years. Can you help us out: is that similar in the States, or do you have reasons why that happens?

Dr Belzer : It's similar. Our small folks, I suppose, are the owner-operators, but it gets more complicated because they have to have operating authority. You don't have operating authority in Australia, and we do. Maybe it's an operating authority of one. But what happens is there are companies—Hunt was a good example—paying the right wage and doing the right thing. It had a terrific result—more productivity, safer, less turnover; what's not to like here? It was really hard for them to measure the profitability. It took a long time to try to work this through—my apologies to Hunt; their CEO asked me to figure this out about 15 years ago, and it took a long time to figure it out with the data that we had—but the bottom line is that it does make sense. The return on the investment—that is, the ROI—is positive if you pay people more money. The reason for safe rates is that you have to have a common denominator, not the lowest common denominator but one that achieves safety. If you achieve that, you're going to be much more able to attract drivers into the industry. So I think there is no such thing as a driver shortage; it's a shortage of pay. If we pay people more money and we make the conditions better, it will attract people in to do the work. That's really, I think, the bottom line.

CHAIR: I agree with you, but there also needs to be an enforcer there so the good operators, the good employers and the drivers who want to make a decent wage, aren't penalised, because here in Australia we have a shocking record of enforcement. They're missing in action. Is that the same in the States?

Dr Belzer : I don't want to say anything too impolitic, but I think the FMCSA has been barking up the wrong tree, and I think it's because they've been required to. They're doing a really good job, so I'll say this: especially compared to European Union enforcement, which is abysmal, the United States comes out smelling like a rose. I was rather shocked. We have an enforcement regime and we hold people to it. It's still hard to do. It's kind of like whack-a-mole. I don't know if you have the term 'whack-a-mole', when there's always somebody popping up to become the dirtbags or something like that. There's always somebody out there. They do a pretty good job of that, but my argument has been that, because they don't look at how drivers are paid and they don't look at the economics of the industry as driving the safety, they're missing the biggest factor.

I was on a panel, the National Academy of Sciences panel that reported to them, that did an analysis of their data and said, 'This is what you've got to do to improve your data quality,' and I think they're improving their data quality a lot. Again, they really are. It's quite astounding data and it's all public. It's all available. It's not secret. The European Union data—they won't give it to me, because they think I might come up with something they don't want to see. Okay. Fine. But in the US, they're really doing all that. But, because of this blind spot on pay, they're not collecting that information.

We recommended that they collect information, at least the average pay per full-time driver at the firm. If we had that by company, we would be able to do more precisely what Shenyang Ju and I did in the paper I just showed you that is going to go out to review shortly. That paper predicts really strongly that pay is the biggest factor. I think the way you solve the problem is you collect data on the pay and then you run those estimations and see just what that number is. I think you'll find that the biggest bang for the buck is going to come from the buck. Well, I use the term 'buck' and it's an American term, but you know perfectly well what I mean: it's all about the money.

CHAIR: It is exactly and I can't argue with you. Dr Belzer, I'm sorry but we have run out of time. Thank you so much, and greetings from your friends here in Australia. I look forward to when our earth returns to normality and we can catch up.

Dr Belzer : I've had my vaccine. I'd bring you one too.

CHAIR: I'm not allowed to have one for a few months. There are people who need to get it before me. But don't worry, I will take it as soon as it comes. Thank you so much. Hopefully, we'll catch up in the near future.