Online content filtering exists everywhere, with priorities and definitions differing across countries. In the obscenity case of Jacobellis v. Ohio (1964), Justice Potter Stewart wrote in his short concurring opinion that "hard-core pornography" was hard to define, but that "I know it when I see it".

Sometimes, copyright infringements, rumors and defamation need filtering under certain circumstances. However, there could be "unwanted" consequences to such a method of Internet management, irrespective of time and place. There are two types of error: Type 1 errors are unwanted information filtered in, and Type 2 errors are wanted information filtered out.

The choice of an Internet management method usually depends on the tolerance level of the above-mentioned errors, the cost of filtering implementation, and the speed and extent of the dissemination of information. Not only governments are choosing to filter content, but also Internet platforms, service providers, and families are doing so. 

Once a policy is in place, the move is usually to strengthen the filtering policy over time, even when social benefits exceed social costs because the one who removes it will be accountable for possible future problems.

This is why it is important to engage different stakeholders in the decision-making process. It is unimaginable how a desirable filtering policy could be formulated without delegations to the local government as well as market.

China is big enough to allow policy experiments to be conducted in different sectors, groups and cities. This is in line with other decision-making processes such as housing and traffic policies. In such cases, the central government has allowed or encouraged local governments to implement the policies based on their situations, and then selected the best model to promulgate nationwide.

For instance, the household responsibility system, which allowed villagers to keep excess income after making contributions to the government, was an innovation made at the village level 40 years ago. It was applied to the whole country about five years later.

There is no reason why Internet management policies could not follow such an example; especially that the similar experimentation approach in other policy areas has been well received. Over the past decade, we have seen an evolution of innovative filtering ideas which were added to China’s Internet regulations. 

Taking a website down has been changed to blocking pages with undesirable content. During some key international events, some participants are allowed to access the Internet in a whole other different way.

On the other hand, Internet platforms could be guided to increase social welfare in a more cost-effective way. Different versions of safe harbor policies exist in different countries for Internet providers. The idea is to not penalize Internet providers too heavily as it is not possible to filter undesirable content every time. 

Clear guidelines and contingencies should be made and jointly designed with the government. For example, some countries require Internet platforms to remove undesirable content within a short time after receiving complaints.

It is important, however, to maintain a high alert level and keep routine checks on Internet platforms to make sure that proper procedures could be effectively executed should there be an emergency, such as a national security breach. It is also important to start thinking about pricing mechanisms to set the right penalty fee for violations or access price for selected content.

For example, instead of a "cold turkey" policy to set up a blacklist of websites or keywords for university libraries, the government could encourage the librarians to jointly maintain such lists locally in their academic institutions. Librarians could choose to accept a penalty if they think the official lists are too strict and the violations seem very unlikely. 

In another example, a researcher could opt for a paid service which enables him to authorize a licensed agent to monitor his browsing activities when he searches for scientific materials on the Internet. These are but examples of open communication channels in the academia which inspire market solutions to information access.

Building up a socially-engaged filtering innovation mechanism, especially through experimentation, is  important. With the speed of Internet, "unwanted" innovation could have dramatic effects that make previous efforts obsolete. For example, Wi-Fi networks provided by foreign satellites could render most filtering policies in each country futile. Internet is not a separate space. There should be more communication between different stakeholders for better regulatory policies.

(Benjamin Chiao is a professor at the Shanghai International Banking and Finance Institute and Paris School of Business. The article reflects the author's opinion, and not necessarily the views of CGTN.)