Big data is one of the most important drivers of the digital economy and the fourth industrial revolution. The rational use of big data can greatly promote economic development, but if not used properly, it can also cause a lot of problems.

If big data is applied in a proper way, it can help businesses expand scale, increase efficiency, improve experiences, reduce costs and control risks. These five positive effects may change some laws of current economic operations.

For example, with the support of big data analysis, the boundary of the 80/20 rule may become less prominent. The 80/20 rule, also known as the Pareto Principle, is a theory that finds that, for any given situation, roughly 80 percent of consequences come from 20 percent of causes.

In other words, it is entirely possible for financial institutions to offer financial services to 80 percent of mass customers under the premise of low costs and high efficiency through the use of big data.

Thanks to the application of big data and large-scale technology platforms, the common phenomenon of diminishing returns to scale may also change. The marginal cost for various businesses will decline, or even reach zero. All these may bring revolutionary changes to the economy and finance.

There is a concept of production function in economics, which is determined by several production factors, including land, capital and labor. If data are added to this production factor, it may change the future marginal return of each factor, which may eventually lead to changes in some basic characteristics of the production function.

For less developed countries, it might be a new opportunity, as it usually takes quite some time for factors, including human capital, to accumulate. But with big data, it is entirely possible for a country to leapfrog, if it can collect and analyze data well.

However, there are also differences between data and traditional production factors. In terms of the definition of rights, data factors are not scarce, so their use is not exclusive. This might become a benefit, but there are also objective difficulties in terms of transactions and pricing.

What's more, land, labor and capital can be allocated. Although land cannot be moved, labor and capital can be allocated to form new production units and launch new production. For data, some can be configured but others cannot.

To sum things up, data governance has a series of problems to be solved, including confirmation of rights, transactions, pricing and usage.

To collect and analyze data, it is also important to maintain the balance between protecting rights and how to give full play to value. Good governance can protect rights and interests as well as realize sharing, reasonable pricing and scientific allocation, thereby creating the greatest economic benefits.

However, there might be more difficulty in reality. I suggest governments should adopt a pragmatic strategy to achieve a balance between protecting privacy, data security and exerting value.

First, strike a balance between safety and innovation. Data protection, or data security in a broad sense, includes national security and personal privacy protections. In terms of data protection, Europe has done the best job but it is also because Europe doesn't have a particularly successful platform economy or digital economy companies.

China's digital economy, however, has helped birth a number of industries, but also led to some problems. The United States is in the middle, but it is hard to judge whether the US is in the best condition.

Therefore, data protection must be strengthened, but it should not be overprotected, like that of Europe. It is necessary to reasonably achieve a balance and make a reasonable distinction between different types of data.

Some data involving private rights should be controlled more strictly, while other data is expected to be controlled appropriately, because the ultimate goal is to give full play to the value of big data.

Second, the need to maintain a balance between sharing and efficiency. Big data is continuous in time and space rather than a single piece of stand-alone datum. Data that are really meaningful should be integrated and analyzed to break information silos and achieve data sharing.

But to share data, there are two bottlenecks. One of the biggest difficulties for many financial institutions in serving micro, small and medium-sized enterprises is the asymmetry of information and lack of sufficient data to judge the credit status of such enterprises.

Provinces like Guangzhou, Zhejiang and Shandong have made good attempts to integrate data to support financial services. They have established a comprehensive financial services platform that shares information regarding social security, taxation, water and electricity.

To solve the problem, more top-level policies are needed to ensure the safety of sharing existing data so that we can achieve stability in the provision of financial services.

The second bottleneck is that big data credit investigations should deal with data iteration and benefit distribution. There are two big data credit investigation companies in the country, but they are facing difficulties.

We have worked with the International Monetary Fund and the Bank for International Settlements to study whether big data can be used in credit risk assessment. The answer is yes, but there are certain problems. It is limited to small and short-term loans and it is more difficult to increase the amount. If the amount increases, this set of evaluation methods may not be equally effective.

In short, financial data-sharing should adapt to different data conditions. For some data suitable for sharing, we create opportunities to maximize benefits. Vice versa, if data are not suitable for sharing, then we should find mechanisms to maximize its economic and social benefits as much as possible.

The global digital economy can be roughly divided into three markets, the United States, China and the rest of the world. Among the top 20 unicorn digital technology companies in the world, China accounts for half of them, which is quite remarkable.

However, if we do an in-depth analysis, we will find that China's advantages are the demographic dividend, the separation dividend and the innovation opportunities brought about by relatively inadequate data protection. Whether these can continue to support the development of the country's digital economy is questionable.

If the current development is incorporated into the global structure, it is necessary to consider that it will be technology rather than scale that will lead the country's future development. If we want to compete in the future, we need to know how to support the next innovations.

There is another interesting observation. The general public is often fed up with large companies, especially when income distribution continues to deteriorate.

The general direction of big data governance is to support innovation and standardized behavior, but the general direction of common prosperity needs to be particularly highlighted. The result of innovation cannot just create thousands of billionaires.

So, data governance should be considered in an innovative way. It might not be very appropriate to simply follow the traditional governance methods or learn from the European or US methods.

Taking antimonopoly efforts as an example, the traditional judgment standards of antimonopoly efforts involve market share and pricing. But now, this idea may need to be reconsidered as the basic feature of the platform economy is scale.

The fundamental issue behind it is in fact the ability to compete, that is, whether there is market competition. To judge whether there is a monopoly, it might be more reliable to look at competitiveness rather than market share.

The writer is director of the Institute of Digital Finance at Peking University in Beijing.

The views don't necessarily reflect those of China Daily.