Category: Technology

Competition law and policy can help ensure open and accessible markets with fair and reasonable terms for businesses

Digital platforms are at the centre of the global economy and daily lives of consumers.

A handful of these platforms have become dominant in specific markets without facing meaningful competition. They include Amazon as a marketplace, Facebook in social networking, Google in search engines and Apple and Google in application stores.

Digital platforms rely on big data and are characterized as multisided markets with economies of scale, network effects and winner-takes-all features.

These firms offer their products for “free” on one side of the market and earn revenues from online advertising and selling user data on the other side of the market.

Download:

• Restoring competition in “winner-took-all” digital platform markets

The growing market power of these platforms raises concerns not only for consumers and smaller businesses but also for competition authorities.

Consumers not in control

Consumers can no longer control the use of their data.

Smaller businesses face unfair market conditions, where they compete with big platforms that offer services by self-preferencing their own products. It is now widely recognized that these markets cannot self-correct.

What needs to be done?

One effective response is competition law and policy that promotes open and accessible markets with fair and reasonable terms for businesses.

This goal is more pronounced in highly concentrated digital markets, where large platforms’ market power is enduring.

The most important competitive threats to monopolists are likely to come from new entrants, which are vulnerable to exclusionary conduct or anticompetitive acquisitions.

Governments should have in place relevant policies and legal frameworks to overcome different challenges of the platform economy. These include competition, consumer protection and data protection policies and legislation.

Adapt to new realities

There is a need for adapting competition law enforcement tools to new business realities by revising laws like in Germany and Austria or issuing regulations or guidelines as has been done in Kenya and Japan.

A 2017 law revision in Germany incorporated in the assessment of the market power of firms in the digital economy such criteria as direct and indirect network effects, parallel use of services from different providers and switching costs for users.

It also factored in economies of scale in connection with network effects, access by firms to data relevant for competition and innovation-driven competitive pressure.

This amendment allowed the Federal Cartel Office in Germany to consider these criteria in analyzing Facebook’s dominance in the social network market during its investigation into Facebook between March 2016 and February 2019. 

Merger control regimes should enable competition authorities to scrutinize the acquisition of start-ups by major platforms.

Merger analysis needs to incorporate the role of data in acquiring and sustaining market power and establishing entry barriers to new firms, thereby affecting future competition and innovation.

Not only free but also fair competition

It is important to ensure not only free but also fair competition. This is more so in digital markets, where smaller firms face challenges in their contractual relationship with big platforms.

Competition law provisions on unfair trade practices and abuse of superior bargaining position, as found in competition laws of Japan and the Republic of Korea, would empower competition authorities in protecting the interests of smaller firms vis-à-vis big platforms. 

Developing countries could consider this policy measure in revising their competition legislation or introduce a separate regulation concerning digital platforms’ dealings with their business users.

Such measures could facilitate entry of local small and medium-sized enterprises (SMEs) to platform markets, thereby allowing developing countries to reap the benefits of the digital economy.

This is important as SMEs are crucial to job creation and innovation.

Both the implementation of fair competition legislation and review of acquisitions of startups by dominant platforms could play an important role in maintaining an inclusive, competitive and fair business environment in the digital economy. This might eventually enhance innovation.

Apt taxation policy needed

Another critical element needed to ensure fair competition is an appropriate taxation policy. A significant proportion of the value created in the digital economy results from users who provide data.

The current international corporate tax system is not adapted to the digital economy. There is not yet a common understanding of “value creation” for taxation purposes in the digital economy.

This leads to a disconnect between where value is generated and where taxes are paid. According to the UNCTAD Digital Economy Report 2019, taxes paid abroad by Facebook represented only 2.9% of the profits it generated outside the United States in 2017.

Ideally, an international taxation system, which is agreed upon by all countries, and recognizes the main aspects of digital businesses that have significant implications for taxation, should be put in place.

By Georgios Kapetanvasileiou, Analytical Consultant at SAS

Most insurance companies depend on human expertise and business rules-based software to protect themselves from fraud. However, people move on. And the drive for digital transformation and process automation means data and scenarios change faster than you can update the rules.

Machine learning has the potential to allow insurers to move from the current state of “detect and react” to “predict and prevent.” It excels at automating the process of taking large volumes of data, analysing multiple fraud indicators in parallel – which taken individually may often be quite normal – and finding potential fraud. Generally, there are two ways to teach or train a machine learning algorithm, which depend on the available data: supervised and unsupervised learning.

Predictive modelling

In predictive modelling or supervised learning, algorithms make predictions based on a set of examples from historical data. You can present an algorithm with historical claims information and associated outcomes often called labelled data. It will attempt to identify the underlying patterns in fraudulent cases. Once the algorithm has been trained on past examples, you can use it to infer the probability of a new claim being fraudulent. AKSigorta Insurance is using advanced predictive modelling as part of its investigation process. The company has managed to increase its fraud detection rate by 66% and prevent fraud in real time.

There is a wide variety of predictive modelling algorithms to choose from, so users should take into account issues such as accuracy, interpretability, training time and ease of use. There is no single approach that works universally. Even experienced data scientists have to try different methods to find the right algorithm for a specific problem. It is, therefore, best to start simple and explore more advanced machine learning methodologies later. Decision trees, for example, are an excellent way to start exploring complex relationships within data. They are relatively easy to implement and fast to train on large volumes of data. More importantly, they are very easy to understand or interpret, and can be a good starting point for new business rules.

Other options for more accuracy

Decision trees can, however, become unstable over time. When accuracy becomes a priority, practitioners should look at other options. Support vector machines (SVMs) and neural networks are capable of learning complex class boundaries and generalise well to unseen cases. They have been extensively used for fraud detection. Tree-based algorithms, such as gradient boosting and random forests, have also become more popular in recent years. Ideally, analysts should try multiple approaches in parallel before deciding what works best.

Supervised learning is effective in identifying familiar cases of fraudulent activity but cannot uncover new patterns. Another challenge is the limited numbers of fraud examples with which to train the algorithm. Fraud is a relatively rare event, after all. The ratio between fraud and nonfraud cases can sometimes be as much as 1 to 10,000. This means that predictive algorithms tend to be overwhelmed by the sheer volume of nonfraud cases, and may miss the fraudulent ones. Labelling new data for training a model can also be time consuming and expensive.

Unsupervised learning

Unsupervised learning algorithms are trained against data with no historical labels. In other words, the algorithm is not given the answer or outcome beforehand. It is merely asked to explore the data and uncover any “interesting” structures within them. For example, given certain behavioural information, unsupervised learning algorithms can identify groups (or clusters) of customer transactions that appear similar. Anything that appears different or rare could be flagged as an anomaly (or an outlier) for further investigation.

Unsupervised learning methods can, therefore, identify both existing and new types of fraud. They are not restricted to predefined labels, so can quickly adapt to new and emerging patterns of dishonest behaviour. For example, a New Zealand health insurer used unsupervised learning methods to identify cases where practitioners were deliberately overcharging patients for a particular procedure or providing unnecessary treatment for certain diagnoses.

Unsupervised anomaly detection methods include univariate outlier analysis or clustering-based methods such as k-means. However, the recent move towards digitalisation means more data, at higher volumes, from a wider range of data sources. New algorithms, such as Support Vector Data Description, Isolation Forest or Autoencoders, have been introduced to address this. These may be a more efficient way of detecting anomalies and allow for faster reaction to new fraud.

Social network analysis

These methods are useful for identifying opportunistic fraud. However, many fraudsters today operate as part of professional, organised rings. Activity may include staged motor accidents to collect on premiums, ghost brokering, or collusion between patients and health practitioners to inflate claim amounts. These career fraudsters can repeatedly disguise their identities and evolve their way of operating over time.

Social network analysis is a tool for analysing and visually representing relationships between known entities. Examples of shared entities could be different applicants using the same telephone number or IP address, or a motor accident involving multiple people. Social network methods can automate the process of drawing connections from disparate data sources and visually representing them as a network. This significantly reduces the investigation time – in one case, from 10 days to just two hours. In the UK, a large P&C insurer made £7 million savings per annum by uncovering groups of collaborating fraudsters using network analytics.

A hybrid approach

No single technique, however, is capable of systematically identifying all complex fraud schemes. Instead, insurers need to combine sophisticated business rules and advanced machine learning approaches. This will allow them to cast the net wide, but improve accuracy and reduce false positives, making fraud detection more efficient.