A Four-Fold Classification of Data

One of the vexing issues while talking about data protection is the various ways in which data can be classified. Public or private, identified or identifiable, data or metadata; the glut of classifications makes it difficult to achieve consistency.

In The Master Algorithm, Pedro Domingos classifies data into four types:

  • Data that is shared with the world at large. This includes blogs, tweets, reviews on websites, etc.
  • Data that is shared with a limited circle of people. The simplest example under this category would be interactions on social media platforms.
  • Data that is shared with companies. This can often happen without the individual knowing that a company has data belonging to him.
  • Data that individuals do not wish to share with anyone, which is quite self-explanatory.

This is a simple enough classification. That said, there is scope for an overlap between the four types of data. As Domingos is quick to point out, the data from the second category could well fit the third as well, since modes of communication are often controlled by a few companies.

Despite this caveat, the classification is useful. Its simplicity works in its favour by helping more people be a part of the conversation around data protection. And one can hope that more conversations will act as a springboard for developing fresh insights into how we look at our data.

What Came First – The Tech or the Reg?

John Frank Weaver, in a 2014 blog post on Slate, argues that it is best to bring in laws to regulate future technology. His reasoning is that when technology begins to unravel, it does so at an exponential speed, making it almost impossible to regulate once this event happens. Instead, Weaver proposes that regulations should already be in place before a new technology is developed, so that it always stays ahead and wary of these speedy developments.

Precautionary as this approach might be, it is more likely to wreak a whole lot of havoc instead. It is undesirable to pass laws before the subject matter to which it would apply even comes into existence. In doing so, we might end up doing more harm than good, creating more obstacles than enabling the prevailing ones being solved organically.

Regulating technology minutely from the get go is inefficient and ineffective. Firstly, it is very difficult to say exactly which direction tech is going to proceed in. We do not know where the next big innovation, next “disruption” is going to come from. By creating a regulation well before it is time to do so, we might end up passing a law which either becomes redundant in the face of unexpected technology or worse, ends up creating a mess of an ecosystem which would have flourished otherwise.

Also, laws are enduring and it takes a lot more time and effort to undo them later. So, it would make sense to regulate things at the margin – beginning with technology that is already existing today, or an innovation on the horizon which poses concerns that are easy to imagine and to solve. For instance, it is easier to regulate a technology concern such as data protection in digital advertising than, say, artificial superintelligence.

The “regulating at the margins” approach is more desirable also because it is bound to contain useful stipulations. It also allows companies, regulators and individuals to adapt themselves to the new legal framework and to rectify their concerns quicker and more effectively.

This has the added benefit of being an innovation-friendly environment – stepping in when needed, but otherwise giving technology a defined, safe space through which it can progress.

 

 

Why has the US Policy Orthodoxy on Iran Sustained for Four Decades?

On 8th May, the US President announced that the US was withdrawing from the Iran nuclear deal. The US would also be reimposing the sanctions on Iran that were in place prior to the deal. Essentially, we are back to a hostile Iran-US relationship after a short break where a change seemed likely. Now, this hostile policy orthodoxy in the US vis-a-vis Iran has sustained itself for nearly 4 decades. And one of the foreign policy mysteries for me has been: why is that the case?

After all, Iran is one of the most “normal” states in West Asia. It is also a regional power and now there is even some alignment between US and Iranian interests in Afghanistan and over ISIS. And yet, the foreign policy of the US towards Iran hasn’t change for nearly forty years. What are the possible reasons? I asked this question to my colleagues. I’m summarising some of their responses and my own views on them.

The oft-repeated reason given is the Iran hostage crisis of 1979. It is argued that this highly televised, 444-day imbroglio is the reason behind the perception of Iran as a ‘rogue state’ in the US. I doubt if that is the case. Even though this crisis might well be the reason that set the current policy orthodoxy in motion, it does not sufficiently explain why the orthodoxy would continue for four decades. In fact, in the same year the US embassy in Islamabad was burnt. Two Americans died as a result. And yet, there was no break in the US-Pakistan relationship. So, it doesn’t seem logical that another contemporary incident of a similar nature, one in which no American hostage was killed, can create and sustain a policy orthodoxy for four decades. 

The second reason given is that the hatred towards Iran is sustained by Iran’s own acts of hostility towards the US. Indeed, Iran has often taken up the gauntlet on various occasions. But again, this reason doesn’t sufficiently explain why the policy orthodoxy did not change even after Iran demonstrated its willingness to change as part of the P5+1 negotiations. The North Korean example shows that the US that a change in relationship terms is possible even with a state belonging to the ‘axis of evil’.

The third reason given is Trump. That’s an easy one to contest though. Long before Trump came into the picture, this policy orthodoxy was still going strong.

The fourth reason give is “follow the money”. The argument is that pro-Israel and pro-Saudi lobbies in the US ensure that there is no foreign policy change in the US on the Iran issue. There is some weight in this argument and it could help explain the longevity of the policy orthodoxy. If that is case, the emergent hypothesis is that the policy change is incumbent on the Iran-Saudi Arabia-Israel triangle. Unless Iran can patch up with at least one of these two West Asian powers, the US will keep the heat on. 

In any case, this question needs methodical research. I think it’s just one of those questions in foreign policy which is not raised enough. Someone should do a study of the kind Nicolas Blarel has done to explain the orthodoxy and change in the India-Israel relationship. Or perhaps, I have been ignorant. If you know of a study that tackles this question systematically, please point me to it!

 

 

A Quid Pro Quo Life

While reading Pedro Domingos’ The Master Algorithm, a book about machine learning, I found the following extract that talks about life in the digital age:

Every transaction works on two levels: what it accomplishes for you and what it teaches the system you just interacted with. Being aware of this is the first step to a happy life in the twenty-first century.

To reassure us that this is a positive development, Domingos goes on to say that it is better to think of a computer as a tool to serve us rather than as an adversary. And what it learns from us helps it to serve us better.

I have three thoughts about this.

One, this element of reciprocity is already present in transactions between humans. A shopkeeper I buy something from, is learning about me and, by extension, his customer base. The presence of machine learning amplifies this part of our lives.

Two, this logic can be applied to the products offered by companies like Google and Facebook. This would be a more nuanced way of looking at our relationship with these enterprises than the simple, and now over-used, notion of data being the new oil.

Three, in the event that Domingos’ optimism is misplaced and a system happens to be malign, we should have the ability to walk away from a transaction.

Elaborately Learned Superstition

Check out this excellent paragraph from that Jane Jacobs masterpiece, The Death and Life of Great American Cities. Talking about the urban planners of her time, she writes:

And to put it bluntly, they are all in the same stage of elaborately learned superstition as medical science was early in the last century, when physicians put their faith in bloodletting, to draw out the evil humors which were believed to cause disease. With bloodletting, it took years of learning to know precisely which veins, by what rituals, were to be opened for what symptoms. A superstructure of technical complication was erected in such deadpan detail that the literature still sounds almost plausible. However, because people, even when they are thoroughly enmeshed in descriptions of reality which are at variance with reality, are still seldom devoid of the powers of observation and independent thought, the science of bloodletting, over most of its long sway, appears usually to have been tempered with a certain amount of common sense. Or it was tempered until it reached its highest peaks of technique in, of all places, the young United States. Bloodletting went wild here. It had an enormously influential proponent in Dr. Benjamin Rush, still revered as the greatest statesman-physician of our revolutionary and federal periods, and a genius of medical administration. Dr. Rush Got Things Done. Among the things he got done, some of them good and useful, were to develop, practice, teach and spread the custom of bloodletting in cases where prudence or mercy had heretofore restrained its use. He and his students drained the blood of very young children, of consumptives, of the greatly aged, of almost anyone unfortunate enough to be sick in his realms of influence. His extreme practices aroused the alarm and horror of European bloodletting physicians. And yet as late as 1851, a committee appointed by the State Legislature of New York solemnly defended the thoroughgoing use of bloodletting. It scathingly ridiculed and censured a physician, William Turner, who had the temerity to write a pamphlet criticizing Dr. Rush’s doctrines and calling ”the practice of taking blood in diseases contrary to common sense, to general experience, to enlightened reason and to the manifest laws of the divine Providence.” Sick people needed fortifying, not draining, said Dr. Turner, and he was squelched.

It is my case that bloodletting in the 19th century was like government regulation in the 20th and 21st. There exists a “superstructure of technical complication” that makes wonks and boffins believe in their omnipotence, but they are as wrong as Dr Rush.

We will look back on this with wonder 100 years from now. Dang, I feel like I’m trapped in the wrong century. That said, better this one than any that came before.

Light-touch Regulation for Social Media

There’s suddenly a lot of talk about how governments need to regulate social media. From a public policy perspective, the immediate cause underlying this policy change is an egregious case of misuse of social media. Policy changes that arise out of crises can often go overboard in the policy instrument deployed.

Using the simple threefold classification of carrots, sticks, and sermons, it means that governments are more likely to use sticks rather than carrots or sermons in such cases. Under the garb of user protection, governments will use the ‘need for regulating the conniving social media‘ narrative to suppress dissent. So, assuming that at least a few governments will choose to intervene, which instrument should be used?

I would advocate for a sermons approach: the government can instruct social media companies to carry a user login banner which explicitly states that

the opinions on your timeline are not be verified and may not be reflective of the truth. User discretion is advised.

Think of the banner that appeared in the beginning of the World Wrestling Entertainment telecasts:

fights are performed by professionals solely for the purpose of entertainment. Any attempt by our fans to emulate our Superstars physicality is extremely dangerous and irresponsible.

The result was that there was no ambiguity in the minds of the viewer that WWE was an entertainment show and not a gladiator fight. Perhaps, a regulation of this kind has some lessons that are relevant now.

Where do you get your money, Mr Politician?

Over the last year, the role of foreign entities in elections has become an important point of consideration. In New Zealand, a national MP Jian Yang came under fire for having undisclosed relations with the Chinese intelligence. The law in Australia has also changed to regulate political funding after revelations that 80 per cent of foreign funding came from China. And of course, there’s the open secret about Russian interference in the 2016 American elections. Rory Medcalf dubs this ‘sharp power‘ but the concept is not new. Countries have been meddling in each other’s affairs while peddling statements about justice and fairness since the conception of sovereignty.

In India, we may think that our elections are messy enough without the influence of foreign hands. Indeed, the Election Commission has its work cut out in regulating illegal cash flow by political parties during elections. But our system has been made all the more vulnerable by the 2018 Finance Bill. A sneaky little clause in the Budget disempowered the existing Foreign Contributions (Regulation) Act, 2010 retrospectively so that both the BJP and the Congress are cushioned from scrutiny. This makes the Indian system all the more susceptible to foreign influence. It may seem counter-intuitive to think of any external strings in the 2019 elections. But if the trends in countries like Australia and the United States are anything to go by, then they are not to be easily dismissed.

India-China Collaboration on Railway Line in Afghanistan?

After the Xi-Modi Wuhan summit, there is a lot of buzz around the possibility that India and China might take up a joint economic project in Afghanistan.

This prospect has got many people excited. All prominent news agencies have reported this and yet there is little clarity on what exactly this project is all about. The MEA’s press release on the Wuhan summit in fact does not mention Afghanistan at all. The MEA spokesperson has been quoted in Times of India saying that the identification of this project is still in progress. The Hindu’s report vaguely mentions the possibility of a road link to Chabahar from Aynak via Hajigak.

But Praveen Swami in Business Standard has the most clear view about what this project might be. He writes:

Earlier this week, President and Prime Minister agreed to explore joint China-India work on a railway line in Afghanistan, with one spur carrying Mes Aynak’s ore to Torkham, and over the into Pakistan; the other in a great north-western arc, into Hairatan.

I created this google map to understand how this railway line might look like. Please note that this is only an illustration – I have no more details other than the above article.

Based on the May 4, 2018 report in the Business Standard [Click to expand]

Won’t such a project pass through Taliban-controlled areas? What is the security situation like in the areas that this railway line might pass through. To check that out, I overlaid the Jan 2018 BBC illustration showing areas under Taliban/government control over the route map. That looks as follows.

The speculated railway line pass through areas with Taliban presence [Click to expand]

The areas marked in brown are under full-Taliban control. Areas in grey are in full government control. The orange areas are government-controlled areas having open Taliban presence. The darker shade indicates higher risk (attacked at least twice a week). Lightest shade of orange represents areas that are attacked once in three months on an average.

As is clear from this graphic, India, China, and the Afghan governments have a tough challenge ahead of them if they are serious about this project.

Formal Employment Data Sources: New and Old

Ajit K Ghose has an article in Business Standard on why payroll estimates data does not accurately represent job creation in the Indian economy.

What I found most interesting was that he has listed all the sources of formal employment data in India in this article. It is very rare to find all employment data sources, old and new, in one place. So, I’m noting them down here:

  1. Monthly estimates of payroll account derived from the database maintained by the Employees’ Provident Fund Organisation.
  2. Surveys of employment and unemployment conducted by National Sample Survey Organisation (NSSO).
  3. The Labour Bureau conducted a survey similar to (2) for 2015-16.
  4. Starting 2017-18, NSSO has launched a periodic labour force survey to produce annual estimates of formal employment in the country.

The Chinese Leadership’s Prowess is Overrated

Earlier this week, we had the opportunity to meet one of India’s most experienced China hands. A diplomat by profession, he has served in China, speaks Mandarin fluently, and follows developments in China even after his retirement.

During his talk, he warned us:

Indian elites tend to see Chinese leaders as superheroes — as ten-feet tall men who can do no wrong. Such a view ignores the many grievous mistakes that the Chinese leadership has committed over the years. And bad memories stick for longer than the good ones.

According to him, the two big mistakes of the Chinese leadership are their treatment of the Uighurs and the Tibetans. I have two more to add to this list.

One, the treatment meted out to rural migrants in cities is fomenting a quiet unrest in several cities. The Economist has an article on the increasing discontent:

The younger generation are products of China’s one-child policy, which went into force nationwide in 1980 (although in the countryside, families were sometimes allowed two). They are among the first to suffer its unintended consequences. The one-child policy contributed to a drastic change in the sex ratio because female fetuses were aborted by parents who wanted their only child to be a boy. The ratio of boys to girls at birth soared in the 1980s, peaking in 2005, when there were 122 baby boys for every 100 baby girls, one of the most distorted ratios ever seen [..]

Among Chinese men generally, a common response to the shortage of women is for prospective grooms to buy an apartment and car before marriage—a sort of reverse dowry. One survey found that three-quarters of young women in big cities took this into account before accepting a man’s offer. Alas for migrant swains, they cannot afford such a bride price, especially in expensive cities such as Beijing and Guangzhou. It is usually difficult for people without a city’s hukou to buy government-subsidised housing there. Young migrants are therefore at a threefold disadvantage. There are fewer women of marriageable age. Those who come from their own background tend to marry richer rivals. And the men cannot compete in the marriage market by buying property.

The earnings of the youngest ones have deteriorated the most. Mr Tian looked at earnings by age. He found that the highest earners are those in their mid-30s (between 32 and 36). That remained constant in all his surveys. But there was a significant change among workers in their mid-20s (22 to 26). In 2008 these younger migrants were earning almost as much as the best-paid. By 2015, they were earning much less.

Mr Tian’s survey includes a question about where respondents place themselves in society on a scale from top to bottom. Between 2006 and 2015 the migrants he questioned gave, on average, ever lower assessments of their social position. Initially, the younger ones (aged between 22 and 26) were the most likely to describe themselves as being in the top half of society. By 2015 they were more inclined than older migrants to put themselves in the bottom half. Mr Tian concludes that those born in the 1990s are the most disappointed of the migrants he has studied [The Economist, 3rd May 2018].

The second mistake is the way China has treated its neighbours. Its arrogant conduct has turned away even potential partners over the last few decades. Nitin Pai had written how alienating a young India has been one of the biggest mistakes of the Chinese leadership.

It is widely accepted that China was the victor of the brief border war of October 1962. While Beijing did achieve its political and military objectives – of teaching the Indian government of Jawaharlal Nehru a lesson – that was a strategic self-defeat for China. Why? Because it turned a country of young people (in 1962, half of India’s population was less than 19) into believing that China is the enemy.

Unfortunately, in the past few weeks, official statements from the Chinese government and commentary appearing in official media are taking us close to another 1962 – even if no shots are fired over the Doklam region. Contrary to what Beijing might think, threats of war and reminders of 1962 strengthen India’s national resolve to stay firm. The more strident the rhetoric from Beijing , the stronger is the public opinion in India to confront China.

Half of India’s population is under 26. Almost 70 per cent of the Indians surveyed in the above-mentioned poll were already “quite concerned” about China’s growing military power and its territorial disputes with India. Whatever the stakes on the remote Himalayan slopes, they are likely to carry an imprint of China as an adversary and an enemy well into the rest of their lives. Is it really in China’s interests to alienate half a billion people across its borders for the next several decades? Will it be easier or more difficult to achieve President Xi Jinping’s “China Dream” under these conditions? How does it help China if India is pushed into a tighter embrace of the United States? [SCMP, 25th July 2017]

These four mistakes need to be kept in mind when we discuss China’s out-of-ordinary development feats. Far too often, we tend to see China through a “rational actor model” lens, to borrow Graham Allison’s landmark classification.The reality is far more complex. The four mistakes show that Chinese leaders have not always made consistent, value-maximising choices. Perhaps it is our lack of knowledge about China that makes us ascribe rationality to every decision of the Chinese leadership.