Both hands on the wheel

The 20-minute version

20 min read · the full book is 55 min

You're on the high street at ten past eight, third car back at the lights. The phone's in the cupholder and it lights up. You don't decide to look. Your hand is already there, and by the time the light goes green you've read the message, thought of a reply and typed half of it. The car in front moves. You move. You finish the reply with your thumb at fifteen miles an hour, looking up between words, and put the phone back with the feeling of having done nothing at all.

Or it's the A-road at fifty. The buzz. You leave it for a bit, because you're not the kind of person who looks at fifty. Then the road straightens and empties, and you hold the phone low against your knee where nobody in the next lane can see it, and read. Two seconds, maybe three. Your eyes come back up and everything's where you left it.

It was a message about a parcel.

Or it's the map, which has told you to turn where there's no turn, and you tap it to see what it meant, and tap it again, and for about six seconds you're a person looking at a screen who is also, in a secondary sort of way, steering.

You know it's dangerous. Everyone does. You've seen the adverts and you may have a rule you keep most of the time. I'm not going to tell you what you already know, and I'm not going to describe a crash. I want to look at the beliefs that let someone who knows it's dangerous do it anyway. That it's only a glance. That you only do it when it's safe. That you're a good driver, and you've done it a thousand times without anything happening, and hands-free is fine, and the message might matter. Each has something true in it. Each, I'll argue, is working for the habit and not for you. Then there's a way of driving where the itch to look doesn't get a vote, because the phone isn't where your hand can reach.

The law, briefly. In Great Britain it's illegal to hold and use a phone or any similar device while driving, including at traffic lights and in queues, and the GOV.UK page on it is the source for anything I say about the law. Hands-free use is allowed.

If you drive for a living, nothing here needs you to be unreachable for a whole shift. If you're the parent of someone who passed their test in the last two years, there's one fact further on you'll want them to know.

One thing before you go on, for a week. Every time you look at the phone while driving, at the lights or on the move, make a note once you've parked. Where you were, roughly what speed, and what the message or the tap turned out to be about. Don't change anything yet.

Twenty-seven metres

Nobody reads a message on their knee at fifty because they've weighed it up. The phone buzzed, the itch arrived, the road looked empty, and looking was the thing that made the itch stop.

It's worth being fair to what the peek gives you. The buzz opens a small question, who is it and what do they want, and the peek closes it. It turns twenty minutes of dead time into twenty minutes in which you can do something. It keeps you reachable, which matters if you have children or a parent who might ring. And every single time, it gives you the experience of nothing happening. You looked, you looked back up, and the road was where it was. That last one is the strongest.

Start with the glance itself, because the whole case for it is that it's short.

Bruce Simons-Morton and colleagues fitted the cars of forty-two newly licensed teenage drivers with cameras and sensors and recorded them for eighteen months. Then they looked at the six seconds before every crash and near-crash, and at random stretches of ordinary driving, and measured how long eyes were off the road. The single longest glance was what mattered. When it went over two seconds, the odds of a crash or near-crash were 3.8 times higher, and 5.5 times higher when the task involved a phone. Sheila Klauer and colleagues ran the same kind of study with those novices alongside a hundred and nine experienced drivers. For the novices, dialling raised the odds more than eightfold. For the experienced drivers it raised them two and a half times. Experience shrank the number and didn't remove it.

Now some arithmetic of my own. At thirty miles an hour you cover about thirteen metres a second, so a two-second glance is twenty-seven metres of road you didn't see, about six car lengths. At fifty it's forty-five metres. The glance feels like nothing because nothing was in those metres this time, and the road doesn't send a reminder when something nearly was.

"I only look when it's safe."

This is the belief that does most of the work. You don't look on a roundabout, or in the rain with a lorry alongside. You choose the straight, empty road. The trouble is who's doing the choosing.

David Sanbonmatsu, David Strayer and colleagues put drivers in a simulator, some talking on a hands-free phone and some not, and afterwards asked everyone how safely they'd driven. The drivers off the phone had a fair idea, and their ratings matched their errors. The drivers on the phone made more errors, and their ratings had no relationship to their errors at all. The phone took away their ability to tell.

Paul Atchley and colleagues found something stranger in a survey of three hundred and forty-eight young drivers. When drivers chose to start a text themselves, they rated the road at the time as safer. The road hadn't changed. Once you've decided to look, the road agrees with you. Those drivers were surveyed, and Sanbonmatsu's were talking, so I'm extending both. But think about it honestly. Have you ever looked at the phone and thought, this is a bad moment? The moments always seem fine, because the part of you that judges has already been recruited.

"I'm a good driver." You probably are. In 1981 Ola Svenson asked drivers to rate their skill and safety against the others in the room, and most put themselves above the middle, which can't be true of most people. Lina Koppel and colleagues repeated it in 2023 with just over twelve hundred people and got the same result. So the feeling is one nearly every driver has, including the ones it isn't true of. And Sanbonmatsu and Strayer, in another study, found that the university students who multitasked most, including those reporting the most phone use while driving, scored worst on a test of actually doing two things at once and had the most inflated view of their own ability. That's one study of students. Put it next to Svenson and the picture holds. Being good at driving means being good at what you do with your eyes on the road. Nobody's reactions cover twenty-seven metres they didn't see.

Under all of these sits the clean sheet. You've done it a thousand times and nothing has happened. What better evidence is there than your own record?

Redelmeier and Tibshirani, in 1997, took six hundred and ninety-nine drivers who'd had a collision and owned a mobile phone, and got the billing records for the day of the collision and the same time a week earlier. A driver on a call was about four times as likely to crash as the same driver not on a call. Each driver was compared with themselves, which takes away the argument that phone users are simply worse drivers.

Think about what fourfold means for a rare thing. A collision on any given drive is very unlikely, and four times very unlikely is still unlikely. So someone who looks on every drive will, most years, have nothing happen, and their clean sheet gets a year longer, and they feel surer than ever that they're the exception. That's what a fourfold risk of a rare event looks like from the inside, right up to the drive where it doesn't. Neil Weinstein described the general pattern in 1980. People rate their own chances of bad events as below average, and personal experience feeds it. The more often something hasn't happened to you, the less likely it feels.

The thousand times are a sample of what happens when you look and nothing's there. The one that matters isn't in the sample yet.

Loopholes

The law allows hands-free, so the problem must have been the hand. A phone in a cradle, on speaker, leaves your hands on the wheel and your eyes ahead. What's left?

Attention, which was doing the driving. In 2001 David Strayer and William Johnston had people drive a simulator while doing various things. Listening to the radio didn't hurt their driving. Listening to a book didn't. What hurt was conversation, on either a handheld or a hands-free phone. It doubled the number of signals drivers missed and slowed their reactions to the rest. Two years later Strayer, Frank Drews and Johnston found drivers on a hands-free call couldn't recognise roadside billboards they'd looked straight at. Eye tracking showed their eyes had been on the billboards. Their attention hadn't.

The crash figures agree. Redelmeier and Tibshirani's drivers showed no safety advantage for hands-free. Suzanne McEvoy and colleagues studied four hundred and fifty-six drivers in Perth who'd ended up in hospital after a crash, and found phone use in the ten minutes before was associated with about four times the risk, with no significant difference between hands-free and handheld. Their conclusion was that hands-free is not any safer. Two meta-analyses, one of twenty-three studies and one of thirty-three, found the same pattern for both kinds of phone, mostly in slower reactions, and the second found drivers on the phone didn't make up for it by slowing down or leaving more room. The law draws its line at the hand, where a police officer can see it. Your attention doesn't stop there.

"It's no different from talking to a passenger." Drews, Monisha Pasupathi and Strayer had people drive while telling a friend about a time their life had been in danger, the friend either in the passenger seat or on the phone. Phone drivers made the most errors. Passengers talked about the traffic, and both people simplified what they were saying when the road got harder. Samuel Charlton found passengers slowed their talking as a hazard came up, and callers didn't, even callers who could see the driver's view through a window. A passenger is a second pair of eyes with a stake in the outcome. The person on the phone doesn't know you're at a junction, and keeps talking when the van pulls out.

"I use voice, so I'm not looking." Strayer, Joel Cooper and colleagues had two hundred and fifty-seven people use the voice systems in ten new cars, then gave them five days to practise. The mental workload was rated moderate to high, practice didn't fix it, and attention took time to come back even after the task had ended. A similar study of voice assistants on phones, on suburban roads, found talking to the phone was as demanding as a hard memory test. Voice to answer a call beats a hand on the phone, and the law agrees. It still isn't free.

"It's not like drink-driving." Strayer, Drews and Dennis Crouch compared drivers on the phone with drivers at 0.08 per cent blood alcohol, the legal limit in England and Wales, in a simulator. The drinkers drove more aggressively. The phone drivers braked later and had more collisions, and the authors concluded the impairment from phone use can be as profound as driving at the limit. One simulator study isn't a road, and a text isn't a drink. But the feeling that they're in different categories doesn't come from a measurement either. And notice the rule you already keep about drink. It doesn't bend for an empty road or a short trip or a good driver. You don't say you only drink-drive when it's safe.

"It might be urgent." Get out your list. I'd guess none of the messages needed an answer in the next five minutes, and most didn't need one in the next hour. A parcel. A group chat. A friend asking if you're around. The boss saying call me when you can, which carries its own answer. The buzz sounds the same for all of them, and you didn't know what the message said when your hand moved. You looked because it buzzed. Urgent was the story afterwards. The law's one emergency exception is calling 999 or 112 when it's unsafe or impractical to stop, and that's a call you'd make, not one you'd receive.

Then the comfort. If it were really dangerous, you'd have stopped.

Atchley's young drivers rated texting while driving as very risky, riskier than a phone call. Ninety-two per cent of them read texts while driving, eighty-one per cent replied, and seventy per cent started them. How risky they thought it was barely predicted whether they started texts and didn't predict reading or replying at all. Heidi Nemme and Katherine White, following a hundred and sixty-nine students for a week, found texting was predicted by attitude, by what friends did, and by what people felt they ought to do. Believing it was risky didn't move much.

So knowing doesn't stop it, because the habit isn't run by what you know. It's run by the buzz, the hand, and the thousand times nothing happened. That's why the adverts haven't worked, and why this isn't trying to be one. The knowledge is filed. The hand doesn't check the file. Whatever changes the habit has to be something other than more knowing.

"It's only at the lights." The car's stopped, so nothing can happen. The law disagrees, and the offence applies at traffic lights, in queues, and when the car has switched its own engine off at a stop. And watch what a look at the lights actually does. It starts stopped and finishes moving, because the light changes when it changes and not when you reach the end of the message. Thomas Dingus and colleagues put cameras in the cars of more than three and a half thousand drivers for three years and found that reaching for a handheld phone carried nearly five times the crash risk of ordinary driving. Reaching, the thing you do when the light goes green and the phone needs to go somewhere.

"Everyone does it." The same cameras found handheld phone use took up more than six per cent of driving time, with the crash risk 3.6 times ordinary driving while it went on. Distraction was present in sixty-eight per cent of the nine hundred and five crashes they caught. The people in those cars had looked around at the lights and seen everyone doing it too. Nemme and White found that what friends did predicted texting over and above people's own attitude. So the norm is a lever, and it's currently pulling the wrong way. You're one of the everyone. When you stop, the norm in your car moves by one, and the child in the back is watching.

The glovebox

The law first, once, because one part of it matters more than the rest.

It's illegal to hold and use a phone, sat-nav, tablet or any device that can send or receive data while driving, for any reason, online or offline, including in flight mode. The 2022 change closed the old gap where only calls and texts counted. The exceptions are calling 999 or 112 when it's unsafe or impractical to stop, being safely parked, a contactless payment while the vehicle isn't moving, and parking the car remotely. Hands-free is allowed as long as you don't hold the device at any point, and it mustn't block your view. The police can also stop you if they think you aren't in proper control because you're distracted. The penalty for holding and using is six points and a two-hundred-pound fine, and in court the fine can reach a thousand pounds, or two and a half thousand for a lorry or bus, with a possible ban. Penalties change, so the GOV.UK page is the source.

If you passed your test in the last two years, six points means you lose your licence. You take the test again. A new driver seen reading one message at one red light is, at that moment, back to being a learner.

The law isn't what stops people, though. It's worth counting what the peek costs that isn't a crash.

Strayer's voice studies found attention took a while to come back after a task ended. So a drive with six peeks in it isn't a drive with twelve seconds missing. It's a drive that was never quite all there, each buzz landing on a driver who hadn't fully returned from the last.

And the drive itself used to be something. It was one of the last places where nothing was asked of you except what you were doing. You couldn't answer the email. You could look at the road and think, or not think, and arrive with twenty minutes behind you as a gap between the desk and the kitchen. The phone came into the car because it was in your pocket, and within a year the drive was a place with a phone in it like everywhere else. Here's what I've noticed, and it's an observation, not a finding. With the phone out of reach, the first five minutes are restless. The hand goes to the cupholder and finds nothing. Then it goes quiet, and you arrive less scattered than you're used to.

Now the method, which is small.

The phone goes somewhere your hand can't get to it at a red light. The glovebox, if it's out of reach from the driver's seat while you're driving. The bag on the back seat. The pocket of a coat on the back seat. The boot, if you're someone who'll reach behind you. Anywhere that would mean undoing your seat belt or turning round. The cupholder won't do, and neither will the passenger seat, even face down. Face down is what everyone tries first, and it lasts until the second buzz. A phone within reach is a phone you'll reach for, so the rule has to be something your hand runs into, because your judgement has already been recruited.

It goes there before the engine starts. Once the car's moving, out of reach means stopping, and you won't stop.

Turn on the drive setting. Most phones have one that detects driving, silences notifications and can send an automatic reply. It doesn't replace out of reach, because the itch turns up without a buzz often enough, but it removes most of the buzzes. It usually lets a repeated call through, or can be set to, and that's your emergency channel.

Then tell people. Look at your list and see whose messages you looked at. It's usually four or five people. One sentence each. "I don't see messages when I'm driving, so if you need me in the next half hour, ring twice." Before, a slow reply from the car was you being slow. Now it's you driving, and they knew. Part of the pressure at the buzz was real, because people did expect a reply, and this takes that part away. What's left is the itch on its own, which is much easier to leave in the glovebox. If your job needs you reachable, have the same conversation with whoever needs you. Most managers would rather you weren't reading texts at fifty in the company car.

Anything with a screen happens with the handbrake on. The map is set at the kerb, destination in, route showing, in a cradle at eye level where you can see it without touching it. It's the one bit of kit worth buying. Music gets chosen in the car park, an album long enough for the drive. If the map goes wrong on the move, drive on and let it re-route. If you really need to touch it, pull over, which takes about the minute the peek used to. If someone's in the car, the phone is theirs. They can see the road, so let them do the map.

Calls are the grey area. My view is that hands-free is legal and costs attention, and you should take fewer than you think. The clean rule is that calls happen parked. If you won't have that, answer by voice, keep it short, and say you'll ring back. Don't pick the phone up.

So the rule is one sentence. The phone is out of reach before the engine starts, and anything with a screen happens with the handbrake on.

Nothing in it asks you to judge what's safe or urgent or only a glance. That's the point. Those were the beliefs.

Peter Gollwitzer and Paschal Sheeran's meta-analysis of plans in the form "when this happens, I'll do that" found they reliably help people act on intentions they already hold. Make three. When I sit in the driver's seat, the phone goes in the glovebox before the seat belt. When it buzzes in the glovebox, I say "later" out loud and keep driving. When I need the map or the phone on the move, I pull into the next safe place and put the handbrake on.

Say two things out loud, once. The standing message to your four or five people. And to whoever's usually in the car, "the phone's going in the glovebox from now on, and if I reach for it, say so."

Then start. The next drive.

Later

The first week your hand will go to the cupholder at every red light and find nothing, and you'll feel slightly ridiculous. That's my expectation, not a finding. Phillippa Lally and colleagues found new habits became automatic over weeks, at very different rates for different people, so I won't give you a number of days. This one has an advantage, though. The car is the same every day, and so are the seat and the glovebox. Cues don't get much more reliable.

A few situations come up often.

  • The sat-nav. A phone in a cradle showing the map, set before you move, is legal and about as safe as a phone in a car gets. The risk is the tap. Holding the phone to see the map is the knee with a road on it.
  • Two phones. Both go out of reach. If the job needs you reachable while driving, that's a hands-free call answered by voice and kept short, a known cost you're taking because the job asks it. Take fewer than it strictly needs.
  • On call. The two-call rule is your channel. Calls ring through, messages wait. If the job sends emergencies by text, raise that with the job. A text you must read within a minute and a car you must drive can't both be done, and it shouldn't be you choosing at fifty.
  • The child in the back. The dropped bottle and the out-of-reach toy get the same answer as the phone, at the kerb. A child can survive four minutes without a bottle, and one old enough to hold the phone is old enough to be the navigator.
  • Driving for a living. The auto reply sends your standing message a hundred times a day, and your licence is your income. The rule's the same. The reason to keep it is bigger.
  • The driver beside you. If you're the passenger and their phone is on their knee, one plain sentence. "Can I do the phone? I'd rather you drove." Most people hand it over, because being given a reason to put it down is easier than being told to.

And one day you'll pick it up anyway. You got in mid-call and the phone went in the cupholder, and you'd read the message at the lights before you remembered there was a rule. Or three weeks in, the glovebox quietly stopped being where the phone goes, and you can't say when.

Lally's volunteers missed days without it materially affecting how the habit formed, and I'd extend that to one look at the phone. What matters is the next drive. Once you've parked, ask which belief was running. "Only a glance" at the lights, "might be urgent" if you'd been waiting on someone, "safe here" on an empty road. That tells you which one still has life in it.

Don't turn it into a resolution. It came back because the phone was within reach, which is a fact about where the phone was. Fix the where. If the glovebox is too easy to skip, use the bag on the back seat. If the phone's in your hand when you get in, finish the call before the door shuts. And don't count clean days. A clean sheet was the problem to begin with.

A Tuesday, a month or so on. You get in, the phone's in your hand from the call you just finished, and you put it in the glovebox without thinking, because that's where it goes now. Seat belt. Engine. The high street, the lights, third car back. Your hand still goes to the cupholder sometimes, finds the empty plastic, and comes back to the wheel.

The A-road. Fifty. A buzz, muffled, from the glovebox. Something in you sits up, the same something that always did, and then sits back down, because there's nothing to do about it. The road straightens and empties, and you notice it's the safe moment, and you notice that noticing it is what you used to do just before looking. The road stays the road for the whole forty-five metres.

Nine minutes later, handbrake on in the car park, you get the phone out. It was your friend asking if you're around at the weekend.

The next time you get in the car, before the seat belt, put the phone somewhere you can't reach it. Just once, and see what the drive is like.