Monday, November 12, 2018

System transformations in development

Systems change is the buzzword in development today. The presumption is that wicked problems in development cannot be solved with piecemeal projects, however innovative, and demand simultaneous system-wide interventions to change both processes and behaviours and attitudes.

While I am in agreement with the need to reform systems to realise truly sustainable change and the desired development outcomes, I am less sanguine about its prospects. For a start, there are truly very few examples of such changes from any country. Two, I think very few stakeholders actually appreciate what it takes to realise system-wide change. 

The danger with such efforts is that it can get sucked into a rabbit hole with the strong likelihood of little to show for even after a reasonable period of experimentation. We know what a good system looks like and maybe its ingredient elements but know very little about the relative importance associated with each element much less how to get there.

Take the example of systems change in bureaucracies. It would involve having in place attributes like goal clarity, motivated officials, systemic appetite for change, performance accountability, guidance and monitoring systems, stable leadership, and political and bureaucratic commitment. Addressing each one of these is a challenge in itself and we know very little about how to achieve that. Further, creating an objective function of all these to assess their relative importance would be almost God-like. Compounding problems, this transformation has to be achieved in an environment of acutely weak state capacity. 

The only practical response to this problem is to start somewhere promising, a minimum viable product (MVP), and opportunistically engage very actively to steer the course. The starting point is therefore very critical and has to be a powerful enough lever with potential to trigger off system-wide ripples as well as be ripe enough in time to for the change momentum to be strong enough.  

Once the starting point is identified, the next step is to figure out the pathways to change. In the context of developing countries like India, I can think of at least four pathways to change.

1. Positive deviances effect - Spot champions of change at a reasonably small but systemically important enough level and support them to emerge as positive deviances. This is a two-dimensional exercise - identify appropriate level, and spot champions - which require the deepest contextual and systemic understanding. Once we have a small sample of such champions out there, one can only hope that some of them are replaced by equally good successors and the reforms have enough time to stick. 

The whole e-governance eco-system in India can trace its origins to several positive deviances, both on use-cases as well as work-flow innovations, that emerged over the early noughties across districts of India. Over time the best features of different positive deviances got consolidated to form robust e-governance solutions across the entire spectrum of development activities. 

Globally, the use of PPPs in infrastructure contracting has evolved gradually through the emulation of positive deviances in different types of contracting structures and across different sectors, especially in UK, Canada, Australia, and Latin America. Resources like the World Bank's PPIAF have helped enormously in consolidating the learnings and providing resources which interested parties could take off the shelf and use with minimal customisation. 

2. Anchor interventions - Given the amorphous nature of what is required to realise system transformation, it is useful to have a clear and actionable starting point. Technology platforms (and not solutions) - by consolidating data and making available actionable information as decision-support and monitoring system, and which can be used to gradually build out modules covering other use-cases - are good examples of likely anchor interventions. 

For example, India's Swachh Bharat Mission or Smart Cities programs are good examples of anchors for systemic changes on sanitation and urbanisation. Similarly, the Ease of Doing Business (EoDB) in improving business environment and National Assessment Survey (NAS) and School Education Quality Index (SEQI) in improving learning outcomes are other good examples. But as all these show, the anchors by themselves are no good without good execution and opportunistic engagement to attract and integrate other elements of system transformation as we go along. 

3. Mutations - Another way to realise systemic change is through sudden regime shifts which replace legacies - new rules of the game, new personnel, new entities, new political regimes etc. They are essentially mutations which cannot be planned for, though one could work actively to create the conditions for them. The more wicked and universal the problem, the greater the difficulty of having this pathway to change.  

The Indian Insolvency and Bankruptcy Code (IBC) is a recent example of how a mutation can trigger the emergence of conditions that aligns incentives of lenders and borrowers to create a strong bankruptcy and resolution environment. The most prominent examples of such mutations over the past three decades are China's and India's liberalisation, the collapse of the Communist regimes etc. 

4. Cumulative effect - Finally, things just happen. In such cases it is difficult to attribute the change to any proximate causes but more likely the cumulative effect of several diffuse factors over a long period of time. The problem with such changes is that not only can it be planned but also once the change happens it is unlikely to have champions ready to lead the change. 

For example, the shift in the debate from inputs to learning outcomes in India over the past 3-4 years does not have any immediate cause. It has been the cumulative result of several factors, the most prominent being the receding importance of inputs and the growing salience of poor learning outcomes and its adverse impact on productivity and economic growth. But the cumulative nature of the change has also meant that the agenda on what and how to move to addressing deficient learning outcomes has remained still-born and without any champions to lead it. 

It is also the case that many changes are path dependent - the change requires traversing the path of exhausting alternative options and demonstrating their inadequacy or failure, thereby creating the conditions for acceptability of the desired change element. This is the political reality of any system. In fact, in many cases the desired change elements themselves emerge only as we progress with implementation. 

Take the example of focus on access and inputs in case of health and education. In the absence of buildings, teachers, and student attendance, focusing on quality could not have been a bureaucratic of political imperative. Or without the difficult experience and struggles of (and memory thereof) having done engineering works through government agencies, it would be impossible for systems to embrace PPPs contracts and concessions. Or of austerity, despite the knowledge of near certain suffering and damage, invariably preceding expansionary policies since the moral hazard concerns in its absence are considered to be prohibitive. Or the political will to embrace delegation of power requires a history of failure with centralised administration. 

One would have noticed that I have not mentioned evidence as a pathway to significant change in case of complex development challenges. Its role lies more in being one of the many contributors to creating the conditions required for systemic change. By itself, I cannot think of any evidence (or even a set of consciously created evidences) having been significant proximate contributors to triggering decisions that led to transformative systemic changes. 

Actually presence of right people at the right place at the right moment in time - or Overton windows - is perhaps the most important requirement to realise systemic transformations. Funny that in this age of evidence-based policy making, very little of development thinking and engagement goes in this direction. 

Each of these pathways to change are not mutually exclusive and they often co-exist and get strengthened by association of some or all of others. 

As a note of caution, while I can now think of only these four factors, it is entirely likely that there are other pathways to change. Further, while the post is likely to be relevant in most country contexts, I have written it from the perspective of system change in India. 

In another post, I shall focus on the distinction between leader's personality-induced perception of systems transformation and real systems transformation.  

Sunday, November 11, 2018

Weekend reading links

1. On the traffic related negative externalities associated with ride hailing services,
Far from reducing congestion by encouraging people to give up their cars, as many had hoped, ride-hailing seems to increase it. Bruce Schaller, a transport consultant, estimates that over half of all Uber and Lyft trips in big American cities would otherwise have been made on foot or by bike, bus, subway or train. He reckons that ride-hailing services add 2.8 vehicle miles of driving in those cities for every mile they subtract. A new working paper by John Barrios of the University of Chicago and Yael Hochberg and Hanyi Yi of Rice University spells out one deadly consequence of this increase in traffic. Using data from the federal transport department, they find that the introduction of ride-sharing to a city is associated with an increase in vehicle-miles travelled, petrol consumption and car registrations—and a 3.5% jump in fatal car accidents. At a national level, this translates into 987 extra deaths a year.
As I have blogged earlier, this really should not have been a matter of debate. Ride-hailing services clearly induce more passenger vehicles into the roads. And that is precisely what public policy should strive to discourage.

2. This timeline of events on the insolvency resolution process in India is very informative.
One of the true "big bang" reforms and a genuine success of the government. The challenge is now its implementation - both in terms of supply (capacity and integrity of NCLT, IRP, CoC etc) and demand (market absorption, debtors gaming etc). 

And it also raises the question about whether such reforms could not have been pushed without the gravity of the crisis being so. 

3. Good illustration of the challenges associated with infrastructure project bond financing,
A 2016 bond issued for the construction of a public‐private partnership (PPP) hospital in Turkey is frequently mentioned as a model for how MDBs can credit‐enhance project bonds, but is in many ways more illustrative of its difficulties. The project involved a risk guarantee provided by the Multilateral Investment Guarantee Agency (MIGA) for €208 million (out of a total €288 million bond) in the event of expropriation, transfer restriction or breach of contract. This was supplemented with a liquidity facility by the EBRD for €89 million to mitigate project construction risk and potential payment delays from MIGA’s guarantee caused by required arbitration procedures. Even with this credit enhancement package, investor interest was insufficient to float the bond publicly. It was in the end a private deal, and a substantial portion was purchased by other development finance institutions (DFIs): IFC for €80 million, France’s Agence Française de Développement‐ Proparco for €40 million and Holland’s Nederlandse Financierings‐Maatschappij voor Ontwikkelingslanden (FMO) for €20 million.
The EBRD support was effectively a guarantee on a guarantee!

4. Juan Ketterer and Andrew Powell from IADB are the latest to highlight the unbundling of construction and maintenance phases of an infrastructure asset and finance them accordingly. They also suggest refinancing of assets post-construction with bond issuances and establishment of national infrastructure development funds that aggregate projects and fund them. 

5. The Indian Economic Survey 2017-18 said that 12% of urban housing stock is vacant. It triggered a flurry of responses to address this "problem".

So how does this 12% compare with that in other countries. The Bloomberg has an article which shows that over 50 million units or 22% of China's urban housing stock is unoccupied and 12-15% is pretty much the norm across countries.

Clearly there is a natural dynamic associated with all systems. In this case, it is clear that even in  a reasonably efficient steady state 12% of houses are likely to remain vacant for various reasons - temporary migration, second property purchases, speculation, and just the flow from purchase to moving in. Further, the numbers are likely to be even lower in average percentage terms across Indian cities if we take out the largest 4-5 cities whose absolute house numbers skew the percentage.

6. Too many of canonical experiments that purport to illustrate evidence on something are being debunked. The Easterlin Paradox (and this) and Marshmallow Test (and this) are but just two examples. The latest is new research which traces a biochemical cause for the Placebo effect - fake treatments with no plausible reason to create an effect. 
Depression, back pain, chemotherapy-related malaise, migraine, post-traumatic stress disorder: The list of conditions that respond to placebos — as well as they do to drugs, with some patients — is long and growing. But as ubiquitous as the phenomenon is, and as plentiful the studies that demonstrate it, the placebo effect has yet to become part of the doctor’s standard armamentarium — and not only because it has a reputation as “fake medicine” doled out by the unscrupulous to the credulous. It also has, so far, resisted a full understanding, its mechanisms shrouded in mystery. Without a clear knowledge of how it works, doctors can’t know when to deploy it, or how... theories, which posit that the mind acts upon the body to bring about physical responses, tend to strike doctors and researchers steeped in the scientific tradition as insufficiently scientific to lend credibility to the placebo effect...


Aided by functional magnetic resonance imaging (f.M.R.I.) and other precise surveillance techniques, Kaptchuk and his colleagues have begun to elucidate an ensemble of biochemical processes that may finally account for how placebos work and why they are more effective for some people, and some disorders, than others. The molecules, in other words, appear to be emerging. And their emergence may reveal fundamental flaws in the way we understand the body’s healing mechanisms, and the way we evaluate whether more standard medical interventions in those processes work, or don’t. Long a useful foil for medical science, the placebo effect might soon represent a more fundamental challenge to it... the placebo effect is a biological response to an act of caring; that somehow the encounter itself calls forth healing and that the more intense and focused it is, the more healing it evokes... “Rituals trigger specific neurobiological pathways that specifically modulate bodily sensations, symptoms and emotions,” he wrote. “It seems that if the mind can be persuaded, the body can sometimes act accordingly.” 
Does this also mean that Eastern systems of medicine which focus as much on the mind as on the body itself is correct on purely technical basis as well?

7. FT covers how the infamous Malaysian 1MDB scandal is engulfing Goldman Sachs with the DoJ indictment of three of its senior executives, including the co-head of Asia-Pacific investment banking division. The investment bank had raised $6.5 bn for the Malaysian state investment fund in 2012-13 pocketing nearly $600 m in fees in the process.

Goldman is presenting the case legally as one of 'rogue employees' and not known by its leadership. But it is a very hard act to pull off since these 'rogue' employees were themselves part of the leadership.

8. Finally, Times has a long overdue post on the corrosive political implications of business concentration. Tim Wu writes,
Over the last two decades, more than 75 percent of United States industries have experienced an increase in concentration, while United States public markets have lost almost 50 percent of their publicly traded firms. There is a direct link between concentration and the distortion of democratic process. As any undergraduate political science major could tell you, the more concentrated an industry — the fewer members it has — the easier it is to cooperate to achieve its political goals. A group like the middle class is hopelessly disorganized and has limited influence in Congress. But concentrated industries, like the pharmaceutical industry, find it easy to organize to take from the public for their own benefit. Consider the law preventing Medicare from negotiating for lower drug prices: That particular lobbying project cost the industry more than $100 million — but it returns some $15 billion a year in higher payments for its products.

Wednesday, November 7, 2018

Adults in the Room - observations

Read Yanis Varoufakis' chronicle of his tenure as Finance Minister of Greece. 

Several interesting bits. Great account of how European institutions actually work and the power balance within. Clearly Germany matters and those at the helm don't try to hide it.

Two points stand out. 

1. He clearly is someone who believes he cannot do wrong and his views and plans are the fail-proof. Not one admission of goof-ups that I can remember from the entire book. 

2. Resolution of any problem or situation for him is more about its technical dimension. Sample the numerous instances of the pride with which he talks about his proposals and his speeches, compared with the grudging acceptance of compromises and concessions to the Troika and other interlocutors.

Without in anyway condoning the pig-headedness of the institutions on austerity and the failings of European politics, it is no surprise that Yanis failed. Just those behavioural attributes meant that he had to. A perfect illustration of a brilliant technocrat struggling with navigating his ideas by overcoming real-world challenges. 

Two standout extracts

On insiders and outsiders,
“There are two kinds of politicians: insiders and outsiders. The outsiders prioritise their freedom to speak their version of the truth. The price of their freedom is that they are ignored by the insiders, who make the important decisions. The insiders, for their part, follow a sacrosanct rule: never turn against other insiders and never talk to outsiders about what insiders say or do. Their reward? Access to inside information and a chance, though no guarantee, of influencing powerful people and outcomes. So Yanis, which of the two are you?”
On conspiracy theories (applies as much to claims of grand strategies),
“There are what I called “super black boxes”, whose size and import is so great that even those who created and control them cannot fully understand their inner workings : for example, financial derivatives whose effects are not truly understood even by the financial engineers who designed them, global banks and multinational corporations whose activities are seldom grasped by their CEOs, and of course governments and supranational institutions like the IMF, led by politicians and influential bureaucrats who may be in office but are rarely in power. They too convert inputs – money, debt, taxes, votes – into outputs – profit, more complicated forms of debt, reductions in welfare payments, health and education policies...
When a large-scale crisis hits, it is tempting to attribute it to a conspiracy between the powerful. Images spring to mind of smoke-filled rooms with cunning men (and the occasional woman) plotting how to profit at the expense of the common good and the weak. These images are, however, delusions. If our sharply diminished circumstances can be blamed on a conspiracy, then it is one whose members do not even know that they are part of it. That which feels to many like a conspiracy of the powerful is simply the emergent property of any network of super black boxes. The keys to such power networks are exclusion and opacity. Recall the ‘Greed if great’ ethos of Wall Street and the City of London in the years before the 2008 implosion. Many decent bank employees were worried sick by what they were observing and doing. But when they got their hands on evidence or information foreshadowing terrible developments, they faced Summers’s dilemma: leak it to outsiders and become irrelevant; keep it to themselves and become complicit; or embrace their power by exchanging it for other information held by someone else in the know, resulting in an impromptu two-person alliance that turbocharges both individuals’ power within the broader network of insiders. As further sensitive information is exchanged, this two-person alliance forges links with other such alliances. The result is a network of power within other pre-existing networks, involving participants who conspire de facto without being conscious conspirators. Whenever a politician in the know gives a journalist an exclusive in exchange for a particular spin that is in the politician’s interest, then journalist is appended, however unconsciously, to a network of insiders. Whenever a journalist refuses to slant their story in the politician’s favour, they risk losing a valuable source and being excluded from that network. This is how networks of power control the flow of information: through co-opting outsiders and excluding those who refuse to play ball. They evolve organically and are guided by a supra-intentional drive that no individual can control, not even the President of the US, the CEO of Barclays or those manning the pivotal nodes in the IMF or national governments.”

Sunday, November 4, 2018

Weekend research papers reading links

1. Erik Brynjolfsson, Chad Syverson, and Daniel Rock find a productivity J-curve as General Purpose Technologies (GPTs) evolve,
General purpose technologies (GPTs) such as AI enable and require significant complementary investments, including business process redesign, co-invention of new products and business models, and investments in human capital. These complementary investments are often intangible and poorly measured in the national accounts, even if they create valuable assets for the firm. We develop a model that shows how this leads to an underestimation of output and productivity in the early years of a new GPT, and how later, when the benefits of intangible investments are harvested, productivity will be overestimated... The error in measured total factor productivity therefore follows a J-curve shape, initially dipping while the investment rate in unmeasured capital is larger than the investment rate in other types of capital, then rising as growing intangible stocks begin to affect measured production...


This period can be of considerable length. For example, the technologies driving the British industrial revolution led to “Engels’ Pause,” a half-century-long period of capital accumulation, industrial innovation, and wage stagnation. In the later GPT case of electrification, it took a generation as the nature of factory layouts was re-invented.
2. Stephan Heblich, Daniel M Sturm, and Stephen J Redding seek to quantify the impact of transport technologies on the urban economy. Their model uses data for London from 1801-1921 and the introduction of steam railways and finds for the period,
... that removing the entire railway network reduces the population and the value of land and buildings in Greater London by 20 percent or more, and brings down commuting into the City of London from more than 370,000 to less than 60,000 workers.
3. Stephen Cecchetti and Enisse Kharroubi find evidence of the adverse impact of credit growth on productivity,
We examine the negative relationship between the rate of growth in credit and the rate of growth in output per worker. Using a panel of 20 countries over 25 years, we establish that there is a robust correlation: the higher the growth rate of credit, the lower the growth rate of output per worker. We then proceed to build a model in which this relationship arises from the fact that investment projects that are more risky have a higher return. As their borrowing grows more quickly over time, entrepreneurs turn to safer, hence lower return projects, thereby reducing aggregate productivity growth. We take this theoretical prediction to industry-level data and find that credit growth disproportionately harms output per worker growth in industries that have either less tangible assets or are more R&D intensive.
And their conclusions on financial sector growth are very important,
First, the growth of a country's financial system is a drag on productivity growth. That is, higher growth in the financial sector reduces real growth. Financial booms are not, in general, growth-enhancing. Second, using sectoral data, we examine the distributional nature of this effect and find that credit booms harm what we normally think of as the engines for growth – those industries that have either lower asset tangibility or high R&D-intensity. This evidence, together with recent experience during the financial crisis, leads us to conclude that there is a pressing need to reassess the relationship of finance and real growth in modern economic systems.
This graphic on the R&D intensity of various manufacturing industries is interesting

4. Falk Brauning and Victoria Ivashina find significant spillovers from US monetary policy on emerging market economies through the foreign banks' lending channel.
Foreign banks’ lending to firms in emerging market economies (EMEs) is large and denominated predominantly in U.S. dollars... Outstanding shares of foreign banks’ dollar credit for African, American, and Asian emerging economies are over 90 percent. Even for emerging Europe, this number is 60 percent... This creates a direct connection between U.S. monetary policy and EME credit cycles. We estimate that over a typical U.S. monetary easing cycle, EME borrowers experience a 32-percentage-point greater increase in the volume of loans issued by foreign banks than do borrowers from developed markets, followed by a fast credit contraction of a similar magnitude upon reversal of the U.S. monetary policy stance. This result is robust across different geographies and industries, and holds for U.S. and non-U.S. lenders, including those with little direct exposure to the U.S. economy. EME local lenders do not offset the foreign bank capital flows, and U.S. monetary policy affects credit conditions for EME firms, both at the extensive and intensive margin. Consistent with a risk-driven credit-supply adjustment, we show that the spillover is stronger for riskier EMEs, and, within countries, for higher-risk firms.
In case of EM's loans are the dominant form of external liability compared with bonds for developed markets, and foreign banks share of all external liability is for EMs is double that of developed countries.
 5. Raj Chetty et al map children's adult life outcomes based on their childhood circumstances. 
We construct a publicly available atlas of children's outcomes in adulthood by Census tract using anonymized longitudinal data covering nearly the entire U.S. population. For each tract, we estimate children's earnings distributions, incarceration rates, and other outcomes in adulthood by parental income, race, and gender. These estimates allow us to trace the roots of outcomes such as poverty and incarceration back to the neighborhoods in which children grew up. We find that children's outcomes vary sharply across nearby areas: for children of parents at the 25th percentile of the income distribution, the standard deviation of mean household income at age 35 is $5,000 across tracts within counties. We illustrate how these tract-level data can provide insight into how neighborhoods shape the development of human capital and support local economic policy using two applications. First, the estimates permit precise targeting of policies to improve economic opportunity by uncovering specific neighborhoods where certain subgroups of children grow up to have poor outcomes. Neighborhoods matter at a very granular level: conditional on characteristics such as poverty rates in a child's own Census tract, characteristics of tracts that are one mile away have little predictive power for a child's outcomes. Our historical estimates are informative predictors of outcomes even for children growing up today because neighborhood conditions are relatively stable over time. Second, we show that the observational estimates are highly predictive of neighborhoods' causal effects, based on a comparison to data from the Moving to Opportunity experiment and a quasi-experimental research design analyzing movers' outcomes. We then identify high-opportunity neighborhoods that are affordable to low- income families, providing an input into the design of affordable housing policies. Our measures of children's long-term outcomes are only weakly correlated with traditional proxies for local economic success such as rates of job growth, showing that the conditions that create greater upward mobility are not necessarily the same as those that lead to productive labor markets.

6. Finally, Tyler Watts, Greg Duncan, and Haonan Quan, have a 900 student sample study which questions the findings of the famous Marshmallow test which appeared to show that children who were able to exercise self-control and resist marshmallows placed before them did better in life. They find limited support for delayed gratification leading to better outcomes and claims that circumstances matter more,
Instead, it suggests that the capacity to hold out for a second marshmallow is shaped in large part by a child’s social and economic background—and, in turn, that that background, not the ability to delay gratification, is what’s behind kids’ long-term success... This new paper found that among kids whose mothers had a college degree, those who waited for a second marshmallow did no better in the long run—in terms of standardized test scores and mothers’ reports of their children’s behavior—than those who dug right in. Similarly, among kids whose mothers did not have college degrees, those who waited did no better than those who gave in to temptation, once other factors like household income and the child’s home environment at age 3 (evaluated according to a standard research measure that notes, for instance, the number of books that researchers observed in the home and how responsive mothers were to their children in the researchers’ presence) were taken into account. For those kids, self-control alone couldn’t overcome economic and social disadvantages.


The failed replication of the marshmallow test does more than just debunk the earlier notion; it suggests other possible explanations for why poorer kids would be less motivated to wait for that second marshmallow. For them, daily life holds fewer guarantees: There might be food in the pantry today, but there might not be tomorrow, so there is a risk that comes with waiting. And even if their parents promise to buy more of a certain food, sometimes that promise gets broken out of financial necessity. Meanwhile, for kids who come from households headed by parents who are better educated and earn more money, it’s typically easier to delay gratification: Experience tends to tell them that adults have the resources and financial stability to keep the pantry well stocked. And even if these children don’t delay gratification, they can trust that things will all work out in the end—that even if they don’t get the second marshmallow, they can probably count on their parents to take them out for ice cream instead.

Saturday, November 3, 2018

Globalisation graphics

PIIE have an excellent feature on globalisation.

1. Global trade has skyrocketed since about the 1950s, and especially since the nineties.
2. This has coincided with steep declines in tariffs across major economies.
3. Four-fifths of world trade is driven by supply chains of multinationals. Trade in intermediate goods is double that of final goods.
4. US had an overall trade deficit of $447 bn in 2017, with a very large deficit in goods offset partially by a smaller surplus in services.
5. Since the war, the share of jobs in US manufacturing has declined steadily, with that of services rising from just over half to 84% from the fifties to 2018. But strikingly, even as the absolute number of employees in manufacturing have declined, the value of output has increased significantly.
6. The balance sheet of post-war global trade expansion on the US economy is unambiguous.
7. But the aggregate hides concentrated particulars. This about the very localised displacement effects of globalisation on population clusters is central to the backlash against it,
Certain manufacturing and industry workers in specific geographic regions lost out, such as those in furniture, apparel, steel, auto parts, and electrical equipment industries in Tennessee, Michigan, and the mid-Atlantic states... The problem is compounded because policymakers have done little to help workers and communities adjust at a time when the wealthiest Americans have gained the most in recent years.
The answer to this problem is well known,
Instead of sacrificing trade gains, many economists recommend domestic policies like wage insurance, expanded tax credits, better unemployment benefits, and subsidies for health insurance for all displaced workers regardless of the cause. Such policies could reduce worker anxiety about job turnover across the board, whether it be from trade or other bigger factors. Currently, there is government support through a program called Trade Adjustment Assistance (TAA), though it only helps workers directly impacted by trade and the amounts paid are limited... Broader domestic policies can also help workers adapt to the continuously changing job market, such as access to higher education and health care, but Americans remain conflicted about the government’s role in these social safety net programs. Other advanced economies have generally increased the size of government programs as they opened up to trade.
But the challenge is with walking the talk with actual policy actions,
The United States spends only a fifth of what other advanced economies spend on average to help people find new jobs through education, training, job search assistance, and other active labor market programs.
The relevance of economy-wide social safety nets to cushion those impacted and vulnerable to various trends assumes even greater importance in light of technology and globalisation induced practices like automation and production unbundling and offshoring. No country, including developing ones, is immune to these forces. 

Friday, November 2, 2018

Tax avoidance in the US

Gabriel Zucman examines the evolution of taxes paid by US multinationals on their foreign profits since 1966 and comes up with two conclusions,
First, taxes paid to oil-producing States have been contained, allowing U.S. petroleum firms to earn high after-tax rates of returns... U.S. multinationals pay much lower tax rates to oil-producing states today than in the 1970s: while this rate averaged 70% from 1966 to 1990, it has averaged 45% since 1991. The foreign tax rates of U.S. oil multinationals fell significantly after the first Gulf War... Although it is not possible to know for sure what caused this decline, a possible interpretation of the fall in the taxes collected by oil-producing countries—and more broadly, of the favorable sharing of oil rents that U.S. multinationals have enjoyed in the long run—is that they reflect a return on military protection granted by the United States to oil-producing States...

Second, the effective foreign tax rate of U.S. multinationals in sectors other than oil has collapsed since the mid-1990s. While part of this decline is due to the fall of corporate tax rates abroad, by our estimates almost half of it owes to the rise of profit shifting to tax havens. In 2015, about half of the foreign profits of non-oil U.S. multinationals are made in non-haven countries where they face effective tax rates of 27%, and about half are booked in tax havens where they face effective rates of 7%. Following the Tax Cuts and Jobs Act of 2017, these profits are only liable for a small amount of residual tax in the United States—and hence profit shifting has proved to be an effective way for U.S. firms to cut their taxes and boost the after-tax returns on their foreign operations... 
Virtually all of the cross-border return differential enjoyed by the United States comes from the yield differential on direct investments, and in particular from the high post-tax yield on U.S. direct investments (DI) abroad. Here we show that about half of the DI yield differential can be explained by the favorable sharing of oil rents and the fact that U.S. non-oil multinationals have successfully avoided paying part of their foreign taxes.
The tax havens include Ireland, Luxembourg, Netherlands, Switzerland, Singapore, Bermuda, and Caribbean havens. Oil sector includes oil and gas extraction, petroleum and coal products, and wholesale petroleum.

In recent years almost half of the foreign profits of US MNCs has been booked in tax haven affiliates, with 18% in Ireland alone.
Reflecting the sudden shift post-1990, the tax rates paid by US oil MNCs to foreign governments declined sharply from over 70% to about 45%. There are perhaps multiple reasons for the decline.
On the non-oil side, the foreign tax rate (foreign corporate income tax payment divided by pre-tax profits net of interest payments and capital depreciation) paid by US MNCs declined from 32% in 2000 to about 17% in 2015. This makes the effective tax rates in non-havens almost 20 percentage points higher in 2015. As to the reasons, Zucman and Wright decompose the 15 percentage points fall to relocating of capital and labour to low-tax places (1 point), fall in foreign statutory tax rates (8 points), and the rise of profit shifting to tax havens (6 points).
Haven non-oil affiliates, with just 10-15% of the wages and tangible capital, make up almost half of all pre-tax profits, which also corresponds to a capital share of net corporate output of 75-90%! In contrast, non-haven non-oil affiliates which employ the bulk of US MNCs tangible capital and workforce, pre-tax profits to wage ratios are less than 35%, which corresponds to a capital share of net output below 25%.
These haven affiliates have pre-tax profits-to-wages ratio of 300-800%