On 18th February, 2016, Uganda Communications Commission, the Telco regulator, ordered all ISPs to sever access to Facebook, Twitter and WhatsApp. 11 million Internet users, including myself were forced to live through a four-day Internet blackout.

With this unprecedented move, Uganda joined Syria, Russia, Egypt, Burundi and other regimes that have weaponized the Internet to curtail free speech and access to information. With the alarming frequency of use, you would think Internet blackouts actually deliver results—but they are not as effective as you might think.

From Uganda’s four-day ordeal, I can tell you this much: Using Internet blackouts to suppress free speech is like chasing a wild-goose: governments do not get the intended results. Instead, the collateral damage foments resentment, mistrust, and wrecks irreparable reputation damage.

When Government of Uganda decided to restrict Internet access, it hoped to control its citizens’ access to information. However, within 48 hours, one-and-a-half million people had circumvented the blockage and were back online. You do not need a military strategist to tell you how ill-conceived this strategy was in the first place.

Governments needs to understand that implementing an effective Internet blackout is next to impossible. Workarounds are pervasive: Egyptians and Syrians used SMS-to-Twitter services, and satellites to circumvent complete Internet outages in 2011, and today VPNs are being used to thwart selective blockages in Uganda and Russia. Effective blackouts are not just impractical, they backfire almost every time.

As a matter of fact, even the most benign actions after Uganda’s blackout was met with paranoia from the 1.5 million people who had circumvented the blockage. Similar reactions have been seen in Sudan, Iran and Russia where Internet blackouts have been used.

Indeed, the Arab Spring in 2011 was fueled by Internet cuts in Egypt. Ordinary citizens were forced to get out on the streets to look for information because of Internet and media blackouts—they ended up joining protest movements. This is another example of how governments lose by perpetrating Internet blackouts.

Moreover, as sure as night follows day, the Streisand Effect, where an attempt to hide or censor information publicizes it more widely, always follows an Internet blackout. Global attention is always drawn to countries that practice Internet blockages. For instance, several parties, including the U.S Secretary of State followed and condemned Internet blackouts in Syria, in 2011 the same way they did with Uganda in 2016.

Additionally, partial Internet blackouts, where only specific sites are unavailable, fragment the Internet into small islands. We witnessed this in Burundi in April 2015, Syria since 2011, and in Uganda in 2016, where one side of the Internet was accessible to everyone without restrictions, and the other side, had a set of websites, like Facebook, Twitter and WhatsApp, et al, which were restricted.

Indeed, Internet experts, including at the World Economic Forum have already opined on the dangers of Internet fragmentation—asserting that practices like blackouts contribute to fragmentation of the Internet which is not in the greater interests of the Internet and its users.

In the end, my thoughts are informed by a desire to preserve the Internet, however the inescapable truth stands: the cost-benefit analysis of Internet blackouts simply do not favor governments. The internet is a force that cannot be controlled by any one government and trying to do so, is a recipe for disaster. In the words of Sun Tzu (The Art of War), “one must avoid battle with the strong,”

I believe international interventions are warranted in the fight to dissuade governments from further weaponizing the Internet. The idea of an international treaty, which has been flaunted by key Internet organizations like ISOC is one such approach that I support. Perhaps such interventions, coupled with the disincentives above could reduce instances of Internet blackouts.