The relationship between the value of the internet and its openness could be symbiotic
“When I use a word,’ Humpty Dumpty said in rather a scornful tone, ‘it means just what I choose it to mean—neither more nor less.’ ‘The question is,’ said Alice, ‘whether you can make words mean so many different things.’ ‘The question is,’ said Humpty Dumpty, ‘which is to be master—that’s all.”
“Internet openness” is tricky. While there seems to be support from all sides for ensuring that the internet remains a democratic and plural space (just last month, the minister for communications and IT stated how “we instinctively value the internet to be open, plural and inclusive”), there is also a growing concern that this “openness” is under increasing threat. In the democratic world, and perhaps elsewhere it is almost an article of faith to accept that the internet has had enormous impact on communications, politics and business both within and across borders. In some instances, the internet has disrupted, creating new ways of doing old things, while in others it has spawned new models unthinkable in the unconnected world. Beliefs however are not enough for crafting good and sustainable policy. The area most fertile for documenting the benefits of more and better internet connectivity is economic well-being.
Thus it adds to GDP, enables efficiency, makes businesses more productive and expands the geographical scope of markets. The tough part however is to try and value its “openness”. How much of the internet’s benefits stem from it being open? In other words, would net value be destroyed if we ‘closed’ it down?
We do believe that openness substantially adds not just to the economic value of the internet, but also its political value. Any attempt to measure this quantitatively however immediately confronts the Humpty Dumpty problem of definition. We mean by openness what we choose to mean, and our choices are not always the same.
A review of the existing work on internet openness along with our discussions with a number of experts revealed that any attempt to define ‘openness’ would quickly become a daunting pursuit, necessitating attention on many fronts, each requiring judgements to be made. Consider the following. Openness of the internet is decidedly affected by the state of freedom of expression (both online as well as offline) and access to knowledge including a reasonable intellectual property rights regime and “fair use” provisions.
How does one proxy these? In addition, it is also influenced by alternate IPR regimes such as “free and open source software”, open standards and open content and the availability and curation of cultures of Open Data and Open Government Data. Measurement anyone? And ditto for policy measures that are asserted in the interest of national sovereignty and the creation of “borders” on the internet. These include mandatory data localisation and local routing—sometimes pejoratively referred to as the “balkanisation of the internet”. The quality and nature of access also counts, for example making the internet available in all languages as well as for those who are deaf or blind. Regulations in the “physical layer” can arguably change the degree of openness—bans on unlicensed spectrum can prevent the deployment of free wi-fi clusters in poor neighbourhoods and onerous Know-Your-Customer requirements for public access points (such as internet cafes) can prevent those without identification from going online. And while access is not the same as openness, there is reason to believe that increase in access positively impacts the value of internet openness. But while such a broad perspective can help one understand the enormous scope of openness, it simultaneously makes the task of measurement herculean.
An alternative approach is to estimate the value that would be unlocked by relaxing an identifiable constraint without defining the entirety of the internet’s form. For example, restrictions on VoIP not only increase the costs of calling but also discourage calls that would otherwise have taken place. Limitations on access to information are costs in themselves but these also reduce the perceived value of the Internet itself and disincentivise activity online. Transactions costs may increase through limitations on how e-commerce is conducted or data collected for advertising, while localisation policies may add significantly through redundant costs of duplicating hardware. Or one could take advantage of episodic disruptions such as ‘killing’ of the internet in Egypt during the Arab Spring for 5 days. OECD estimated the cost of the blackout at about $90 million, or $18 million a day, 3-4% of the country’s GDP for the period of the blackout. The problem with estimating the ‘shadow value’ of individual constraints or episodes is that they cannot be extended to the entire internet because of general equilibrium considerations.
Consequently, putting a number on the exact value “openness” or “closedness” is a near-impossible problem, one that is made worse by a lack of data and dependence on counterfactual reasoning. There are of course strong public-interest reasons for why the Internet must not be completely unregulated—the protection of user rights and ensuring national security are essential public policy objectives—but such decisions must be made in the context of data-based analyses of the costs and benefits.
One possible solution that can help finesse these problems is to instead use relative measures to determine the secular trajectory of an index on openness. Measures such as Freedom House’s “Freedom on the Net” index and the World Economic Forum’s “Network Readiness” index track countries’ internet openness over time, relative to other countries. Building on these can help glean evidence of relative changes in a country’s internet regime over time, as well as determine whether or not there exists a correlated change in internet value. The policy firm Dalberg has attempted this, plotting countries’ “Freedom on the Net” scores over time against the internet’s “economic impacts” on their economies (measured as a sub-index in the WEF’s Network Readiness exercise). The results indicate a remarkably close positive relationship between how intensely countries use the internet and ICT and how “free” and “open” their regimes are as measured by the FoTN index. Replicating this exercise to plot “freedom” scores against measures of the Internet’s contribution to countries’ GDP shows similar results (with China being the one notable outlier).
So, does openness add to internet value? While the answer might depend on the definition more than we care to admit, it appears that increased openness is good for unlocking the internet’s economic potential. Following this up with robust measures of the value of the internet and its impact on a country’s GDP over time can thus help understand the dollar value of such openness, adding data-based robustness to the formulation of internet-policy. Increasing academic and policy focus on understanding the contours of internet openness, its impacts on economic activity, and measuring their magnitude is thus essential for a better understanding of what policies are best in the long term.
It may well be that the relationship between the value of the internet and its openness is symbiotic—greater openness may spur stronger growth in the internet economy and its contribution to GDP, while the increasing intensity of the internet’s role in the economy may simultaneously stimulate demands from users for greater openness. Like Portia’s quality of mercy, openness is likely to be twice blest; It blesseth the government that gives and the internet that takes.
By Rajat Kathuria, Vatsala Shreeti & Parnil Urdhwareshe
Kathuria is director & chief executive at ICRIER, where Shreeti and Urdhwareshe are research assistants.
Views are personal