Mr. Chair, Thank you for the opportunity to appear today.
I am Anthony Durocher, Deputy Commissioner of Monopolistic Practices at the Competition Bureau and I am joined by my colleague Alexa Gendron-O'Donnell, Associate Deputy Commissioner of the Bureau's Economic Analysis Directorate.
The Bureau is an independent law enforcement agency that ensures Canadian businesses and consumers prosper in a competitive and innovative marketplace.
The Bureau administers and enforces Canada's Competition Act, which involves investigating and addressing abuses of market power, anti-competitive mergers, price-fixing and deceptive marketing practices.
Competition law enforcement requires more than theory. Evidence-based enforcement is at the heart of what the Bureau does and this requires that our decisions be based on credible evidence that can withstand judicial scrutiny.
It is also important to recognize that we are enforcers, not adjudicators. The Competition Act requires us to meet several thresholds and standards, such as proving there has been a significant harm to competition.
Regardless of if we want to bring a case forward, we are guided by the decisions of the Competition Tribunal and courts.
It is difficult to turn on the television or read the news without seeing the increasing role of data in our economy. The power that data now represents, and the control that digital platforms have over it, deserves careful consideration. The bureau recognizes its important role in this area and strives to be a leader through both its enforcement work and its policy work, both of which we plan to discuss today.
We understand that this committee is particularly focused on privacy. It is important for me to say from the outset that safeguarding privacy is not an explicit goal under the Competition Act, so our role is limited in this regard. However, there are two ways privacy can be relevant to our work. First, if companies compete to attract users by offering privacy protection, then this dimension of competition can be a relevant factor in reviewing anti-competitive activity. Second, if companies mislead consumers about whether and how their data will be used, this may also raise concerns under the Competition Act.
There are many obvious benefits associated with the collection and analysis of data, particularly for driving innovation, but there are also risks. The bureau has a mandate to safeguard competition in the digital economy, and we continue to prioritize this work. However, it is important to acknowledge that competition law has its limits. It is not a cure-all for the broader threats that data and data-driven platforms may pose for society, such as breaches of privacy, election tampering or manipulation of public opinion. These risks go beyond our legal mandate. Nevertheless, we are happy to bring our competition expertise to bear on this important discussion, as these issues are cross-cutting and will benefit from collaboration across government to protect Canadians.
A little over a year ago, the bureau published a comprehensive white paper entitled, “Big data and Innovation: Implications for Competition Policy in Canada". The purpose of this paper was to engage with stakeholders by prompting a discussion on how the emergence of big data should affect competition law enforcement.
Following an extensive consultation, the bureau found that there's no need for hasty moves in this area. The current framework is up to the task, but our tools must evolve to deal with the complex issues arising from digital platforms, such as those that monetize user data through advertising by offering free services to consumers.
A recurring concern we hear about is the large and growing size of some tech firms, but big doesn't necessarily mean bad. Becoming big is the reward a firm could get for successfully introducing an innovative product. We should not punish this success. Only when we find evidence that a big firm is engaging in harmful anti-competitive conduct should we intervene.
It is important to find the right balance between preventing any competitive behaviour that harms Canadian consumers and avoiding undue over-enforcment and the inadvertent harm this may cause to innovation and the economy. Some of the issues that we have heard about relating to the digital economy and our monitoring include firms buying emerging competitors or excluding disruptive ones; firms that may use artificial intelligence or algorithms to collude and fix prices; and firms misleading consumers about whether and how their data will be used. If we find evidence that any of these practices violate the Competition Act, the bureau will act to protect Canadians.
We have already conducted several notable investigations in the digital economy, including against Google over an alleged abuse of market power related to its search engine, and the Toronto Real Estate Board, or TREB, over its real estate data.
Our case against TREB is a great example. We were able to stop TREB from withholding its real estate data from agents who wanted to offer innovative online services to homebuyers and home sellers. This case exemplifies how we are ensuring that Canadian consumers benefit from the innovation happening in the digital economy.
We welcome the opportunity to discuss the bureau's white paper, as well as these recent cases, in greater detail during the question and answer period.
The digital economy is a top priority for the Bureau. We will continue to monitor the online marketplace, including the conduct of large tech firms.
We will also continue to work closely with our domestic and international partners and carefully review the actions taken by our international counterparts. However, laws and competitive dynamics may differ significantly between countries, and we must remain mindful of that.
The Bureau also encourages all Canadians to reach out to us if they have any evidence of violations of the Competition Act.
Before fielding your questions, I would note that the law requires the Bureau to conduct investigations in private and keep confidential the information we have. This obligation may prevent us from discussing some past or current investigations.
We appreciate the opportunity to appear before you to discuss our work and look forward to your questions.
Thank you, and good morning, Mr. Chair and members of the committee.
My name is Dan Rogers, and I'm the deputy chief of foreign signals intelligence at the Communications Security Establishment. I am responsible for CSE's foreign signals intelligence program. I'm joined today by my colleague André Boucher, the assistant deputy minister of operations at the Canadian Centre for Cybersecurity.
The Canadian Centre for Cybersecurity, which is a part of CSE, is Canada's national authority on cybersecurity and cyber-threat response. It's a pleasure to appear before you as you continue your study.
With regard to the incident involving Cambridge Analytica and Facebook, CSE does not have a mandate to regulate social media, nor is it a law enforcement agency. We have no oversight role with respect to these companies. We do, however, have a role in identifying and helping protect against cyber-threats to Canada's democratic process. Therefore, I would like to focus my remarks on these threats and how they can be mitigated through good cyber and physical security.
I also hope to leave you with a better sense of what CSE does and how we have changed as an organization since CSE officials last appeared before this committee in 2017.
CSE is Canada's national signals intelligence agency for foreign intelligence and the technical authority for cybersecurity and information assurance. I would like to emphasize that CSE only directs its signals intelligence activities at foreign communications. CSE is prohibited by law from directing its activities at Canadians anywhere or at anyone in Canada.
CSE operates at the cutting edge of today's threat environment. Whether providing intelligence on foreign-based terrorism or threats to Canadians abroad, or defending against cyber-attacks, CSE helps to ensure Canada's prosperity, security and stability.
More recently, CSE was asked to assist the with her mandate to lead the Government of Canada's efforts to defend the Canadian electoral process. Specifically, the mandate letter for the Minister of Democratic Institutions directed that she ask CSE to analyze risks to Canada's political and electoral activities from hackers and release this assessment publicly, and to offer advice to Canada's political parties and Elections Canada on best practices when it comes to cybersecurity.
In response, we released a report on cyber-threats to Canada's democratic process in June 2017. While the report is unclassified, key judgments in the assessment rely on multiple sources including classified information from CSE's unique cybersecurity and foreign intelligence expertise. CSE examined cyber-threat activity against democratic processes across Canada at the federal, provincial and territorial, and municipal levels and around the world. The report examined the types of threat actors involved, the targets they are likely to select and the methods they may use to target their victims.
CSE assessed that in the 2015 Canadian federal election, Canada's democratic process was targeted by low-sophistication cyber-threats likely perpetrated by hacktivists and cyber criminals. These activities had no effect on the results of the election and no impact on the privacy of Canadians. CSE has assessed that, at the federal level, political parties and politicians and traditional and social media are more vulnerable to cyber-threats than election activities themselves.
Consistent with the increasing cyber-threat activity against democratic processes worldwide, we expect to see multiple hacktivist groups deploying cyber capabilities in an attempt to influence the democratic process during the 2019 federal election. These will likely be low-sophistication activities, but will be well planned and will target more than one aspect of the democratic process.
CSE has been asked to continue this analysis and expects to release an update to the 2017 report.
While offering mitigation advice was outside the scope of the threat report, to respond to 's second request of CSE, we have held briefings with political parties, provincial and territorial clerks, and Elections Canada to offer best practices when it comes to cybersecurity.
Our key message in all of these briefings is that, while system safeguards are expected to curtail most suspected malicious activity, we cannot rely solely on technical safeguards. Users must also be diligent and have good cybersecurity habits in order to stop the threats of today and to stay ahead of the threats of tomorrow.
CSE has made available on its website several documents, the "Top 10 IT Security Actions", "Cyber Hygiene", "Mobile Security" for IT enterprise, and other resources with user best practices. We'd be happy to speak to any of these in greater detail during the questions and answers.
Cybersecurity is a team sport. We'll continue to work with Elections Canada to ensure that the electoral process is secure and remains a trusted aspect of our democratic process.
CSE will work with and other stakeholders, if requested, to advance the goal of protecting Canada's democratic institutions and electoral processes from cyber-threats.
On October 1, the announced the launch of the cyber centre, Canada's national authority on cybersecurity and on cyber-threat response. The cyber centre, housed at CSE, brings together cyber expertise from Public Safety Canada, Shared Services Canada and CSE all under one roof. A unified government source of expert advice and guidance for the private sector, critical infrastructure owners and operators and all Canadians, the cyber centre will help ensure a safe and secure cyberspace.
This newly established centre will also enable better coordination of efforts in the protection of Canada's democratic institutions from cyber-threats. This includes the period preceding the 2019 federal election.
Again, thank you for inviting us here today. We look forward to answering your questions.
Good day, Mr. Chairman and committee members, and thank you very much for the invitation to be here.
As the managing director of the Canadian economic analysis department of the Bank of Canada, I'm happy to present our views and observations regarding the declining competition in advanced economies. I'll also address the implications for both competition and the longer-term dynamism of the economy related to the emergence of large tech firms as well as the growing importance of big data.
Understanding the impact of digitalization on the Canadian economy is crucial as we seek to achieve our objective of low, stable and predictable inflation. The Bank of Canada does not have a regulatory role with respect to the privacy of citizens' data so I trust you will understand that I will not be able to address the privacy implications of these issues.
The Canadian economy is digitalizing rapidly. Digital disruption is expected to be positive for economic progress overall. New firms are being created and existing ones are being transformed as new technologies change the way businesses operate. For consumers, digitalization means that households can purchase a seemingly ever-widening range of goods and services 24-7 from around Canada and from around the world.
Digitalization will contribute to higher productivity, and hence higher living standards, in the coming years and decades.
There is a lot of concern about the rise of the robots and how they could take away people's jobs. Naturally, we tend to focus on these initial effects. But we also need to be mindful that it takes a long time to fully replace a worker with a robot.
Still, there is no doubt there will be disruption for some, and there is time for society to adjust. People whose jobs are affected will need support. Job training and a strong safety net are key.
We must also remember that digitalization is creating new kinds of jobs and will create some that haven't even been imagined yet. These new jobs will help the economy grow. New jobs mean new incomes, which will be spent not just in the digital economy but across the whole economy, with benefits for workers in traditional jobs too.
One of the driving technologies of digitalization is the application of artificial intelligence and machine learning in conjunction with big data to a wide range of business applications. AI and ML increase firms' productivity in three major ways. First, AI and ML help companies make better products and improve their customers' experiences. Second, they help develop products and services more efficiently and more quickly. Finally, they help firms reach new markets and customers.
There are practically countless examples of such applications.
They include farmers using GPS autopilots to drive their tractors and optimize fertilizer and pesticide use; robots working on factory floors and in warehouses, "driving" forklifts to move goods and digitally track them from supplier to retailer; Al offering up suggestions for products or services you may wish to buy; and having chatbots and robo-advisers standing ready to answer your questions when you visit websites.
By implementing AI and ML with big data, firms can gain a competitive advantage, ultimately through offering a better product or service at a lower price. One of the features of AI and ML, big data and network effects is that there are often significant benefits in being a first mover. In fact, market concentration happens quite naturally in industries with prominent network effects and other scale economies.
In the current environment, this dynamic can lead to the creation of superstar firms. These firms tend to have fewer employees than conventional companies and they often earn impressive monopoly profits.
What is new is that the winner-takes-all effect is magnified in the digital economy because user data has potentially become another source of monopoly power. Data from a large network creates a formidable barrier to entry in some cases. Another barrier to entry can come from firms using the position as gatekeepers of crucial online services to impede their competitors and thwart innovation. In this context, we believe competition policy can be modernized appropriately to help ensure that benefits of digitalization are fully realized.
What do we know? What evidence do we have on the issue of market concentration, markups and prices?
In recent years, economists have paid considerable attention to the secular rise of market concentration in advanced economies. In particular, models have been developed that tie this rise to digitalization. Specifically, these firms are able to capture an increasingly large share of the market because of technological advances, such as AI and ML with big data, thereby increasing concentration. They also have a high share of profits, which can lead to a fall in the labour share of income.
Overall, most industries have seen an increase in their concentration over the last 15 years. Although the evidence is not conclusive, a broad increase of industry concentration across countries suggests that technological change, that is, digitalization, rather than country-specific factors, is perhaps the main driver.
One concern in an environment dominated by superstar firms is that those firms have more power when setting prices, which could lead to an increase in prices. That's why economists have also been looking at the secular rise in market power of firms as measured by markups. For example, researchers documented a rise in average markups in the U.S. from 1980 to 2014. They also found that global markets have risen as well. This increase is also observed in Canada. For Canada, they document a very similar overall trend to the U.S., a finding confirmed by the IMF. This suggests that market power's been rising in many countries over the past few decades.
The next question is whether digitalization has affected consumer prices. This is often referred to as the “Amazon effect”, where competition from digital retailers results in lower prices. It may appear inconsistent that digitalization can lead to both higher markups and lower prices. However, it is simply that the benefits of technology partly go to the customer in the form of lower prices, but also to the firm in the form of higher markup over lower cost.
While the direct evidence of the impact of digitalization on inflation is mixed, it does tend to point to downward pressure overall. In a research paper published last year, bank staff found that the direct evidence pointed to a small negative impact of digitalization on inflation. That is, digitalization was weighing on price increases rather than feeding them. Evidence using online prices data, such as the Billion Prices Project, is mixed. Some find that online prices tend to behave similarly to bricks-and-mortar store prices, while others find big effects on inflation year over year on the downward side. Most of us are familiar with why this might be happening—our ability to check competitor's prices using our smart phones before we head to the checkout.
Finally, when using the framework upon which the bank's main economic models are built to assess the channels through which digitalization may affect inflation, we find that most developments associated with digitalization would put downward pressure on inflation.
Overall, the impact of digitalization on market concentration, and hence competition, remains an open question. The bank will continue to examine the impact of digitalization on the Canadian economy as we pursue our objective of promoting the economic and financial welfare of Canada.
Thank you once again for the invitation to appear.
I appreciate that, Chair. Thank you very much.
Thank you to all for appearing this morning.
I recognize that the Competition Bureau and the bank have only peripheral suggestions that might be applied in the recommendations we make to government on completion of this report on the digital vulnerability of the Canadian electoral system, or threats to the Canadian electoral system. Therefore, I'd like to direct all of my time to the CSE witnesses today.
As a politician, I participate in social media almost entirely for political benefit, and there are significant benefits to using Facebook, Instagram and other social media—Twitter.
This week the digital threat was brought home to me when my Instagram account was seized by someone from outside of the country. My Facebook account was hacked and took some time to be recovered.
It brought to mind the so-called Beyoncé trick, which previous witnesses have spoken to before the committee. In the United States, in the last federal election, a Facebook fan page was created paying tribute to Beyoncé, which accumulated millions of followers. Then, in the final days of the election campaign—and this was set up, we understand, by Russian players at one level or another—messaging went out which, in the end it has been concluded, was aimed at discouraging black voters from voting in that campaign, or in some of the campaigns.
We asked one of our previous witnesses, Dr. Ben Scott, about how Canadians might protect themselves from the sort of Trojan Horse social media time bomb that was set to go off in the decision-making period in an election campaign. He suggested that agencies like the CSE would be playing what he called "red teaming", Cold War game playing, in trying to anticipate threats, how one would respond to threats, how one would see this as a fraudulent attempt to interfere with the election process. He essentially said that security agencies have an ability—and I recognize you have no authority over social media—and certainly American security agencies have an ability, to see foreign intervention or foreign players in the social media sphere.
I'm wondering if you could address what the CSE is doing in that area.
Certainly. Thank you for the question.
I'll answer from the foreign intelligence perspective, which is in my domain. Then I'll invite André to talk to some of the guidance we provide to Canadians to deal with that sort of issue.
As a foreign intelligence-mandated organization, we do track targets on the global information infrastructure, which includes social media. If we have intelligence priorities from the government, which would include things like looking at foreign nations that would have an interest in disrupting our electoral systems, we would look anywhere on the global information infrastructure to identify what those activities, capabilities and intentions of those foreign states are.
I can't in this setting speak to the details of how we do that, but I can definitely say that it is an active part of our role to observe that activity to the extent that we can. Then we roll that up into foreign intelligence products to provide to our partners in the government, and to André's team, which form the basis of some of their cybersecurity advice and work with those who may be affected by these activities.
André, do you wish to speak to that?
Thank you very much for the question. It's a good question.
What we're seeing right now in the Canadian economy is there's a lot of activity in the digital economy. While we don't have an explicit measure of the digital economy, the things that we do look at show there's very robust growth. Taking one measure, for example, if you look at GDP by industry, there's a category called computer systems design and related services. It's been growing more than 7% a year for the last five years. In value-added space, it's as big as autos and aerospace combined. So there's very rapid growth, a lot going on.
If you look more anecdotally in centres like Toronto, Montreal, the Waterloo corridor, Edmonton and other places, there's a lot of digital activity, a lot of investment going on in IP and research and development. Also, this has attracted the interest of large players to bring FDI into Canada, with them locating here, to benefit from the talent pool we have in terms of big data, AI and ML and a lot of artificial intelligence. By one metric, Canada has the third largest number of researchers in AI and ML, so we're well positioned to take advantage of this, and we see this as a really strong driver of the Canadian economy right now.
This goes to the security end of our organization.
Maybe piling on a little bit on what Dan has said about the team and the size of the team, the power of 2,500 is really the power of 2,500 plus our Five Eyes colleagues, and that helps you scale, when you're facing foreign threats of different kinds, of all kinds. We work very closely together. We share advice and guidance, and we share the perspectives on what the threats are, the methods that they use, and what to do about it.
On the specifics of countries or technologies, of course there will be moments when, within the Five Eyes, we might have different views and different opinions, but that's to be expected, because we're from different nations. We have different sovereign rights and different organizations, and we have different systems, in fact, and a different presence already. Where you see perhaps some discussions between us, our situation is different. We take different measures, but at the end of the day, around the table, when we sit down and look at a similar scenario, we always come to the same conclusion. Where there might be some differences on the surface, I assure you that, where it matters at the deep end, we are very closely aligned, and we have been for 70 years.
I have a few questions.
I want to start with you, Mr. Rogers, and it's a really simple question.
Political parties get their backs up when we talk about potentially bringing them under a regulatory framework with respect to privacy or data protection practices. I'm glad you've given advice to all Canadian political parties.
My question is not so much from a privacy protection standpoint, but from ensuring best data management practices. I attended a parliamentary round table of representatives in Washington. was there as well. A number of representatives say that two-factor authentication is necessary in today's day and age, and if it's not a rule for political parties, that's a huge problem.
I read your June 2017 report, and you say you're not worried about Elections Canada, but you're worried about the vulnerability of political parties. When I read that the Democratic Party and the Republican Party were both hacked and that there was selective distribution of the material from that hacking, it's political parties that need to up their game on data management practices.
Shouldn't they be regulated?
Mr. Durocher, in your report, the white paper, it indicates that the purpose of the act and the review, when you look at big data, the goal really is to ensure an innovative, efficient and prosperous economy. There's a conversation about substitutability. It's not just the price; you could talk about quality.
We had some folks here talking about the worries about antitrust. It was on this point, quite apart from pricing. People are put in a position—Ms. Vandenbeld got at this—where you're forced to deal with a monopoly.
We have the Bank of Canada suggesting that the five biggest global tech companies have a market cap of $3.5 trillion U.S. There are certain companies we have to deal with in our day-to-day lives. There's no choice to be made. We have to give up what we give up to access the service. It's not necessarily a price consideration, but there's a quality consideration. Part of that quality of service is the data and the privacy that I potentially give up.
That was not a big part of the conversation when you were looking at big data in the paper. I wonder if you could speak to that.
Yes. That's an excellent question as well.
Data portability of the regulations that we're seeing through the GDPR is the most noteworthy, I think, from a competition perspective. In theory, it can be pro-competitive. It can empower consumers to take their data from one platform to another. Obviously the devil is in the details as to how that's operationalized, but certainly it's something we're taking note of.
We're seeing it in the Canadian banking industry. For instance, the underlying premise of the open banking initiative is enabling people to move their data from one service provider to another. From a competitive perspective it's certainly very interesting. It's something we're monitoring very closely.
By the same token, we have to watch how it's operationalized. From a competitive perspective, when we look at regulations relating to privacy, another competition consideration involves the cost of compliance. It must not be so high as to effectively entrench large players and make it more difficult for smaller players to compete.
Mr. Santor, this is my last question.
Ms. Wilkins, senior deputy governor, remarked in February of this year that data has become another source of monopoly power. She indicated two concerns. One, it might impact innovation in a negative way, and two, it may well return to monopoly pricing in the long run.
There may be some other concerns. We've certainly heard some other potential antitrust concerns from other witnesses here.
She threw out some potential solutions that other people have been talking about regarding how we regulate ownership and the sharing of information, and maybe treating tech platforms as utilities.
When we had the CRTC here on net neutrality, they talked about a section that says companies can't unjustly discriminate or give undue reasonable preference towards themselves—or any person, but including towards themselves—or subject any person to unreasonable disadvantage. That's to get at equal treatment.
Should we regulate Facebook, Google, Apple, Amazon and Microsoft? Shouldn't we treat them the same as Rogers and Bell?