WEBVTT
00:00:00.160 --> 00:00:01.520
There's evidence in the U.S.
00:00:01.600 --> 00:00:12.960
that a platform for nurses to take overtime fees had gone out and purchased credit and debt load data that they were layering in, essentially proxying a desperation score.
00:00:13.119 --> 00:00:17.199
So nurses that had more student debt were offered a cheaper price for their labor.
00:00:17.600 --> 00:00:21.199
Demographics, hosted by Mike Moffitt and Kara Stern.
00:00:22.480 --> 00:00:25.359
The next big hit to your wallet might not be inflation.
00:00:25.519 --> 00:00:30.559
It might be that stores can figure out the maximum you're willing to pay and charge you exactly that price.
00:00:30.800 --> 00:00:35.439
It's called surveillance pricing, and politicians are starting to debate whether Canada needs to ban it.
00:00:35.600 --> 00:00:41.920
Today I'm talking to Vass Bednar, a managing director of the Canadian Shield Institute and author of the book, The Big Fix.
00:00:42.240 --> 00:00:47.439
She's someone who's been writing about the problems with surveillance pricing since before almost any of us knew to worry about it.
00:00:47.520 --> 00:00:49.840
And we're so glad to have her back on the program.
00:00:50.320 --> 00:00:51.439
Thanks for having me.
00:00:51.679 --> 00:00:55.439
There's algorithmic pricing, there's dynamic pricing, and there's surveillance pricing.
00:00:55.520 --> 00:00:57.840
Like, what is surveillance pricing that we're talking about now?
00:00:58.159 --> 00:01:05.359
Surveillance pricing is sort of layers in with that computational data-driven pricing strategy.
00:01:05.519 --> 00:01:12.079
So beyond using a variety of factors like what your competitors are pricing something at, right?
00:01:12.480 --> 00:01:26.560
Um, personalized algorithmic pricing requires excessive data use about who you are, where you live, your behaviors online, sometimes your past purchase history, and sometimes where you are and where you're going.
00:01:26.799 --> 00:01:38.480
So it's a much more invasive pricing tactic that's supercharged by data and much more common in kind of an e-commerce context where there's way more information asymmetry.
00:01:38.560 --> 00:01:43.920
It's kind of just you and a screen, whether it's your phone and a loyalty program or, you know, shopping online.
00:01:44.079 --> 00:01:46.719
I would say that's that's the bigger difference.
00:01:47.040 --> 00:01:56.400
Like the first time I remember hearing about this type of pricing, and it's only sort of similar was with Uber, where you'd have like, you know, I remember people talking about if you're on an iPhone versus an Android, you'd get different prices.
00:01:56.480 --> 00:01:59.359
Like that's like a the bare minimum version of this kind of.
00:01:59.439 --> 00:02:02.719
But then also people started getting used to dynamic pricing with Uber.
00:02:02.799 --> 00:02:04.400
Like what's different with this?
00:02:05.120 --> 00:02:09.120
So dynamic pricing suggests supply and demand constraints, right?
00:02:09.280 --> 00:02:19.360
I think uh a parallel would be airlines, where we became pretty comfortable with price fluctuations that we could monitor, but weren't necessarily tied to the individual.
00:02:19.599 --> 00:02:27.120
So if you were getting a last-minute flight, uh it was maybe more expensive, you're booking far further in advance, there were more seats available, right?
00:02:27.680 --> 00:02:37.599
Now, because of data-driven pricing, there's some evidence that suggests about a thousand different data points are used to calibrate the price of your seat on an airplane.
00:02:37.759 --> 00:02:46.000
And back to the Uber example, you know, also some evidence that if you're uh using a business credit card, Uber will upcharge you, charge you more.
00:02:46.159 --> 00:02:49.680
Again, inferring your willingness to pay and squeezing more there.
00:02:50.000 --> 00:03:28.319
Where I think there's a difference between the kind of supply and demand type of calibrations, or I'll use the euphemism pricing experiments that we've seen with Uber, is external factors like say it's raining, so there's more demand, and they'll tell you, you know, prices higher right now, or um, Blue J just won the World Series, fingers crossed, you know, there's more people looking for rides out there, something like that, versus this ride is going to cost you more vast because you tend to take this route and we know you're very likely to take this route and you're using a business card.
00:03:28.560 --> 00:03:33.039
You already get some places like different apps that'll give you discounts based on things you've purchased before.
00:03:33.199 --> 00:03:43.039
So yeah, I know I get sometimes with like shoppers or Canadian Tire or like Tim Hortons will be like, hey, you have it, I mean we know you like to buy this kind of sandwich and you haven't bought it in a long time.
00:03:43.360 --> 00:03:44.960
What's your favorite thing at Tim Hortons?
00:03:45.199 --> 00:03:50.479
Well, it sometimes it gives me breakfast sandwiches discounts because I like their breakfast sandwiches and they're like good in a pinch, you know?
00:03:50.560 --> 00:03:57.439
So I'm like, sometimes it'll be a little bit of time since I've bought a breakfast sandwich because they tend to be like, you know, I'm going out on a route, have to go drive somewhere far or something.
00:03:57.520 --> 00:03:59.520
Like that's when I get it so it doesn't happen all the time.
00:03:59.599 --> 00:04:04.000
And then I'll get a notification like, hey, you haven't bought a breakfast sandwich in a while.
00:04:04.159 --> 00:04:05.759
Would you like a discount on that?
00:04:05.840 --> 00:04:07.520
If you buy it today, you'll get this discount.
00:04:07.599 --> 00:04:12.479
And to me, I'm like, that feels like a form of surveillance pricing that already exists, right?
00:04:12.800 --> 00:04:28.879
It's a form of it, but I would say the more insidious is something like uh the Taco Bell app being able to infer when it's your payday and increasing the price of your Cordita deal versus, you know, incentivizing you to come back to Tim Hortons.
00:04:29.040 --> 00:04:40.800
You know, Tim Hortons is actually a great example because they were investigated by the Federal Privacy Commissioner and many provincial privacy commissioners for their invasive data collection in their loyalty program, right?
00:04:40.959 --> 00:04:42.959
Tracking you down to where you are.
00:04:43.120 --> 00:04:52.560
Um, an Epic study from last year found that Target would make uh televisions in their store uh more expensive the closer you were to the store.
00:04:52.639 --> 00:04:56.240
So that's kind of inferring a probability that maybe you're going to make a purchase.
00:04:56.319 --> 00:05:00.639
Again, it makes it harder for people to understand what things actually cost.
00:05:00.879 --> 00:05:08.000
We're losing price universality as Canada considers whether and when we're going to tolerate this.
00:05:08.160 --> 00:05:12.879
Because you're right, there may be some instances where we think it's appropriate, say for discounting.
00:05:13.199 --> 00:05:21.040
We need to decide if we're looking at it from a privacy perspective, consumer protections perspective, marketing, you know, deceptive marketing angle.
00:05:21.199 --> 00:05:23.040
There are different ways to kind of get at it.
00:05:23.199 --> 00:05:28.879
But to your point that it's increasingly becoming a feature of how we experience markets, like absolutely.
00:05:29.040 --> 00:05:38.879
Um, even geographic pricing, which is more blunt, but like I used to think that my secret deodorant costs the same at every single shop or struck mark, right?
00:05:39.040 --> 00:05:40.160
And that's just not the case.
00:05:40.240 --> 00:05:40.959
And that's okay.
00:05:41.120 --> 00:05:52.480
And a funny phrase that I I don't know if I'll ever say again in my life, but there was like a good Toronto Sunpiece from a few months ago looking at Big Macs costing different amounts at different McDonald's.
00:05:52.639 --> 00:05:57.040
So it's not that the pricing differential itself is egregious.
00:05:57.279 --> 00:06:07.439
It just reminds us that that pricing strategy defies what we may have as an expectation of what the norm is with pricing when actually something very different is happening.
00:06:07.680 --> 00:06:15.199
So I get that the idea is that like they're using your personal information, and that's kind of like it seems like people think that's a little bit like icky when you people are talking about this.
00:06:15.360 --> 00:06:22.480
I see like Abacus Data did some research into how people felt about it and they found 83% of Canadians wanted it banned or regulated.
00:06:22.639 --> 00:06:27.680
They found only 7% of people are convinced by the argument that this is just supply and demand at work.
00:06:27.920 --> 00:06:30.800
So I'm trying to understand like, what is it about it specifically?
00:06:30.959 --> 00:06:35.920
Is it just that they're using your data in a way that like that is it that they're using more data than they did before?
00:06:36.079 --> 00:06:38.879
Because we've had for a long time things like seniors discounts.
00:06:38.959 --> 00:06:41.920
We've had coupons offered to customers who've purchased before, as I said.
00:06:42.079 --> 00:06:46.800
We have like, there's lots of different ways that the companies have targeted people using their data for so long.
00:06:46.879 --> 00:06:49.839
And usually it's been, I think, you know, you use your loyalty card or their app.
00:06:49.920 --> 00:06:54.560
And so they're collecting the data they get from your app, from the app that they created that you're choosing to use.
00:06:54.720 --> 00:06:56.720
So what is different about it now?
00:06:57.120 --> 00:07:05.120
What's different is that the discounts, let's let's stay with discounts because you know, a senior's discount is a good example, right?
00:07:05.759 --> 00:07:08.560
Is more blunt and uniformly experienced.
00:07:08.720 --> 00:07:18.240
So, you know, if you're over a certain age, and I believe, you know, for shoppers, it has to be a certain day as well, you have equitable access to that discount.
00:07:18.399 --> 00:07:26.800
That discount is not predicated on you having a mobile device, using a mobile device when you shop, you know, sharing additional information.
00:07:27.040 --> 00:07:30.959
Couponing is a very, I'll use the the D word, dumb, right?
00:07:31.199 --> 00:07:34.079
Dumb in the sense that it's not supercharged by data.
00:07:34.240 --> 00:07:39.920
You know, you and I can still access the same deals when they come in a flyer.
00:07:40.160 --> 00:07:50.639
Where we can't see each other's universes, Cara, is in these closed app ecosystems that are more finely calibrated to who we are.
00:07:50.800 --> 00:07:55.279
Are they necessarily discriminating against us, you know, as individuals?
00:07:55.439 --> 00:07:57.600
That's where it gets tricky with privacy law.
00:07:57.759 --> 00:08:01.839
Privacy law is designed for harms that individuals experience.
00:08:02.079 --> 00:08:07.199
But more often than not, personalized pricing is about segmenting, right?
00:08:07.360 --> 00:08:10.639
Creating particular cohorts, particular profiles.
00:08:10.800 --> 00:08:13.120
And that needs to be part of our conversation.
00:08:13.439 --> 00:08:22.800
How comfortable are we with people's information beyond past purchase history and that relationship that you have with the store?
00:08:23.120 --> 00:08:37.919
Another history, again, I think our conversation in Canada on personalized pricing, though, you know, I was trying to like beat a drum or tap my bongos on this, like feels like it very sudden, like feels like it almost came out of nowhere.
00:08:38.080 --> 00:08:40.559
There's been a lot of like learning and kind of overviews.
00:08:40.639 --> 00:08:42.799
And, you know, what do you mean this is happening?
00:08:43.120 --> 00:08:46.000
Also comes out of like the credit industry.
00:08:46.399 --> 00:08:57.919
So in the early 2000s, when Canadian Tire was continuing to develop their loyalty program, which is now called Triangle Rewards, they also had, you know, a credit offering.
00:08:58.159 --> 00:09:10.480
And what they learned from, again, this is more like big data seductive era, where we, I think, questioned less what we were doing with information and were just like amazed by what data science could do.
00:09:10.639 --> 00:09:20.799
Canadian Tire found things like anyone that had ever purchased something with a skull on it in the past was like not as reliable to give credit to.
00:09:20.960 --> 00:09:31.440
And the people that were most reliable, anyone that had ever purchased the little like felt circles that you put under furniture, you know, give them leeway with credit because they're seen as being responsible.
00:09:31.600 --> 00:09:36.240
So in that way, that's sort of interesting because your data can privilege you.
00:09:36.480 --> 00:09:44.240
The personalization of pricing that I think has gotten a bit out of control as a function of our weak privacy law tends to rely on much more.
00:09:44.639 --> 00:09:46.159
What can even be done to stop this?
00:09:46.240 --> 00:09:50.240
Because the hard part I'm getting at with this is that like it seems like there's like, where's the line?
00:09:50.399 --> 00:09:51.840
Sometimes you're like, oh, this is okay.
00:09:52.000 --> 00:09:52.799
Seniors, fine.
00:09:53.039 --> 00:09:57.840
That senior getting a discount over that senior because we know that they don't buy it, they haven't bought it in a while.
00:09:58.080 --> 00:09:58.399
Fine.
00:09:58.559 --> 00:10:02.000
Knowing that they have this much in their bank account, uh, don't like that.
00:10:02.240 --> 00:10:05.200
How do we actually draw the line when you're making policy?
00:10:06.000 --> 00:10:08.240
Okay, I've got a total grab back on this.
00:10:08.399 --> 00:10:10.000
You've like asked me a special question.
00:10:10.159 --> 00:10:19.519
I love being asked about policy solutions, but I just also wanted to offer that algorithmic pricing also happens in labor markets, where we have gig platforms that are monopsinistic.
00:10:19.679 --> 00:10:32.559
There's evidence in the US that a platform for nurses to take overtime fees had gone out and purchased credit and debt load data that they were layering in, essentially proxying a desperation score.
00:10:32.639 --> 00:10:41.919
So nurses that had more student debt were offered a cheaper price for their labor based on this calculation that said they are more desperate for the work and they need it more.
00:10:42.159 --> 00:10:44.960
So we're not going to pay them as much.
00:10:45.200 --> 00:10:50.240
I think that's important in terms of like pricing for labor and also that coming for white-collar work.
00:10:50.399 --> 00:10:56.080
Whereas nursing is something that I think we see is, you know, quite standardized and fairly predictable with work.
00:10:56.240 --> 00:11:04.399
So there are other ways that data-driven pricing is hurting people and maybe not aligned with our values.
00:11:04.559 --> 00:11:05.600
What can we do?
00:11:06.159 --> 00:11:15.120
I think the provincial approach framing as a consumer protection problem, which is what the province of Manitoba is doing, is pretty savvy.
00:11:15.440 --> 00:11:33.120
Federally, if we wanted to, we could, you know, this majority government could snap their fingers and make an amendment to our federal privacy law epipata and say that profiling, discriminating uh based on personal characteristics is just a harmful practice.
00:11:33.200 --> 00:11:35.519
And we can sort of just shut it down that way.
00:11:35.759 --> 00:11:41.919
The competition bureau, if there were a case that came forward, does have some ways they could think about this.
00:11:42.080 --> 00:11:46.080
One way that I think is novel is potentially double ticketing, right?
00:11:46.240 --> 00:11:51.519
Where you're supposed to, if if there's a different two different prices, supposed to give the shorter one.
00:11:51.840 --> 00:11:54.480
Is it a form of deceptive marketing?
00:11:55.039 --> 00:12:09.840
Again, a net new piece of legislation that just says this isn't allowed doesn't get us that far because we have to agree on what kind of practice we are pushing back against, right?
00:12:10.000 --> 00:12:14.639
Without hurting those abilities to, you know, get a legitimate discount.
00:12:14.879 --> 00:12:23.279
And loyalty programs are super interesting because that bargain is more explicit in terms of I'm going to give up some privacy in exchange for points.
00:12:23.360 --> 00:12:33.360
But even that bargain, I do think, is starting to fall down as people realize that many loyalty programs are ingesting more about them than they realize.
00:12:33.519 --> 00:12:44.480
My biggest hope for Canada is that we kind of keep the conversation uh going and don't just get locked in this more blunt, like, yes, no, I support it.
00:12:44.639 --> 00:12:48.240
Or there are some instances where we're comfortable with it and we tolerate it.
00:12:48.399 --> 00:12:51.360
Therefore, it is, you know, inherently good and helpful.
00:12:51.519 --> 00:12:57.840
This is not a pay what you can kind of, we're gonna calibrate it to what's possible for you.
00:12:58.000 --> 00:13:01.039
It's fundamentally something that is extractive.
00:13:01.360 --> 00:13:16.000
And the fact that it's increasingly part of people's everyday lives and purchasing everyday essentials, I think is like particularly egregious and concerning for thinking about Canada's middle class.
00:13:16.320 --> 00:13:21.919
When it comes to like how extractive it is, like I understand that's the goal of it is let's take as much money as we can from that person.
00:13:22.080 --> 00:13:26.480
There is a flip side of it where like when I heard about this, I was like, I'm a pretty frugal person.
00:13:26.559 --> 00:13:35.200
Okay, so you know, like sometimes I'll wait on something, I'll set alerts for when something goes down on Amazon and like it'll tell me, okay, it's dropping to the price you want to pay, buy it now.
00:13:35.360 --> 00:13:41.360
So then I was thinking, oh, well, in this case, like, would it end up that it would know that, like, oh, Kara doesn't like pay this much for this?
00:13:41.440 --> 00:13:43.279
We're gonna have to give her a lower price.
00:13:43.440 --> 00:13:49.600
Um, and someone who doesn't mind as much and has maybe has more money can pay more, they'll end up paying more.
00:13:49.679 --> 00:13:53.519
Like, is that possible that it'll turn turn out that way?
00:13:54.000 --> 00:14:00.960
I'm never a fan of everyday people kind of trying to fight back on these structural failures themselves.
00:14:01.120 --> 00:14:03.679
Uh, but I do recognize there are plugins from Amazon.
00:14:03.759 --> 00:14:04.559
I've tried one.
00:14:04.720 --> 00:14:10.320
I don't know, but Kara, I'm also in a phase of my life where I'm willing to splurge now and then for convenience.
00:14:10.559 --> 00:14:21.759
So I have used Instacart in the past when I'm like desperado, need some diapers, need a few groceries to get me over the line, my husband's traveling, I'm tired, whatever.
00:14:22.080 --> 00:14:25.519
I'm fine paying for a premium.
00:14:25.679 --> 00:14:30.080
I'm fine paying, I'm fine, definitely tipping someone uh very well.
00:14:30.240 --> 00:14:34.399
And I'm fine paying weird fees that Instacart uh thinks are necessary.
00:14:34.559 --> 00:14:35.919
Let's put that aside.
00:14:36.159 --> 00:15:07.759
What I'm not fine with is being put into one of what appears to be uh from research in the US, one of four pricing cohorts and seeing a different price, first of all, a different price than is advertised in store and a different price than other people for milk, eggs, and food based on maybe where I live, what I've purchased in the past, you know, what other people in my neighborhood tend to earn, areas of the city I spend a lot of time in, um, other websites that I go to.
00:15:08.000 --> 00:15:10.399
That's where I think it's a little bit out of hand.
00:15:10.480 --> 00:15:11.840
It's unnecessary.
00:15:12.080 --> 00:15:16.399
I'm also totally fine with companies having pricing power.
00:15:16.559 --> 00:15:18.480
No one is saying that that should be taken away from them.
00:15:18.559 --> 00:15:19.600
They've always had that.
00:15:19.759 --> 00:15:22.559
Or with companies using data to set prices.
00:15:22.879 --> 00:15:28.240
Um, data based on what their competitors are doing is great for a competitive market.
00:15:28.799 --> 00:15:37.440
Maintaining a high degree of information asymmetry with individuals, hiding from them that you're even testing out different prices with them.
00:15:37.840 --> 00:15:47.120
If companies need to do this so badly, and that's what we're going to be hearing from more in Canada, because this is what the opposition is starting to organize with in the US.
00:15:47.440 --> 00:15:49.600
And I'm being very surveillant on that.
00:15:49.759 --> 00:15:50.399
It's not hard.
00:15:50.559 --> 00:16:02.639
I just go around the internet, then they should be going out of their way to brag about when and how they're doing this to us because they're hiding it in privacy policies and they're not, right?
00:16:02.799 --> 00:16:07.840
And I think there is a sense of appropriate shame.
00:16:08.559 --> 00:16:18.879
But also there's there is a novelty factor where it's been really interesting to sort of design these apps and test out these prices and earn more money and learn more at the margins.
00:16:19.120 --> 00:16:20.399
That's where we're at.
00:16:20.879 --> 00:16:21.919
How would that look like?
00:16:22.159 --> 00:16:23.279
Like transparency on it.
00:16:23.360 --> 00:16:26.879
Would they be saying, you know, your price has been adjusted based on this sort of thing?
00:16:26.960 --> 00:16:29.600
Or like I don't know how it would actually play out.
00:16:30.480 --> 00:16:31.279
That's one way.
00:16:31.440 --> 00:16:36.720
A couple of places in the US have this where they just have this baseline of disclosure.
00:16:36.960 --> 00:16:42.879
So Uber does show, you know, this the price for this ride was set by an algorithm using your data.
00:16:43.039 --> 00:16:46.639
Uh, the Washington Post also has this notification now, right?
00:16:46.799 --> 00:16:54.559
So again, special offers, um, trying to test like what will bring you in, it's it's calibrated in a particular way.
00:16:54.879 --> 00:17:03.519
PlayStation was recently, I'll use the word caught, caught running a pricing experiment where, you know, they were advertising different prices for different games.
00:17:03.679 --> 00:17:05.839
It's an interesting question, right?
00:17:05.920 --> 00:17:08.160
For like an economics 101 class.
00:17:08.400 --> 00:17:18.160
Should we charge people who play video games more often, enthusiasts, more money because they're likely to pay more because they play more and they value it more?
00:17:18.960 --> 00:17:23.279
Maybe from like an extractive, let's optimize prices situation.
00:17:23.440 --> 00:17:28.640
But in terms of a what's appropriate in the marketplace and how, you know, how do people budget?
00:17:28.799 --> 00:17:38.240
How do you equitably access these goods that is, it's the exact same thing, I think it becomes harder to rationalize charging someone more.
00:17:38.319 --> 00:18:02.480
And what fascinates me about all this is in this cost of living crisis, opposing personalized algorithmic pricing demands a kind of cross-class solidarity that we probably haven't seen in a while or kind of ever, because we sort of have to unite against companies extracting value from all of us by kind of stepping over the line.
00:18:02.640 --> 00:18:05.359
The question is, where is that line?
00:18:05.519 --> 00:18:07.279
How are we going to set it together?
00:18:07.599 --> 00:18:13.599
Yeah, I guess the algorithm that they use right now, like PlayStation will say, if you're someone who wants on day one, you're spending this amount.
00:18:13.759 --> 00:18:16.960
If you want to play it close to release day, that's that's the cost of it.
00:18:17.119 --> 00:18:23.519
And right now they'll do like, you know, if you're willing to wait a year, wait even six months, a year, two years, it'll get cheaper over time.
00:18:23.680 --> 00:18:26.079
And so, and like that's an algorithm that they're using.
00:18:26.160 --> 00:18:28.240
So I guess we've we're okay with that sort of thing.
00:18:28.319 --> 00:18:31.759
Like, you know what, price goes down over time as more people get to play the game, fine.
00:18:31.920 --> 00:18:38.079
But the difference being that, well, day one price is different for someone who like plays a lot versus someone who doesn't.
00:18:38.480 --> 00:18:38.880
Yeah.
00:18:39.119 --> 00:18:46.400
I mean, look, gas prices, we observe gas prices changing, but you and I can see that or we can track that through the gas buddy app.
00:18:46.480 --> 00:18:52.079
When we go, you and I are both at the gas station, we're paying the same amount for that.
00:18:52.240 --> 00:18:52.559
Okay.
00:18:53.039 --> 00:18:55.359
You and I are both speaking in Ontario.
00:18:55.519 --> 00:18:58.799
Hydro prices, they fluctuate based on time of day.
00:18:58.960 --> 00:19:01.519
You and I experience that universally.
00:19:01.680 --> 00:19:04.000
If we can wait to do our laundry at night.
00:19:04.079 --> 00:19:09.200
So I know I feel like I'm doing laundry all the time and it never ends, and that we could do a whole episode on that.
00:19:09.599 --> 00:19:16.559
Again, that price, there's some price dynamism there, but we all experience it equally and we can all predict it.
00:19:17.279 --> 00:19:22.559
That's not what people are opposing with the kind of personalized pricing calibrations.
00:19:22.640 --> 00:19:24.640
It's that loss of price anchor.
00:19:24.880 --> 00:19:33.519
It is, you know, eroding, kind of taking away all the consumer surplus and also creating really intense information asymmetry.
00:19:33.680 --> 00:19:35.920
You can't discipline a market that you don't know.
00:19:36.160 --> 00:19:42.480
So I do agree with Premier Ford, who a few weeks ago now said he wants the market to be free.
00:19:42.880 --> 00:19:46.720
But he said he he wouldn't oppose personalized algorithmic pricing.
00:19:47.039 --> 00:19:48.720
He thinks this is part of a free market.
00:19:49.119 --> 00:19:50.559
He thinks it's part of a free market.
00:19:50.640 --> 00:19:58.960
And I would say that's an area where we differ, where to my mind, to have a market be more free and fair, we need to have those principles of transparency.
00:19:59.119 --> 00:20:01.119
We need better privacy laws.
00:20:01.279 --> 00:20:03.200
You can't discipline a market that you don't know.
00:20:03.440 --> 00:20:06.640
He said that the federal government can use the Competition Bureau.
00:20:06.880 --> 00:20:08.319
Provinces have been talking about this.
00:20:08.400 --> 00:20:13.119
As you said, Manitoba is, I think they were the first one in Canada to say something publicly that I am aware of.
00:20:13.200 --> 00:20:15.920
But premiers are talking like, what can be done at the different levels?
00:20:16.079 --> 00:20:18.960
Is there is there one level that really needs to take ownership?
00:20:19.119 --> 00:20:19.759
Is that federal?
00:20:19.839 --> 00:20:22.640
Or are there things the provincial government or municipal governments can do?
00:20:23.039 --> 00:20:33.839
Provincial governments could do something with the like residential and tenancy acts if they wanted to sort of say we don't want commercial landlords to use pricing algorithms or we want them to disclose, et cetera.
00:20:34.000 --> 00:20:36.079
That is like one potential tool.
00:20:36.319 --> 00:20:48.400
I used to think that Manitoba was the only province thinking about it, but I'll correct myself, fact check on myself, and say that the province of Quebec had updated their privacy law fairly recently.
00:20:48.559 --> 00:20:58.319
Their provincial privacy policy basically prohibits you from discriminating against someone based on their personal characteristics.
00:20:58.720 --> 00:21:03.680
That is a way to say you cannot personalize pricing.
00:21:04.000 --> 00:21:07.759
But to my knowledge, that law has not been tested.
00:21:07.839 --> 00:21:09.759
So a case has not come forward.
00:21:10.000 --> 00:21:12.640
We could better resource these institutions.
00:21:13.200 --> 00:21:15.759
Maybe a private action case could come forward.
00:21:16.000 --> 00:21:27.519
And though I love competition law and competition policy, the competition bureau is not my first pick because it's more of a watchdog policing function of the marketplace.
00:21:27.680 --> 00:21:34.000
Whereas privacy law and consumer protection feels more like an upstream market-shaping carving conversation.
00:21:34.240 --> 00:21:37.920
Of course, downstream, that might be an interesting lever to get at this.
00:21:38.000 --> 00:21:42.559
I'd love to see a private action case come forward to the Competition Bureau testing this.
00:21:42.720 --> 00:21:44.319
I think that would be really interesting.
00:21:44.559 --> 00:21:45.759
Haven't seen this yet.
00:21:45.920 --> 00:21:51.119
I'm not directly planning one, but like on my mood board and like Vass as side schemes, yeah, that would be cool.
00:21:51.279 --> 00:21:58.000
Frances Lankin, before she became a senator, one of the incredible things she was doing was running the United Way.
00:21:58.559 --> 00:22:02.960
And she Had a hot dog stand, maybe a food truck for the day.
00:22:03.200 --> 00:22:15.039
She set it up in front of Queen's Park and she was protesting the gender wage gap and gender, pink tax, also a place we tolerate differential pricing in ways that are weird.
00:22:15.279 --> 00:22:15.599
Okay.
00:22:17.200 --> 00:22:23.599
She sold hot dogs to boys for a dollar and hot dogs to girls for 75 cents.
00:22:24.240 --> 00:22:25.200
I thought that was great.
00:22:25.599 --> 00:22:30.640
You know, I'd love to have a food truck for a day.
00:22:30.799 --> 00:22:35.359
This is about it's like you didn't ask me what all my fantasies are, but I'm unleashing them on you.
00:22:35.599 --> 00:22:39.599
I just want to look at someone be like, oh, you wear glasses, your hot dog's a dollar fifty.
00:22:39.839 --> 00:22:42.799
Oh, you've bought a hot dog before, your hot dog's two dollars.
00:22:43.119 --> 00:22:52.480
And people would be upset and annoyed, and it would seem random and discriminatory because it is random and discriminatory.
00:22:52.799 --> 00:22:56.160
And I would also calibrate the prices in like fairly fine ways.
00:22:56.240 --> 00:23:05.200
So it's like, again, it's not necessarily not the monetary value, it's the principle that I think we need to be reacting to.
00:23:05.839 --> 00:23:08.400
Sure, if you're old, I could charge you a little bit less.
00:23:09.039 --> 00:23:23.839
Like again, I'm interested in confronting people with what it feels like to be seen by a commercial actor in a way that is invasive.
00:23:24.559 --> 00:23:28.480
So sorry I took you on that side quest, but you and I both have young kids.
00:23:28.559 --> 00:23:32.640
And I the place where I was willing to spend the most money is when you're like, you're up at 4 a.m.
00:23:32.799 --> 00:23:35.519
in the first like few months, and you're they're like, you're like, I don't know what to do.
00:23:35.599 --> 00:23:39.279
You're Googling something and you're like, I'll buy anything that'll get them to go to sleep, right?
00:23:41.119 --> 00:23:41.200
Yeah.
00:23:41.440 --> 00:23:49.119
You can send up so many ridiculous things, and I could just think, oh my God, if they have surveillance pricing allowed and they know that you have a young baby at home, like you are screwed.
00:23:49.200 --> 00:23:54.480
All of your life savings are going into like, I don't know, yoga balls that you can bounce on or whatever else that they say will help with it.
00:23:54.640 --> 00:23:55.359
It's horrible.
00:23:55.680 --> 00:24:00.160
Hopefully, we see some politicians pay attention to this and listen to you on how they can fix it.
00:24:00.240 --> 00:24:02.079
Because it would really be nice to see some protection.
00:24:02.160 --> 00:24:06.319
Like we need more competition, but in as you said, a fair way would be nice.
00:24:06.400 --> 00:24:08.559
If it was transparent, that'd be lovely.
00:24:08.880 --> 00:24:09.279
Yeah.
00:24:09.519 --> 00:24:12.160
More competition, but you know, what kind and how?
00:24:12.319 --> 00:24:25.759
Hopefully, we can be as detailed and kind of attentive as the software programs and proprietary algorithms are when it comes to us uh as we approach this policy problem.
00:24:26.400 --> 00:24:27.680
Thanks so much for joining us.
00:24:27.920 --> 00:24:29.119
Really appreciate having you on.
00:24:29.359 --> 00:24:30.480
I'm a huge fan of the show.
00:24:30.559 --> 00:24:31.599
I'm a huge fan of the show.
00:24:31.759 --> 00:24:34.400
Long time, long time listener, second time caller.
00:24:35.119 --> 00:24:36.160
Third time, I think.
00:24:36.319 --> 00:24:36.640
No?
00:24:36.960 --> 00:24:38.400
Third time caller?
00:24:38.720 --> 00:24:40.559
Yeah, because you were the you were on our pilot.
00:24:40.640 --> 00:24:42.400
You were here for talking about your book, The Big Fix.
00:24:42.559 --> 00:24:44.000
I think recommend people go and watch that one.
00:24:44.240 --> 00:24:44.799
That was really good.
00:24:45.680 --> 00:24:46.319
I love that book.
00:24:46.480 --> 00:24:49.599
It's like one of those books where I was reading it and I was like, yes, I'm so angry.
00:24:49.759 --> 00:24:52.160
I can't believe this is all happening and no one cares about this.
00:24:52.240 --> 00:24:54.079
Or some people care, but not enough people care about this.
00:24:54.160 --> 00:24:56.079
Get a it's a great book.
00:24:56.160 --> 00:24:57.680
And it's it's like the perfect length.
00:24:57.759 --> 00:24:59.920
You're like, I get into it, and I here's a solution.
00:25:00.160 --> 00:25:01.200
And I loved it.
00:25:01.359 --> 00:25:02.160
Snack size.
00:25:02.240 --> 00:25:02.640
Exactly.
00:25:02.720 --> 00:25:03.119
Let's go.
00:25:03.200 --> 00:25:04.319
Let's make Canada better.
00:25:04.559 --> 00:25:06.240
Thanks so much for watching and listening.
00:25:06.319 --> 00:25:09.279
Our producer is Meredith Martin, and our editor is Sean Foreman.
00:25:09.440 --> 00:25:15.599
If you have any questions about what calms a crying baby at 4 a.m., you can send an email to missing middle podcast at gmail.com.
00:25:15.759 --> 00:25:17.119
And we'll see you next time.
00:00:00.160 --> 00:00:01.520
There's evidence in the U.S.
00:00:01.600 --> 00:00:12.960
that a platform for nurses to take overtime fees had gone out and purchased credit and debt load data that they were layering in, essentially proxying a desperation score.
00:00:13.119 --> 00:00:17.199
So nurses that had more student debt were offered a cheaper price for their labor.
00:00:17.600 --> 00:00:21.199
Demographics, hosted by Mike Moffitt and Kara Stern.
00:00:22.480 --> 00:00:25.359
The next big hit to your wallet might not be inflation.
00:00:25.519 --> 00:00:30.559
It might be that stores can figure out the maximum you're willing to pay and charge you exactly that price.
00:00:30.800 --> 00:00:35.439
It's called surveillance pricing, and politicians are starting to debate whether Canada needs to ban it.
00:00:35.600 --> 00:00:41.920
Today I'm talking to Vass Bednar, a managing director of the Canadian Shield Institute and author of the book, The Big Fix.
00:00:42.240 --> 00:00:47.439
She's someone who's been writing about the problems with surveillance pricing since before almost any of us knew to worry about it.
00:00:47.520 --> 00:00:49.840
And we're so glad to have her back on the program.
00:00:50.320 --> 00:00:51.439
Thanks for having me.
00:00:51.679 --> 00:00:55.439
There's algorithmic pricing, there's dynamic pricing, and there's surveillance pricing.
00:00:55.520 --> 00:00:57.840
Like, what is surveillance pricing that we're talking about now?
00:00:58.159 --> 00:01:05.359
Surveillance pricing is sort of layers in with that computational data-driven pricing strategy.
00:01:05.519 --> 00:01:12.079
So beyond using a variety of factors like what your competitors are pricing something at, right?
00:01:12.480 --> 00:01:26.560
Um, personalized algorithmic pricing requires excessive data use about who you are, where you live, your behaviors online, sometimes your past purchase history, and sometimes where you are and where you're going.
00:01:26.799 --> 00:01:38.480
So it's a much more invasive pricing tactic that's supercharged by data and much more common in kind of an e-commerce context where there's way more information asymmetry.
00:01:38.560 --> 00:01:43.920
It's kind of just you and a screen, whether it's your phone and a loyalty program or, you know, shopping online.
00:01:44.079 --> 00:01:46.719
I would say that's that's the bigger difference.
00:01:47.040 --> 00:01:56.400
Like the first time I remember hearing about this type of pricing, and it's only sort of similar was with Uber, where you'd have like, you know, I remember people talking about if you're on an iPhone versus an Android, you'd get different prices.
00:01:56.480 --> 00:01:59.359
Like that's like a the bare minimum version of this kind of.
00:01:59.439 --> 00:02:02.719
But then also people started getting used to dynamic pricing with Uber.
00:02:02.799 --> 00:02:04.400
Like what's different with this?
00:02:05.120 --> 00:02:09.120
So dynamic pricing suggests supply and demand constraints, right?
00:02:09.280 --> 00:02:19.360
I think uh a parallel would be airlines, where we became pretty comfortable with price fluctuations that we could monitor, but weren't necessarily tied to the individual.
00:02:19.599 --> 00:02:27.120
So if you were getting a last-minute flight, uh it was maybe more expensive, you're booking far further in advance, there were more seats available, right?
00:02:27.680 --> 00:02:37.599
Now, because of data-driven pricing, there's some evidence that suggests about a thousand different data points are used to calibrate the price of your seat on an airplane.
00:02:37.759 --> 00:02:46.000
And back to the Uber example, you know, also some evidence that if you're uh using a business credit card, Uber will upcharge you, charge you more.
00:02:46.159 --> 00:02:49.680
Again, inferring your willingness to pay and squeezing more there.
00:02:50.000 --> 00:03:28.319
Where I think there's a difference between the kind of supply and demand type of calibrations, or I'll use the euphemism pricing experiments that we've seen with Uber, is external factors like say it's raining, so there's more demand, and they'll tell you, you know, prices higher right now, or um, Blue J just won the World Series, fingers crossed, you know, there's more people looking for rides out there, something like that, versus this ride is going to cost you more vast because you tend to take this route and we know you're very likely to take this route and you're using a business card.
00:03:28.560 --> 00:03:33.039
You already get some places like different apps that'll give you discounts based on things you've purchased before.
00:03:33.199 --> 00:03:43.039
So yeah, I know I get sometimes with like shoppers or Canadian Tire or like Tim Hortons will be like, hey, you have it, I mean we know you like to buy this kind of sandwich and you haven't bought it in a long time.
00:03:43.360 --> 00:03:44.960
What's your favorite thing at Tim Hortons?
00:03:45.199 --> 00:03:50.479
Well, it sometimes it gives me breakfast sandwiches discounts because I like their breakfast sandwiches and they're like good in a pinch, you know?
00:03:50.560 --> 00:03:57.439
So I'm like, sometimes it'll be a little bit of time since I've bought a breakfast sandwich because they tend to be like, you know, I'm going out on a route, have to go drive somewhere far or something.
00:03:57.520 --> 00:03:59.520
Like that's when I get it so it doesn't happen all the time.
00:03:59.599 --> 00:04:04.000
And then I'll get a notification like, hey, you haven't bought a breakfast sandwich in a while.
00:04:04.159 --> 00:04:05.759
Would you like a discount on that?
00:04:05.840 --> 00:04:07.520
If you buy it today, you'll get this discount.
00:04:07.599 --> 00:04:12.479
And to me, I'm like, that feels like a form of surveillance pricing that already exists, right?
00:04:12.800 --> 00:04:28.879
It's a form of it, but I would say the more insidious is something like uh the Taco Bell app being able to infer when it's your payday and increasing the price of your Cordita deal versus, you know, incentivizing you to come back to Tim Hortons.
00:04:29.040 --> 00:04:40.800
You know, Tim Hortons is actually a great example because they were investigated by the Federal Privacy Commissioner and many provincial privacy commissioners for their invasive data collection in their loyalty program, right?
00:04:40.959 --> 00:04:42.959
Tracking you down to where you are.
00:04:43.120 --> 00:04:52.560
Um, an Epic study from last year found that Target would make uh televisions in their store uh more expensive the closer you were to the store.
00:04:52.639 --> 00:04:56.240
So that's kind of inferring a probability that maybe you're going to make a purchase.
00:04:56.319 --> 00:05:00.639
Again, it makes it harder for people to understand what things actually cost.
00:05:00.879 --> 00:05:08.000
We're losing price universality as Canada considers whether and when we're going to tolerate this.
00:05:08.160 --> 00:05:12.879
Because you're right, there may be some instances where we think it's appropriate, say for discounting.
00:05:13.199 --> 00:05:21.040
We need to decide if we're looking at it from a privacy perspective, consumer protections perspective, marketing, you know, deceptive marketing angle.
00:05:21.199 --> 00:05:23.040
There are different ways to kind of get at it.
00:05:23.199 --> 00:05:28.879
But to your point that it's increasingly becoming a feature of how we experience markets, like absolutely.
00:05:29.040 --> 00:05:38.879
Um, even geographic pricing, which is more blunt, but like I used to think that my secret deodorant costs the same at every single shop or struck mark, right?
00:05:39.040 --> 00:05:40.160
And that's just not the case.
00:05:40.240 --> 00:05:40.959
And that's okay.
00:05:41.120 --> 00:05:52.480
And a funny phrase that I I don't know if I'll ever say again in my life, but there was like a good Toronto Sunpiece from a few months ago looking at Big Macs costing different amounts at different McDonald's.
00:05:52.639 --> 00:05:57.040
So it's not that the pricing differential itself is egregious.
00:05:57.279 --> 00:06:07.439
It just reminds us that that pricing strategy defies what we may have as an expectation of what the norm is with pricing when actually something very different is happening.
00:06:07.680 --> 00:06:15.199
So I get that the idea is that like they're using your personal information, and that's kind of like it seems like people think that's a little bit like icky when you people are talking about this.
00:06:15.360 --> 00:06:22.480
I see like Abacus Data did some research into how people felt about it and they found 83% of Canadians wanted it banned or regulated.
00:06:22.639 --> 00:06:27.680
They found only 7% of people are convinced by the argument that this is just supply and demand at work.
00:06:27.920 --> 00:06:30.800
So I'm trying to understand like, what is it about it specifically?
00:06:30.959 --> 00:06:35.920
Is it just that they're using your data in a way that like that is it that they're using more data than they did before?
00:06:36.079 --> 00:06:38.879
Because we've had for a long time things like seniors discounts.
00:06:38.959 --> 00:06:41.920
We've had coupons offered to customers who've purchased before, as I said.
00:06:42.079 --> 00:06:46.800
We have like, there's lots of different ways that the companies have targeted people using their data for so long.
00:06:46.879 --> 00:06:49.839
And usually it's been, I think, you know, you use your loyalty card or their app.
00:06:49.920 --> 00:06:54.560
And so they're collecting the data they get from your app, from the app that they created that you're choosing to use.
00:06:54.720 --> 00:06:56.720
So what is different about it now?
00:06:57.120 --> 00:07:05.120
What's different is that the discounts, let's let's stay with discounts because you know, a senior's discount is a good example, right?
00:07:05.759 --> 00:07:08.560
Is more blunt and uniformly experienced.
00:07:08.720 --> 00:07:18.240
So, you know, if you're over a certain age, and I believe, you know, for shoppers, it has to be a certain day as well, you have equitable access to that discount.
00:07:18.399 --> 00:07:26.800
That discount is not predicated on you having a mobile device, using a mobile device when you shop, you know, sharing additional information.
00:07:27.040 --> 00:07:30.959
Couponing is a very, I'll use the the D word, dumb, right?
00:07:31.199 --> 00:07:34.079
Dumb in the sense that it's not supercharged by data.
00:07:34.240 --> 00:07:39.920
You know, you and I can still access the same deals when they come in a flyer.
00:07:40.160 --> 00:07:50.639
Where we can't see each other's universes, Cara, is in these closed app ecosystems that are more finely calibrated to who we are.
00:07:50.800 --> 00:07:55.279
Are they necessarily discriminating against us, you know, as individuals?
00:07:55.439 --> 00:07:57.600
That's where it gets tricky with privacy law.
00:07:57.759 --> 00:08:01.839
Privacy law is designed for harms that individuals experience.
00:08:02.079 --> 00:08:07.199
But more often than not, personalized pricing is about segmenting, right?
00:08:07.360 --> 00:08:10.639
Creating particular cohorts, particular profiles.
00:08:10.800 --> 00:08:13.120
And that needs to be part of our conversation.
00:08:13.439 --> 00:08:22.800
How comfortable are we with people's information beyond past purchase history and that relationship that you have with the store?
00:08:23.120 --> 00:08:37.919
Another history, again, I think our conversation in Canada on personalized pricing, though, you know, I was trying to like beat a drum or tap my bongos on this, like feels like it very sudden, like feels like it almost came out of nowhere.
00:08:38.080 --> 00:08:40.559
There's been a lot of like learning and kind of overviews.
00:08:40.639 --> 00:08:42.799
And, you know, what do you mean this is happening?
00:08:43.120 --> 00:08:46.000
Also comes out of like the credit industry.
00:08:46.399 --> 00:08:57.919
So in the early 2000s, when Canadian Tire was continuing to develop their loyalty program, which is now called Triangle Rewards, they also had, you know, a credit offering.
00:08:58.159 --> 00:09:10.480
And what they learned from, again, this is more like big data seductive era, where we, I think, questioned less what we were doing with information and were just like amazed by what data science could do.
00:09:10.639 --> 00:09:20.799
Canadian Tire found things like anyone that had ever purchased something with a skull on it in the past was like not as reliable to give credit to.
00:09:20.960 --> 00:09:31.440
And the people that were most reliable, anyone that had ever purchased the little like felt circles that you put under furniture, you know, give them leeway with credit because they're seen as being responsible.
00:09:31.600 --> 00:09:36.240
So in that way, that's sort of interesting because your data can privilege you.
00:09:36.480 --> 00:09:44.240
The personalization of pricing that I think has gotten a bit out of control as a function of our weak privacy law tends to rely on much more.
00:09:44.639 --> 00:09:46.159
What can even be done to stop this?
00:09:46.240 --> 00:09:50.240
Because the hard part I'm getting at with this is that like it seems like there's like, where's the line?
00:09:50.399 --> 00:09:51.840
Sometimes you're like, oh, this is okay.
00:09:52.000 --> 00:09:52.799
Seniors, fine.
00:09:53.039 --> 00:09:57.840
That senior getting a discount over that senior because we know that they don't buy it, they haven't bought it in a while.
00:09:58.080 --> 00:09:58.399
Fine.
00:09:58.559 --> 00:10:02.000
Knowing that they have this much in their bank account, uh, don't like that.
00:10:02.240 --> 00:10:05.200
How do we actually draw the line when you're making policy?
00:10:06.000 --> 00:10:08.240
Okay, I've got a total grab back on this.
00:10:08.399 --> 00:10:10.000
You've like asked me a special question.
00:10:10.159 --> 00:10:19.519
I love being asked about policy solutions, but I just also wanted to offer that algorithmic pricing also happens in labor markets, where we have gig platforms that are monopsinistic.
00:10:19.679 --> 00:10:32.559
There's evidence in the US that a platform for nurses to take overtime fees had gone out and purchased credit and debt load data that they were layering in, essentially proxying a desperation score.
00:10:32.639 --> 00:10:41.919
So nurses that had more student debt were offered a cheaper price for their labor based on this calculation that said they are more desperate for the work and they need it more.
00:10:42.159 --> 00:10:44.960
So we're not going to pay them as much.
00:10:45.200 --> 00:10:50.240
I think that's important in terms of like pricing for labor and also that coming for white-collar work.
00:10:50.399 --> 00:10:56.080
Whereas nursing is something that I think we see is, you know, quite standardized and fairly predictable with work.
00:10:56.240 --> 00:11:04.399
So there are other ways that data-driven pricing is hurting people and maybe not aligned with our values.
00:11:04.559 --> 00:11:05.600
What can we do?
00:11:06.159 --> 00:11:15.120
I think the provincial approach framing as a consumer protection problem, which is what the province of Manitoba is doing, is pretty savvy.
00:11:15.440 --> 00:11:33.120
Federally, if we wanted to, we could, you know, this majority government could snap their fingers and make an amendment to our federal privacy law epipata and say that profiling, discriminating uh based on personal characteristics is just a harmful practice.
00:11:33.200 --> 00:11:35.519
And we can sort of just shut it down that way.
00:11:35.759 --> 00:11:41.919
The competition bureau, if there were a case that came forward, does have some ways they could think about this.
00:11:42.080 --> 00:11:46.080
One way that I think is novel is potentially double ticketing, right?
00:11:46.240 --> 00:11:51.519
Where you're supposed to, if if there's a different two different prices, supposed to give the shorter one.
00:11:51.840 --> 00:11:54.480
Is it a form of deceptive marketing?
00:11:55.039 --> 00:12:09.840
Again, a net new piece of legislation that just says this isn't allowed doesn't get us that far because we have to agree on what kind of practice we are pushing back against, right?
00:12:10.000 --> 00:12:14.639
Without hurting those abilities to, you know, get a legitimate discount.
00:12:14.879 --> 00:12:23.279
And loyalty programs are super interesting because that bargain is more explicit in terms of I'm going to give up some privacy in exchange for points.
00:12:23.360 --> 00:12:33.360
But even that bargain, I do think, is starting to fall down as people realize that many loyalty programs are ingesting more about them than they realize.
00:12:33.519 --> 00:12:44.480
My biggest hope for Canada is that we kind of keep the conversation uh going and don't just get locked in this more blunt, like, yes, no, I support it.
00:12:44.639 --> 00:12:48.240
Or there are some instances where we're comfortable with it and we tolerate it.
00:12:48.399 --> 00:12:51.360
Therefore, it is, you know, inherently good and helpful.
00:12:51.519 --> 00:12:57.840
This is not a pay what you can kind of, we're gonna calibrate it to what's possible for you.
00:12:58.000 --> 00:13:01.039
It's fundamentally something that is extractive.
00:13:01.360 --> 00:13:16.000
And the fact that it's increasingly part of people's everyday lives and purchasing everyday essentials, I think is like particularly egregious and concerning for thinking about Canada's middle class.
00:13:16.320 --> 00:13:21.919
When it comes to like how extractive it is, like I understand that's the goal of it is let's take as much money as we can from that person.
00:13:22.080 --> 00:13:26.480
There is a flip side of it where like when I heard about this, I was like, I'm a pretty frugal person.
00:13:26.559 --> 00:13:35.200
Okay, so you know, like sometimes I'll wait on something, I'll set alerts for when something goes down on Amazon and like it'll tell me, okay, it's dropping to the price you want to pay, buy it now.
00:13:35.360 --> 00:13:41.360
So then I was thinking, oh, well, in this case, like, would it end up that it would know that, like, oh, Kara doesn't like pay this much for this?
00:13:41.440 --> 00:13:43.279
We're gonna have to give her a lower price.
00:13:43.440 --> 00:13:49.600
Um, and someone who doesn't mind as much and has maybe has more money can pay more, they'll end up paying more.
00:13:49.679 --> 00:13:53.519
Like, is that possible that it'll turn turn out that way?
00:13:54.000 --> 00:14:00.960
I'm never a fan of everyday people kind of trying to fight back on these structural failures themselves.
00:14:01.120 --> 00:14:03.679
Uh, but I do recognize there are plugins from Amazon.
00:14:03.759 --> 00:14:04.559
I've tried one.
00:14:04.720 --> 00:14:10.320
I don't know, but Kara, I'm also in a phase of my life where I'm willing to splurge now and then for convenience.
00:14:10.559 --> 00:14:21.759
So I have used Instacart in the past when I'm like desperado, need some diapers, need a few groceries to get me over the line, my husband's traveling, I'm tired, whatever.
00:14:22.080 --> 00:14:25.519
I'm fine paying for a premium.
00:14:25.679 --> 00:14:30.080
I'm fine paying, I'm fine, definitely tipping someone uh very well.
00:14:30.240 --> 00:14:34.399
And I'm fine paying weird fees that Instacart uh thinks are necessary.
00:14:34.559 --> 00:14:35.919
Let's put that aside.
00:14:36.159 --> 00:15:07.759
What I'm not fine with is being put into one of what appears to be uh from research in the US, one of four pricing cohorts and seeing a different price, first of all, a different price than is advertised in store and a different price than other people for milk, eggs, and food based on maybe where I live, what I've purchased in the past, you know, what other people in my neighborhood tend to earn, areas of the city I spend a lot of time in, um, other websites that I go to.
00:15:08.000 --> 00:15:10.399
That's where I think it's a little bit out of hand.
00:15:10.480 --> 00:15:11.840
It's unnecessary.
00:15:12.080 --> 00:15:16.399
I'm also totally fine with companies having pricing power.
00:15:16.559 --> 00:15:18.480
No one is saying that that should be taken away from them.
00:15:18.559 --> 00:15:19.600
They've always had that.
00:15:19.759 --> 00:15:22.559
Or with companies using data to set prices.
00:15:22.879 --> 00:15:28.240
Um, data based on what their competitors are doing is great for a competitive market.
00:15:28.799 --> 00:15:37.440
Maintaining a high degree of information asymmetry with individuals, hiding from them that you're even testing out different prices with them.
00:15:37.840 --> 00:15:47.120
If companies need to do this so badly, and that's what we're going to be hearing from more in Canada, because this is what the opposition is starting to organize with in the US.
00:15:47.440 --> 00:15:49.600
And I'm being very surveillant on that.
00:15:49.759 --> 00:15:50.399
It's not hard.
00:15:50.559 --> 00:16:02.639
I just go around the internet, then they should be going out of their way to brag about when and how they're doing this to us because they're hiding it in privacy policies and they're not, right?
00:16:02.799 --> 00:16:07.840
And I think there is a sense of appropriate shame.
00:16:08.559 --> 00:16:18.879
But also there's there is a novelty factor where it's been really interesting to sort of design these apps and test out these prices and earn more money and learn more at the margins.
00:16:19.120 --> 00:16:20.399
That's where we're at.
00:16:20.879 --> 00:16:21.919
How would that look like?
00:16:22.159 --> 00:16:23.279
Like transparency on it.
00:16:23.360 --> 00:16:26.879
Would they be saying, you know, your price has been adjusted based on this sort of thing?
00:16:26.960 --> 00:16:29.600
Or like I don't know how it would actually play out.
00:16:30.480 --> 00:16:31.279
That's one way.
00:16:31.440 --> 00:16:36.720
A couple of places in the US have this where they just have this baseline of disclosure.
00:16:36.960 --> 00:16:42.879
So Uber does show, you know, this the price for this ride was set by an algorithm using your data.
00:16:43.039 --> 00:16:46.639
Uh, the Washington Post also has this notification now, right?
00:16:46.799 --> 00:16:54.559
So again, special offers, um, trying to test like what will bring you in, it's it's calibrated in a particular way.
00:16:54.879 --> 00:17:03.519
PlayStation was recently, I'll use the word caught, caught running a pricing experiment where, you know, they were advertising different prices for different games.
00:17:03.679 --> 00:17:05.839
It's an interesting question, right?
00:17:05.920 --> 00:17:08.160
For like an economics 101 class.
00:17:08.400 --> 00:17:18.160
Should we charge people who play video games more often, enthusiasts, more money because they're likely to pay more because they play more and they value it more?
00:17:18.960 --> 00:17:23.279
Maybe from like an extractive, let's optimize prices situation.
00:17:23.440 --> 00:17:28.640
But in terms of a what's appropriate in the marketplace and how, you know, how do people budget?
00:17:28.799 --> 00:17:38.240
How do you equitably access these goods that is, it's the exact same thing, I think it becomes harder to rationalize charging someone more.
00:17:38.319 --> 00:18:02.480
And what fascinates me about all this is in this cost of living crisis, opposing personalized algorithmic pricing demands a kind of cross-class solidarity that we probably haven't seen in a while or kind of ever, because we sort of have to unite against companies extracting value from all of us by kind of stepping over the line.
00:18:02.640 --> 00:18:05.359
The question is, where is that line?
00:18:05.519 --> 00:18:07.279
How are we going to set it together?
00:18:07.599 --> 00:18:13.599
Yeah, I guess the algorithm that they use right now, like PlayStation will say, if you're someone who wants on day one, you're spending this amount.
00:18:13.759 --> 00:18:16.960
If you want to play it close to release day, that's that's the cost of it.
00:18:17.119 --> 00:18:23.519
And right now they'll do like, you know, if you're willing to wait a year, wait even six months, a year, two years, it'll get cheaper over time.
00:18:23.680 --> 00:18:26.079
And so, and like that's an algorithm that they're using.
00:18:26.160 --> 00:18:28.240
So I guess we've we're okay with that sort of thing.
00:18:28.319 --> 00:18:31.759
Like, you know what, price goes down over time as more people get to play the game, fine.
00:18:31.920 --> 00:18:38.079
But the difference being that, well, day one price is different for someone who like plays a lot versus someone who doesn't.
00:18:38.480 --> 00:18:38.880
Yeah.
00:18:39.119 --> 00:18:46.400
I mean, look, gas prices, we observe gas prices changing, but you and I can see that or we can track that through the gas buddy app.
00:18:46.480 --> 00:18:52.079
When we go, you and I are both at the gas station, we're paying the same amount for that.
00:18:52.240 --> 00:18:52.559
Okay.
00:18:53.039 --> 00:18:55.359
You and I are both speaking in Ontario.
00:18:55.519 --> 00:18:58.799
Hydro prices, they fluctuate based on time of day.
00:18:58.960 --> 00:19:01.519
You and I experience that universally.
00:19:01.680 --> 00:19:04.000
If we can wait to do our laundry at night.
00:19:04.079 --> 00:19:09.200
So I know I feel like I'm doing laundry all the time and it never ends, and that we could do a whole episode on that.
00:19:09.599 --> 00:19:16.559
Again, that price, there's some price dynamism there, but we all experience it equally and we can all predict it.
00:19:17.279 --> 00:19:22.559
That's not what people are opposing with the kind of personalized pricing calibrations.
00:19:22.640 --> 00:19:24.640
It's that loss of price anchor.
00:19:24.880 --> 00:19:33.519
It is, you know, eroding, kind of taking away all the consumer surplus and also creating really intense information asymmetry.
00:19:33.680 --> 00:19:35.920
You can't discipline a market that you don't know.
00:19:36.160 --> 00:19:42.480
So I do agree with Premier Ford, who a few weeks ago now said he wants the market to be free.
00:19:42.880 --> 00:19:46.720
But he said he he wouldn't oppose personalized algorithmic pricing.
00:19:47.039 --> 00:19:48.720
He thinks this is part of a free market.
00:19:49.119 --> 00:19:50.559
He thinks it's part of a free market.
00:19:50.640 --> 00:19:58.960
And I would say that's an area where we differ, where to my mind, to have a market be more free and fair, we need to have those principles of transparency.
00:19:59.119 --> 00:20:01.119
We need better privacy laws.
00:20:01.279 --> 00:20:03.200
You can't discipline a market that you don't know.
00:20:03.440 --> 00:20:06.640
He said that the federal government can use the Competition Bureau.
00:20:06.880 --> 00:20:08.319
Provinces have been talking about this.
00:20:08.400 --> 00:20:13.119
As you said, Manitoba is, I think they were the first one in Canada to say something publicly that I am aware of.
00:20:13.200 --> 00:20:15.920
But premiers are talking like, what can be done at the different levels?
00:20:16.079 --> 00:20:18.960
Is there is there one level that really needs to take ownership?
00:20:19.119 --> 00:20:19.759
Is that federal?
00:20:19.839 --> 00:20:22.640
Or are there things the provincial government or municipal governments can do?
00:20:23.039 --> 00:20:33.839
Provincial governments could do something with the like residential and tenancy acts if they wanted to sort of say we don't want commercial landlords to use pricing algorithms or we want them to disclose, et cetera.
00:20:34.000 --> 00:20:36.079
That is like one potential tool.
00:20:36.319 --> 00:20:48.400
I used to think that Manitoba was the only province thinking about it, but I'll correct myself, fact check on myself, and say that the province of Quebec had updated their privacy law fairly recently.
00:20:48.559 --> 00:20:58.319
Their provincial privacy policy basically prohibits you from discriminating against someone based on their personal characteristics.
00:20:58.720 --> 00:21:03.680
That is a way to say you cannot personalize pricing.
00:21:04.000 --> 00:21:07.759
But to my knowledge, that law has not been tested.
00:21:07.839 --> 00:21:09.759
So a case has not come forward.
00:21:10.000 --> 00:21:12.640
We could better resource these institutions.
00:21:13.200 --> 00:21:15.759
Maybe a private action case could come forward.
00:21:16.000 --> 00:21:27.519
And though I love competition law and competition policy, the competition bureau is not my first pick because it's more of a watchdog policing function of the marketplace.
00:21:27.680 --> 00:21:34.000
Whereas privacy law and consumer protection feels more like an upstream market-shaping carving conversation.
00:21:34.240 --> 00:21:37.920
Of course, downstream, that might be an interesting lever to get at this.
00:21:38.000 --> 00:21:42.559
I'd love to see a private action case come forward to the Competition Bureau testing this.
00:21:42.720 --> 00:21:44.319
I think that would be really interesting.
00:21:44.559 --> 00:21:45.759
Haven't seen this yet.
00:21:45.920 --> 00:21:51.119
I'm not directly planning one, but like on my mood board and like Vass as side schemes, yeah, that would be cool.
00:21:51.279 --> 00:21:58.000
Frances Lankin, before she became a senator, one of the incredible things she was doing was running the United Way.
00:21:58.559 --> 00:22:02.960
And she Had a hot dog stand, maybe a food truck for the day.
00:22:03.200 --> 00:22:15.039
She set it up in front of Queen's Park and she was protesting the gender wage gap and gender, pink tax, also a place we tolerate differential pricing in ways that are weird.
00:22:15.279 --> 00:22:15.599
Okay.
00:22:17.200 --> 00:22:23.599
She sold hot dogs to boys for a dollar and hot dogs to girls for 75 cents.
00:22:24.240 --> 00:22:25.200
I thought that was great.
00:22:25.599 --> 00:22:30.640
You know, I'd love to have a food truck for a day.
00:22:30.799 --> 00:22:35.359
This is about it's like you didn't ask me what all my fantasies are, but I'm unleashing them on you.
00:22:35.599 --> 00:22:39.599
I just want to look at someone be like, oh, you wear glasses, your hot dog's a dollar fifty.
00:22:39.839 --> 00:22:42.799
Oh, you've bought a hot dog before, your hot dog's two dollars.
00:22:43.119 --> 00:22:52.480
And people would be upset and annoyed, and it would seem random and discriminatory because it is random and discriminatory.
00:22:52.799 --> 00:22:56.160
And I would also calibrate the prices in like fairly fine ways.
00:22:56.240 --> 00:23:05.200
So it's like, again, it's not necessarily not the monetary value, it's the principle that I think we need to be reacting to.
00:23:05.839 --> 00:23:08.400
Sure, if you're old, I could charge you a little bit less.
00:23:09.039 --> 00:23:23.839
Like again, I'm interested in confronting people with what it feels like to be seen by a commercial actor in a way that is invasive.
00:23:24.559 --> 00:23:28.480
So sorry I took you on that side quest, but you and I both have young kids.
00:23:28.559 --> 00:23:32.640
And I the place where I was willing to spend the most money is when you're like, you're up at 4 a.m.
00:23:32.799 --> 00:23:35.519
in the first like few months, and you're they're like, you're like, I don't know what to do.
00:23:35.599 --> 00:23:39.279
You're Googling something and you're like, I'll buy anything that'll get them to go to sleep, right?
00:23:41.119 --> 00:23:41.200
Yeah.
00:23:41.440 --> 00:23:49.119
You can send up so many ridiculous things, and I could just think, oh my God, if they have surveillance pricing allowed and they know that you have a young baby at home, like you are screwed.
00:23:49.200 --> 00:23:54.480
All of your life savings are going into like, I don't know, yoga balls that you can bounce on or whatever else that they say will help with it.
00:23:54.640 --> 00:23:55.359
It's horrible.
00:23:55.680 --> 00:24:00.160
Hopefully, we see some politicians pay attention to this and listen to you on how they can fix it.
00:24:00.240 --> 00:24:02.079
Because it would really be nice to see some protection.
00:24:02.160 --> 00:24:06.319
Like we need more competition, but in as you said, a fair way would be nice.
00:24:06.400 --> 00:24:08.559
If it was transparent, that'd be lovely.
00:24:08.880 --> 00:24:09.279
Yeah.
00:24:09.519 --> 00:24:12.160
More competition, but you know, what kind and how?
00:24:12.319 --> 00:24:25.759
Hopefully, we can be as detailed and kind of attentive as the software programs and proprietary algorithms are when it comes to us uh as we approach this policy problem.
00:24:26.400 --> 00:24:27.680
Thanks so much for joining us.
00:24:27.920 --> 00:24:29.119
Really appreciate having you on.
00:24:29.359 --> 00:24:30.480
I'm a huge fan of the show.
00:24:30.559 --> 00:24:31.599
I'm a huge fan of the show.
00:24:31.759 --> 00:24:34.400
Long time, long time listener, second time caller.
00:24:35.119 --> 00:24:36.160
Third time, I think.
00:24:36.319 --> 00:24:36.640
No?
00:24:36.960 --> 00:24:38.400
Third time caller?
00:24:38.720 --> 00:24:40.559
Yeah, because you were the you were on our pilot.
00:24:40.640 --> 00:24:42.400
You were here for talking about your book, The Big Fix.
00:24:42.559 --> 00:24:44.000
I think recommend people go and watch that one.
00:24:44.240 --> 00:24:44.799
That was really good.
00:24:45.680 --> 00:24:46.319
I love that book.
00:24:46.480 --> 00:24:49.599
It's like one of those books where I was reading it and I was like, yes, I'm so angry.
00:24:49.759 --> 00:24:52.160
I can't believe this is all happening and no one cares about this.
00:24:52.240 --> 00:24:54.079
Or some people care, but not enough people care about this.
00:24:54.160 --> 00:24:56.079
Get a it's a great book.
00:24:56.160 --> 00:24:57.680
And it's it's like the perfect length.
00:24:57.759 --> 00:24:59.920
You're like, I get into it, and I here's a solution.
00:25:00.160 --> 00:25:01.200
And I loved it.
00:25:01.359 --> 00:25:02.160
Snack size.
00:25:02.240 --> 00:25:02.640
Exactly.
00:25:02.720 --> 00:25:03.119
Let's go.
00:25:03.200 --> 00:25:04.319
Let's make Canada better.
00:25:04.559 --> 00:25:06.240
Thanks so much for watching and listening.
00:25:06.319 --> 00:25:09.279
Our producer is Meredith Martin, and our editor is Sean Foreman.
00:25:09.440 --> 00:25:15.599
If you have any questions about what calms a crying baby at 4 a.m., you can send an email to missing middle podcast at gmail.com.
00:25:15.759 --> 00:25:17.119
And we'll see you next time.
