Author: conversableeconomist

The fundamental power granted by holding patent is to block others from making use of the invention, unless they pay you a licensing fee. Inevitably, this power looks pretty good when you are the patent-holder, and not so good when you are being sued for infringing on someone else\’s patent. Just to complicate matters, some companies that bring suit for patent infringement didn\’t actually do the inventing themselves, nor do these companies actually make a product. The neutral name for these firms are \”patent assertion entities.\” A less-neutral name, when some of these firms act in a way that seems to misuse patent law as a genteel method of extorting payments from other firms, is to call them \”patent trolls.\”

There is nothing illegal about buying patents, or enforcing patent rights by requiring a license fee. Indeed, it can be a useful service for a firm to collect a group of closely-related patents and then license all of those patents as a group, so that firms wishing to use those patents can negotiate with a single entity. Is it possible to divide up the \”patent assertion entities\” in some defensible way, to distinguish the trolls from the rest?

The Federal Trade Commission has the authority to collect confidential business information. \”[T]he FTC analyzed information from 22 Responding PAEs and over 2,500 of both their Affiliates and other related entities.\” The FTC describes it findings in an October 2016 study: Patent Assertion Entity Activity. The report runs well over 200 pages, including a review of existing research literature and the relevant court cases, together with past FTC thinking on this subject over the last decade or so. It you want to get up to speed on this topic, this report is the place to start.The FTC defines the subject of its study this way:

The term “patent assertion entity,” or PAE, as used by the Commission in this report and elsewhere, refers to a firm that primarily acquires patents and seeks to generate revenue by asserting them against accused infringers. As the term underscores, PAE business models focus on asserting patents that the firm has acquired from third parties, rather than obtained from the U.S. Patent and Trademark Office (USPTO) through prosecution. Patents are a PAE’s principal asset; a PAE does not manufacture, distribute, or sell products. Merely holding a patent, however, does not generate revenue for a PAE. Instead, the firm generates revenue by licensing that patent or, more rarely, by obtaining court-ordered damages in successful patent infringement litigation. Furthermore, a PAE generally initiates negotiations that may lead to a license by communicating a demand for payment to, or filing an infringement suit against, an accused infringer.

Basically, the FTC argues that the \”patent assertion entities\” can be divided fairly neatly into two \”two distinct PAE business models: Portfolio PAEs and Litigation PAEs.\” The FTC steers away from using the term \”patent trolls,\” which in this report mainly comes up in quotations from other articles buried in the footnotes. But \”litigation PAEs\” is the category that most people are thinking of when they refer to \”patent trolls.\” Here is how \”portfolio PAEs\” functioned:

\”Portfolio PAEs most closely resembled the licensing arms of manufacturing firms; they were highly capitalized and often raised money from investors that included both investment funds and manufacturing firms. Typically, these investors received a share of the Portfolio PAE’s future revenue and a license to the Portfolio PAE’s patents. …

\”Portfolio PAEs typically conducted business in the following manner. First, they acquired portfolios of patents. Portfolio PAEs frequently acquired patents from manufacturing firms by making large up-front payments to the owner. Some Portfolio PAEs acquired hundreds or thousands of patents in individual transactions, often purchasing these patents from manufacturing firms. Other Portfolio PAEs acquired smaller numbers of patents per transaction and aggregated them into larger portfolios. Regardless of acquisition model, Portfolio PAEs then organized acquired patents into one or more portfolios, each containing hundreds if not thousands of patents and offered these portfolios for licensing. …

\”Portfolio PAEs negotiated licenses covering large portfolios, often containing hundreds or thousands of patents, frequently without first suing the alleged infringer. The value of these licenses was typically in the millions of dollars. Although Portfolio PAEs accounted for only 9% of the reported licenses in the study, they generated 80% of the reported revenue, or approximately $3.2 billion.  …

In particular, notice that the investors in these \”portfolio PAEs\” are often companies, which the get access to the patents in the portfolio. Thus, these institutions can be viewed as ameliorating the \”patent thicket\” problem, where there are so many closely related and interlocking patents in some areas that it becomes difficult for innovators to function.

Litigation PAEs look different.

\”The Litigation PAE business model frequently employed one or more affiliate entities, usually set up as limited liability companies (LLCs), each created to acquire and assert a small portfolio of patents, without bundling or aggregating acquired patents into larger portfolios. … Regardless of whether they sent demand letters, Litigation PAEs almost always sued potential licensees in district court before beginning license negotiations.  … Litigation PAEs tended to be thinly capitalized. Many had between one and three individual owners, often with no other employees and no offices outside of their owners’ homes. In fact, several Litigation PAEs were simply individual entrepreneurs who relied entirely on outside attorneys and professionals to maintain records regarding their assertion activity. …

\”Litigation PAEs typically sued potential licensees and settled shortly afterward by entering into license agreements with defendants covering small portfolios, often containing fewer than ten patents.  The licenses typically yielded total royalties of less than $300,000.  … The American Intellectual Property Law Association (AIPLA), which periodically surveys the costs of patent litigation, recently reported that defending an NPE patent lawsuit through the end of discovery costs between $300,000 and $2.5 million, depending on the amount in controversy. By this estimate, 77% of Litigation PAEs’ settlements fell below a de facto benchmark for the nuisance cost of litigation. This suggests that discovery costs, and not the technological value of the patent, may set the benchmark for settlement value in Litigation PAE cases. … Given the relatively low dollar amounts of the licenses, the behavior of Litigation PAEs is consistent with nuisance litigation.

\”For each separate patent portfolio that they acquired, Litigation PAEs characteristically
created a new affiliate entity, which often held ten patents or less. They generally operated with little or no working capital and relied on agreements to share future revenue with patent sellers to fund their businesses. Litigation PAEs filed 96% of the cases in the study and accounted for 91% of the reported licenses, but only 20% of the reported revenue, or approximately $800 million.\”

The report also notes that this problem is really centered in the electronics and software industries. \”Of all the patents held by PAEs in the FTC’s study, 88% fell under the Computers & Communications or Other Electrical & Electronic technology categories, and more than 75% of the Study PAEs’ overall holdings were software-related patents.\”

The policy question of how to draw legal distinctions put the brakes on the patent trolls without unduly hindering legitimate protection of patent rights raises some tricky questions, but the report offers some general guidance:

The FTC recognizes that infringement litigation plays an important role in protecting patent rights, and that a robust judicial system promotes respect for the patent laws. Nuisance infringement litigation, however, can tax judicial resources and divert attention away from productive business behavior. With this balance in mind, the FTC proposes reforms to: 1) address discovery burden and cost asymmetries in PAE litigation; 2) provide the courts and defendants with more information about the plaintiffs that have filed infringement lawsuits; 3) streamline multiple cases brought against defendants on the same theories of infringement; and 4) provide sufficient notice of these infringement theories as courts continue to develop heightened pleading requirements for patent cases.

For a previous discussion of these issues, I recommend \”The New Patent Intermediaries: Platforms, Defensive Aggregators, and Super-Aggregators,\” by Andrei Hagiu and David B. Yoffie, which appeared in the Winter 2013 issue of the Journal of Economic Perspectives.

The venture capital industry is fairly well-known: they raise funds from investors, often pension funds or university endowments, to invest in start-up companies at an early stage, knowing that a most of those companies won\’t do very well, but hoping to make a good return when about one in ten of those companies hits it big. The National Venture Capital Association estimates that VC funds raised $28 billion in 2015, and have about $168 billion under management. This isn\’t a large amount in the context of an $18 trillion US economy, but the effects of VC firms in choosing and nurturing start-ups are much larger than the dollar value would suggest.

But behind or perhaps along-side of the VC firms are a more shadowy group known as \”angel investors.\” which may well be investing more money than their better-known VC cousins. Angel investors are not raising a fund from others; instead, they are investing their own money in small start-ups.  Not much is known about them. But Josh Lerner and Antoinette Schoar present some evidence in their report \”Rise of the Angel Investor: A Challenge to Public Policy,\” written for the Third Way think tank (September 23, 2016). They describe the subject this way:

Angel investors are high-net-worth individuals, often (but not exclusively) former entrepreneurs and corporate executives, who make private investments in start-up companies with their own money. While individual angel investors have a long history—for instance, Naomi Lamoreaux and co-authors highlight how Cleveland’s angel investors played a critical role in financing the early electricity and automotive industries—organized angel groups are a quite recent phenomenon. Beginning in the mid-1990s, angels began forming groups to collectively evaluate and invest in entrepreneurial ventures. …  

Angels typically invest at the seed funding stage, making them among the first equity investors in a company beyond its founders. … Angels invested a total of $24.6 billion in 2015 with an average deal size of $345,390, according to the Center for Venture Research. …  The Angel Capital Association (ACA) lists more than 300 U.S. groups in its database. The average ACA angel group in 2015 had 68 member angels and invested a total of nearly $2.5 million in 10.3 deals in 2007. At least between 10,000 and 15,000 angels are believed to belong to angel groups in the U.S. … 

The precise measurement of the total size of the angel investment market is difficult to ascertain due to the fact that most angel investments are made on an individual basis and thus typically are not subject to regulatory disclosure requirements. But estimates suggest that the total size of angel investment has long surpassed venture capital investment in the U.S. and increasingly in some other countries as well. For instance, survey estimates suggest the projected size of the total angel market in the U.S. grew from $17.6 billion in 2009 to $24.1 billion in 2014. The estimated capital
deployed by angel groups in Europe has almost doubled over the past five years, and in Canada, it almost tripled. Some estimates suggest that these investors are as important for high-potential start-up investments as venture capital firms. But despite their
rapid growth, we know very little about the role that angels play internationally and the type of firms in which they invest.

 As Lerner and Schoar describe them, angel investors are \”a growing form of start-up investing that is less formal than the VC market but more professional than receiving funding from friends and family.\”  Venture capitalists often have a few seats on a company\’s board of directors, but angel investors are more likely to be personal mentors to entrepreneurs who are starting a company and to play a fairly direct role in using their connections and experience to help the company grow. Their average investment in a company is fairly small, measured in  hundreds of thousands of dollars, not millions. Here\’s a description of the process for a typical angel group:

Angel groups follow mostly similar templates. Entrepreneurs typically begin the process by submitting to the group an application that may also include a copy of their business plan or executive summary. The firms, after an initial screening by the staff, are then invited to give a short presentation to a small group of members, followed by a question-and-answer session. Promising companies are then invited to present at a monthly meeting (often a breakfast or dinner). The presenting companies that generate the greatest interest then enter a due diligence review process by a smaller group of angel members, although the extent to which due diligence and screening leads or follows the formal presentation varies across groups. If all goes well, this process results in an investment one to three months after the presentation.

The authors describe some of their recent research on angel investors. For one study, they got detailed data on two groups of angel investors, so they could compare companies that just barely made the cut to receive funding with companies that just barely missed the cut–and thus firms whose prospects  looked quite similar–the firms that got funding were measurably (although not extremely) more successful over the next few years. For another study, they looked at \”the records of 13 angel investment groups based in 12 nations and with applicants for financing transactions from 21 nations,\” and again find that firms which just made the cut for funding do better over time. 

The health care costs that Americans pay out-of-pocket are rising–both in total amount, which is perhaps not a surprise, but also as a share of incomes, which is perhaps more disturbing. Ann Foster discusses the pattern in \”Household healthcare spending in 2014,\”  written as a \”Beyond the Numbers\” short essay for the US Bureau of Labor Statistics (August 2016, vol. 5, no. 13). She has lots of details of spending in particular areas, but here\’s an overall figure showing out-of-pocket spending by households as a total amount (left axis) and as share of total expenditures (right-hand axis).

Part of the reason for this rise is that more Americans have health care insurance with high deductibles: that is, the amount you pay out of pocket before the insurance kicks in. Here\’s some information from the most recent Kaiser Family Foundation 2016 Employer Health Benefits Survey.  The report notes:

\”[T]he share of covered workers in plans with a general annual deductible has increased significantly over time: from 55% in 2006, to 74% in 2011, to 83% in 2016, as have the average deductible amounts for covered workers in plans with deductibles: from $584 in 2006, to $991 in 2011, to $1,478 in 2016.\” 

Another simpler way to describe this information is just to look at what share of workers now have a deductible above a certain amount, like $1,000.  This share has been rising for firrms of all sizes, but especially for smaller firms.

I\’ll add that I\’m not automatically opposed to  higher out-of-pocket spending for certain kinds of health costs. One way of holding down the rise in  health care costs is for patients and health care providers to be more sensitive to the cost implications of their choices. Most kinds of insurance against certain costs or damages–auto insurance, homeowner\’s insurance, and so on–have some form of out-of-pocket cost sharing for this reason. In the big picture, being protected against the risk of health care costs in the tens of thousands of dollars, or more, is quite important, even if it means you aren\’t protected against annual health care costs in the range of a few hundred or even several thousand dollars.

But it\’s also wise to remember that the relatively modest rise in out-of-pocket health care costs above is a mixture of households with relatively little in such costs in most years, and other families that are consistently paying the full deductible–an amount which is rising fast–every year. High and rising deductibles will make a lot of people feel as if their health care doesn\’t offer them much benefit in a typical year.

US housing prices have recovered so that they are more-or-less back to the long-run trend that existed before the price bubble of the early 2000s. The total amount of equity that US households have in the form of housing has largely recovered too. Here, with the help of the FRED website run by the Federal Reserve Bank of St. Louis, is a quick overview of the recent roller coaster ride for US housing markets–which leads to the case for why state and local government should be tweaking their priorities to encourage more homebuilding.

Here\’s the bounceback in housing prices since about 2012. There are rises and falls over time, but the current value isn\’t too far from the trend over the last four decades.

The amount of equity that homeowners have in real estate peaked at about $13 trillion at the top of the housing bubble in 2006, then dropped by $7 trillion by 2009 (!), and since 2012 has rebounded back up to almost $13 trillion again.

Another perspective is to look at homeowners\’ equity in housing divided by GDP, which adjusts for changes in inflation and in the size of the real economy over time. With this adjutment, you can see that homeowner equity was mostly in the range of 50-70% of GDP for the second half the 20th century. The housing bubble pushed homeowner equity up to nearly 100% of GDP in 2006, and the bursting of the bubble dropped it to 40% 2009.

Yet another way to slice this data on homeowner\’s equity is to look at it relative to the total value of the real estate. For example, though the 1970s and into the early 1980s, it was common that about 70% of the total value of real estate was equity for homeowners–while the rest was presumably mortgage debt owed to lenders. But even with the rise in housing prices in the 1990s and early 2000s, the share of that value which was homeowner equity was falling, as homebuyers borrowed more aggressively. when the housing bubble burst and the value of homes dropped, the size of the accumulated borrowing remained. The share of housing value that was home equity dropped almost to 35%, which would have been half the level prevailing back in the 1970s and early 1980s. The share of real estate that is homeowner equity has now bounced back, but is still under 60%.

The movements in housign prices also affect home construction. Here\’s the pattern of new housing starts since the late 1950s. These figures are measured in units of housing starts–that is, not relative to the size of the US population, which was about 180 million back in 1960 compared with 320 million today. The data on housing starts is spiky, but the long run-up for the construction industry from about 1990-2005 is clear enough, as is the dramatic fall in construction after that to the lowest levels seen in this data. . There\’s been a bounceback, but in historical context (and given the larger US population and smaller family sizes), it seems rather modest.

The recent bounce-back in housing prices seems to me different from the rise during the bubble. During the housing bubble in the early 2000s, there was a large amount of building (as shown by housing starts), which was sitting vacant (as shown by the vacancy rate), and financed by borrowed money (as shown by the low level of homeowners\’ equity relative to real estate values at a time when prices were rising). The recent bounce-back in housing prices is happening at a time when construction levels remain relatively low, vacancy rates are at historically common levels, the total value of housing relative to GDP is at historically common levels.

Thus, my sense is that the bounce-back in housing prices isn\’t about a borrowing-driven boom in demand, but is more about limitations on housing supply. Those limitations are not typically a matter of federal policy, but are imposed by state and local governments, and by neighborhood activists and courts, whose efforts have the effect of discouraging construction and keeping housing prices higher than they would otherwise be. Jason Furman, the head of President Obama\’s Council of Economic Advisers since 2013, has forthrightly made this point. Last November, for example, he gave a talk called \”Barriers to Shared Growth: The Case of Land Use Regulation and Economic RentsRemarks,\” where he said: \”I will focus on how excessive or unnecessary land use or zoning regulations have consequences that go beyond the housing market to impede mobility and thus contribute to rising inequality and declining productivity growth.\”

Similarly, the 2016 Economic Report of the President, published last January by the CEA, had this to say (pp. 87-89):

\”Supply constraints provide a structural challenge in the housing market, particularly in high-mobility, economically vibrant cities. When housing supply is constrained, it has less room to expand when demand increases, leading to higher prices and lower affordability. Limits on new construction can, in turn, impede growth in local labor markets and restrain aggregate output growth. Some constraints on the supply of housing come from geography, while others are man-made. Constraints due to land-use regulations, such as minimum lot size requirements, height restrictions, and ordinances prohibiting multifamily housing, fall into the man-made category and thus could be amended to support more inclusive growth. While these regulations can sometimes serve legitimate purposes such as the protection of human health and safety and the prevention of environmental degradation, land-use regulations can also be used to protect vested interests in housing markets. …

\”In addition to housing affordability, these regulations have a range of impacts on the economy, more broadly. Reduced housing affordability—whether as an ancillary result of regulation or by design—prevents individuals from moving to high productivity areas. …  Ensuring that zoning and other constraints do not prevent housing supply from growing in high productivity areas will be an important objective of Federal as well as State and local policymakers.\”

Policymakers and economists have been saying for 10-15 years that China\’s economy needs to shift away from being driven by extraordinarily high levels of investment, together with growing exports, and instead shift to being driven by growth of China\’s own domestic consumption. Maybe it\’s actually starting to happen. Jun Nie and Andrew Palmer discuss some recent evidence in \”Consumer Spending in China: The Past and the Future,\” published in the Economic Review of the Federal Reserve Bank of Kansas City (Third Quarter 2016, pp. 25-49). They begin this way:

\”After steadily declining for nearly half a century, the share of consumer spending in China’s GDP has recently increased. Economists and policymakers widely agree that the share of consumer spending must increase for China to continue its economic development; sustaining growth primarily on exports and investment will become more difficult in the longer run. … Although the recent rise in the consumption share has allayed some concerns about slowing growth in China, it has also spurred discussion over whether this trend is sustainable and whether China will truly become a consumption-driven economy.\”

To put China\’s low consumption patterns in perspective, it\’s useful to remember that in the US, about 65% of GDP is usually accounted for by consumption. In China, consumption was only about half of GDP in 1970s and 1980s–and then the share dropped even lower, falling below 40% of GDP. What\’s happening here is that China\’s consumption was growing, but GDP was growing even faster, so the ratio of consumption/GDP was falling, right up until about 2010.

Nie and Palmer discuss the evidence from economic research on what caused China\’s consumption to fall so low, and whether the modest rise of the last few years seems likely to continue. They focus on a few factors that have been evolving over the decades: the age distribution and \”dependency ratios\” in China\’s population, the trend to urbanization, and China\’s housing boom since around 2000–which is about when the consumption/GDP ratio plummeted even lower. I\’d summarize their story this way.

The \”dependency ratio\” refers to what share of the population is either too young or too old to work, as opposed to being working age. The dependency ratio and consumption tend to rise and fall together. Thus, the dramatic fall in birthrates in China as a result of economic growth (and greater education and incomes for women) along with the one-child policy meant a lower dependency ratio. Basically, fewer children means less consumption spending on children, and more saving.

China has seen a dramatic rise in urbanization. As Nie and Palmer write: \”Over the past five decades, the share of China’s population living in urban areas has more than tripled, rising from 18 percent in 1960 to 56 percent in 2015.\” The common pattern is that people moving to urban areas in emergin economies increase both their production and their consumption, but they increase production by more, and so their rate of saving rises.

Nie and Palmer look at patterns of dependency ratios and urbanization across a \”sample of 24
countries including most Asian countries and large developing countries.\” Based on the common patterns, they find that China\’s fall in consumption/GDP ratio from 1970-2000 is almost completely explained by changes in its dependency ratio and urbanization. But the additional drop in consumption/GDP after 2000 is not explained by these patterns, and they instead attribute this fall to a shift in China\’s housing market. Here\’s their explanation (citations and references to charts omitted):

\”The large jump in the household saving rate from 2000 to 2010 is largely related to development in China’s housing market during this period. Before 1998, most Chinese families lived in government-provided houses; after economic reforms in 1998 removed this benefit, however, most Chinese families needed to buy their own homes. This change triggered rapid growth in the Chinese real estate sector, causing home prices to rise tremendously. Furthermore, as house prices started to increase quickly, housing became a popular investment for wealthy Chinese households, raising demand even further and exacerbating house price increases. Indeed, from 2000 to 2010, house prices
in China increased by about 161 percent. In addition to rapidly increasing prices, Chinese homebuyers faced large required down payments—typically 30–40 percent, but in some major cities as high as 50 percent. As a result, middle-class Chinese households were forced to save a disproportionately large share of their earnings to purchase a home. The saving didn’t end with the purchase of a home—the subsequent mortgage payments constituted additional saving as they increased the homeowners’ housing equity. As Rosenzweig and Zhang argue, this housing market dynamic helps explain the rising saving rate associated with rising housing prices from 1998 to 2010.

\”As the scramble to buy homes after the 1998 reform faded, so, too, did the desire to save for homes. Moreover, young people currently do not need to save as aggressively as the previous generation for home purchases because their parents and grandparents often help them buy a home. Due to the one-child policy, which was introduced in 1979, a typical young couple in China are the sole descendants of four parents and eight grandparents. In China, it is common for parents and grandparents to help their children by paying the high down payment or even paying for the entire house. Small family sizes under the one-child policy allow families to focus their resources rather than spreading them among multiple children or extended family members.

\”Large financial contributions from older generations are feasible for two additional reasons. First, after aggressively saving for several decades, many parents and grandparents are wealthy enough to afford this gift to the younger generation. Second, many households own multiple homes. Based on the 2014 China Household Finance Survey, one out of five urban households owns a second home, meaning that on average, a young couple’s parents and grandparents have one extra home between them in addition to their primary residences. Overall, less pressure for young Chinese couples to save helps explain why the Chinese saving rate has started to decline since 2010.\”

What do these factors imply for the future? China\’s old-age population is beginning to rise, and in October 2015 China\’s government announced that it was shifting to a \”two-child policy.\” China\’s dependency ratio seems set to rise, which should cause consumption to continue rising. Urbanization in China also seems likely to continue, which has tended to hold down consumption in China. The original huge boost in saving from the shift to private housing market has largely run its course (although whether the ongoing boom in China\’s housing prices is sustainable is different issue). Mixing this together and stirring a few times, Nie and Palmer suggest that China\’s consumption/GDP ratio will keep rising in the next few years, perhaps to between 45-50% by 2020–that is, to say, almost back to levels prevailing in the 1980s.

When economists and policymakers talk about \”rebalancing\” China\’s economy toward faster growth of domestic consumption, I think they have a bigger shift in mind. Nie and Palmer describe the changes in China\’s household spending this way:

\”[R]eal [household] spending on transportation and communication increased sevenfold over the last two decades compared with a threefold increase in overall household spending. Transportation and communication is a broad category that encompasses many high-end goods and services that have recently become available to middle-class Chinese households, such as smartphones, laptops, and air travel. Travel services in particular appear to be growing quickly: the number of Chinese outbound travelers has been growing at a double-digit pace for many years, and 2015 marked the fourth consecutive year of China as the world’s top tourism source market … [A]nother component of consumption, educational services, is also growing rapidly. Chinese families now are spending more resources on their child’s education both inside and outside the classroom. In addition to paying tuition both domestically and at foreign colleges and universities, Chinese families are increasingly paying for tutoring and other educational supplements for their children.\”

If China\’s government wishes to boost consumption, it could do more to encourage spending in areas like education and health care, as well as to support consumption and health services for its growing population of elderly.

In the long-ago days of the 20th century, it was difficult to find out about small or medium companies that were far away: hard to discover they existed, hard to get details about their products and pricing, hard to place an order, and especially if the firm was in another country, hard to pay for an order and track shipping. That\’s all becoming easier, which suggests that small and medium enterprises may be able to draw on wider markets, and not be so focused on local markets. Thus, the World Trade Organization in its World Trade Report 2016 focuses on the theme of \”Leveling the Playing Field for SMEs\”–that is, \”small and medium enterprises.\”

Much of the report focuses on practical issues affecting small and medium enterprises: connectedness, international trade rules, finance and others. Here, I\’ll just point to two patterns that I found intriguing. Here\’s a figure showing the share of exports and imports attributable to small and medium enterprises, which are defined as as employing 10-250 people (the blue bars), and also micro-enterprises (the gray dots), which are defined as employing 0-9 people.

Unsurprisingly, the developed economies where small and medium enterprises have the largest share of their trade tend to be smaller European economies, like Estonia, Cyprus, and Ireland. It\’s interesting that Italy, which is the fourth-largest economy in Europe, gets more than half its exports from micro, small and medium enterprises. It does jump out at me that the US economy is by far the lowest for small and medium enterprises as a share of its trade. Traditionally, small and medium US firms focused on the enormous US market.

The pattern is different if you look at not the share of exports from small and medium firms, but instead at the share of exporting firms that are small and medium firms. In the US, for example, the micro, small, and medium firms make up about 26-28% of the value of trade. But of the total number of US firms involved in international trade, micro, small, and medium firms make up more than 90%. Here\’s the figure:

It will be interesting to see how these patterns evolve over time. It seems as if the division of production across national borders into global value chains and the ability to export and import services, not just goods, should mean a larger role for small and medium enterprises in global trade. Or course, it will cause some of them to shift categories and turn into larger companies, as well.  

\”Shadow banking\” refers to financial organizations that in various ways receive funds from savers and lend money in financial markets–but are not banks. For example, a money market mutual fund receives money from investors, who can be thought of as \”depositors,\” and then invest the money in bonds, which can be thought of as lending the money to whatever government or private entity issued the bonds. But it\’s not a bank! Alex Muscatov and Michael Perez of the Federal Reserve Bank of Dallas offer a nice quick overview of this sector  in \”Shadow Banking Reemerges, Posing Challenges to Banks and Regulators\” (Economic Letter, July 2016).

They offer a helpful table of sic types of nonbank financial entities: retirement funds, mutual funds, broker-dealers, alternative investment funds, financing firms, and insurance companies. Here\’s a chart with a short description of each category and how it can be connected to traditional banks.

Our national conversation about financial regulation is often focused on banks, but when you think back to the big meltdowns of financial firms that stressed the economy back in 2008, there are a lot of non-bank financial companies like Lehman Brothers, Bear Stearns, the insurance company AIG, the mutual fund Reserve Primary Fund, and others. Indeed, the importance of shadow banking is growing. Muscatov and Perez write:

The  NBI’s [non-bank intermediaries] importance has increased over the past four decades (Chart 1). In 1980, it accounted for roughly 40 percent of the domestic financial sector. As mutual funds’ prominence increased and life insurance companies fueled the markets for corporate bonds and commercial real estate, NBI growth greatly outpaced that of banks. By 1990, NBI accounted for two-thirds of the intermediation market and has continued to slowly gain share.

Here\’s the Chart 1 to which they were referring. The red dashed line is the total liabilities of traditional banks. The shaded areas show the shadow banking or \”nonbank intermediation\” firms. Notice that before the financial crisis in 2008, liabilities of banks don\’t soar; after the crisis, they don\’t fall. The financial crisis instead happened in the shadow banking sector, where you can see the sharp rise in liabilities before the crisis circa 2008 and the sharp fall afterwards.

The basic lesson here is that if you still think banks are the core representative institutions in the financial system of high-income economies, you are a few decades out of date. If you are concerned about the dangers of financial sector risks cartwheeling into the real economy, you need to think about the shadow banking sector. Muscatov and Perez point out that while banking regulators do try to think about risks from the shadow banking sector, \”Still, many areas of NBI remain obscured from regulators’ view, and not all NBI is subject to supervision.\” For a specific warning call, the Global Financial Stability Report published by the IMF in April 2016 includes a chapter on \”The Insurance Sector — Trends and Systemic Risk Implications.\”  The report says:

\”The chapter shows that across advanced economies the contribution of life insurers to systemic risk has increased in recent years, although it clearly remains below that of banks. This increase is largely due to growing common exposures to aggregate risk, caused partly by a rise in insurers’ interest rate sensitivity. Thus, in the event of an adverse shock, insurers are unlikely to fulfill their role as financial intermediaries precisely when other parts of the financial system are failing to do so as well.\”  

For those interested in shadow banking, here are a few previous posts on the subject:

• \”A Shadow Banking Schematic\” (May 27, 2014)
• \”Shadow Banking: US Risks Persist\” (October 8, 2014)
• \”Credit Without Banks: Shadow Banking\” (November 6, 2014)

A US presidential election is, as the political science geeks like to point out, really 50 separate state elections, each one in a winner-take-all format. With the US national election day less than two months away, it seemed like a good time to review some (occasionally controversial) differences across how these state elections are conducted, which are compiled by the National Council of State Legislatures and available on its website.

Voter ID

The first five states to require voter ID were South Carolina (in 1950), Hawaii (1970), Texas (1971), Florida (1977) and Alaska (1980).  But during the last 15 years or so, the trend is clearly toward a a rising number of states imposing such requirements.

The NCSL makes a couple of key distinctions between these laws. One is whether the requirement is for a photo ID (like a driver\’s license) or a non-photo ID (like a bank statement). Another distinction is whether the law is \”non-strict\” or \”strict.\” Pretty much all voter ID laws offer some ways in which people without an ID can cast a provisional ballot, but the difference is whether that ballot will then be counted without further action by the voter (\”non-strict\”), or whether the person casting the ballot needs to do something after election day, like return to an election office and show a valid ID, before their ballot will be counted (\”strict\”). A state-by-state list of voter ID rules is available here at the NCSL website.

Absentee and Early Voting

States do offer some options for those who can\’t or don\’t want to vote on election day, but these options vary considerably. The NCSL summarizes in this way:

1. Early Voting: In 37 states (including 3 that mail ballots to all voters) and the District of Columbia, any qualified voter may cast a ballot in person during a designated period prior to Election Day. No excuse or justification is required. … 
2. Absentee Voting: All states will mail an absentee ballot to certain voters who request one. The voter may return the ballot by mail or in person. In 20 states, an excuse is required, while 27 states and the District of Columbia permit any qualified voter to vote absentee without offering an excuse. Some states offer a permanent absentee ballot list: once a voter asks to be added to the list, s/he will automatically receive an absentee ballot for all future elections.
3. Mail Voting: A ballot is automatically mailed to every eligible voter (no request or application is necessary). In-person voting sites may also be available for voters who would like to vote in-person and to provide additional services to voters. Three states mail ballots to all eligible voters for every election. Other states may provide this option for some types of elections.

The specific rules for early and absentee voting vary considerably across states, as well. Fopr example, the average starting time for early voting is 22 days before an election, but it varies across states from as much as 45 days before to just four days before. The average early voting period average is 19 days, but again, this varies across states from four to 45 days. Some states require that you give an approved reason for requesting an absentee ballot, while other will give an abstentee ballot to anyone who requests one. A state-by-state list of these rules is available here.

Same Day or Online Registration

\”Thirteen states plus the District of Columbia presently offer same-day registration (SDR), allowing any qualified resident of the state to go to the polls or an election official\’s office, either before or on Election Day, register to vote, then cast a ballot, all in that day. California, Hawaii and Vermont have enacted same-day registration but have not yet implemented it.  In most other states, voters must register by a deadline prior to Election Day. The deadline varies by state, with most falling between eight and 30 days before the election. …\”

\”As of June 14, 2016 a total of 31 states plus the District of Columbia offer online registration, and another seven states have passed legislation to create online voter registration systems, but have not yet implemented them.\”

Rules about Recounts

\”43 states permit a losing candidate, a voter, a group of voters or other concerned parties to petition for a recount. In a few states, the vote totals for the top two candidates must be within a specified margin in order for the losing candidate to be able to request a recount. For example, in Idaho, a candidate may petition for a recount if the difference between the requesting and winning candidates is less than 0.1 percent of the total votes cast for the office. In at least five states, a political party officer can request a recount, and in at least 17 states, a voter can petition for a recount. … In most of the states that permit a candidate or other interested party to demand a recount, the petitioner is required to pay a deposit toward the cost of conducting the recount. If the recount reverses the result of the election, that person’s deposit is refunded. If the recount does not change election results, the petitioner is required to pay most of the costs associated with the recount. Automatic recounts are paid for by the state or county that conducts the recount.\”

There are many other differences in how states conduct elections: for example, differences in voter registration, how lists of eligible voters are maintained, the rules about primary elections, the number of polling stations on election day and the hours they are open, the qualifications for poll-watchers, the ways in which paper or electronic voting equipment is used, and more.

It\’s useful to consider this wide variation across states, in part because for many of us it tends to challenge our preconceptions about how voting should happen.  For example, it seems to me that voting by mail has the potentially important problem that many voters will find it harder to cast a truly secret ballot. But clearly residents of Colorado, Oregon, and Washington disagree. I\’m not a big fan of early voting rules, because I think there\’s some value to people having a chance to change their minds right up to election day, rather than being pressured to lock in their vote early, but many states clearly disagree with this view. My home state of Minnesota tends to pat itself on the back for not having a voter ID law, but it\’s a useful exercise in humility to remember that most other states disagree on the merits of such a rule.

In some ways, the differences across states in voting are a litmus test for how you feel about a federalist country in which state and local governments have a substantial degree of autonomy on many issued, including administration of elections, or whether you tend to favor greater control by the central government.

Every now and again, I\’ll post a cartoon suitable for tacking up on a bulletin board, or blending into an economics lecture. This one is from back in 2008, from Dale Everett at the Anarchy In Your Head website, but I just saw it for the first time, so I pass it along.

Addendum: A helpful reader points out that a similar comment was attributed to Ronald Coase by David Landes, who said in a 1983 symposium published in the Journal of Law and Economics: \”Ronald said he had gotten tired of antitrust because when the prices went up the judges said it was monopoly, when the prices went down, they said it was predatory pricing, and when they stayed the same, they said it was tacit collusion.\”

The citation is Edmund W. Kitch, \”The Fire of Truth: A Remembrance of Law and Economics at Chicago, 1932-1970,\” Journal of Law & Economics, Vol. 26, No. 1 (Apr., 1983), pp. 163-234. Quotation is on p. 193.

\”More than 40 million people living in the United States were born in other countries, and almost an equal number have at least one foreign-born parent. Together, the first generation (foreign-born) and second generation (children of the foreign-born) comprise almost one in four Americans. It comes as little surprise, then, that many U.S. residents view immigration as a major policy issue facing the nation. Not only does immigration affect the environment in which everyone lives, learns, and works, but it also interacts with nearly every policy area of concern, from jobs and the economy, education, and health care, to federal, state, and local government budgets.\”

That\’s the beginning of the just-released 500+ page report from the National Academy of Sciences on \”The Economic and Fiscal Consequences of Immigration,\” edited by Francine D. Blau and Christopher Mackie. A prepublication copy of the report (essentially, uncorrected proofs) can be downloaded for free here. The conventional approach, followed in this report, is to divide up the effects of immigration into two areas; effects on native jobs and wages, and effects on government budgets and services. But before getting to those issues, I found some of the basic findings about immigration over the last couple of decades to be intriguing. Quoting from the summary:

• The number of immigrants living in the United States increased by more than 70 percent—from 24.5 million (about 9 percent of the population) in 1995 to 42.3 million (about 13 percent of the population) in 2014; the native-born population increased by about 20 percent during the same period.
• Annual flows of lawful permanent residents have increased. During the 1980s, just under 600,000 immigrants were admitted legally (received green cards) each year; after the 1990 Immigration Act took effect, legal admissions increased to just under 800,000 per year; since 2001, legal admissions have averaged just over 1 million per year.
• Estimates of the number of unauthorized immigrants in the United States roughly doubled from about 5.7 million in 1995 to about 11.1 million in 2014. Gross inflows, which had reached more than 800,000 annually by the first 5 years of the 21st century, decreased dramatically after 2007; partly as a result, the unauthorized immigrant population shrank by about 1 million over the next 2 years. Since 2009, the unauthorized immigrant population has remained essentially constant, with 300,000-400,000 new unauthorized immigrants arriving each year and about the same number leaving.
• The foreign-born population has changed from being relatively old to being relatively young. In 1970 the peak concentration of immigrants was in their 60s; in 2012 the peak was in their 40s.
• Educational attainment has increased steadily over recent decades for both recent immigrants and natives, although the former still have about 0.8 years less of schooling on average than do the latter. Such averages, however, obscure that the foreign born are overrepresented both among those with less than a high school education and among those with more than a 4-year college education, particularly among computer, science, and engineering workers with advanced degrees. The foreign and native born populations have roughly the same share of college graduates.
• As time spent in the United States lengthens, immigrants’ wages increase relative to those of natives and the initial wage gap narrows. However, this process of economic integration appears to have slowed somewhat in recent decades; the rate of relative wage growth and English language acquisition among the foreign-born is now slightly slower than it was for earlier immigrant waves. The children of immigrants continue to pick up English language skills very quickly.
• Geographic settlement patterns have changed since the 1990s, with immigrants increasingly moving to states and communities that historically had few immigrants. Nonetheless, the majority of the foreign-born population continues to reside in large metropolitan centers in traditional gateway states.

The conventional wisdom on the economic effects of immigration is that the effect on jobs is minimal. The number of jobs in a developed economy expands over time as the population expands–whether the growth in population is from native-born workers or from immigration.  Unemployment rates rise and fall with recessions and upswings, but there is no long-term trend to higher unemployment rates over time. On how immigrants affect the total number of jobs for native workers, the NAS report puts it this way:

\”The literature on employment impacts finds little evidence that immigration significantly affects the overall employment levels of native-born workers. However, recent research finds that immigration reduces the number of hours worked by native teens (but not their employment rate).\”

However, immigration can potentially have an effect on the distribution of wages, potentially substituting for some kinds of native workers and thus holding down their wages, but also potentially complementing other native workers and leading to higher wages for them. The wage effects of immigration is a tough topic for research. For example, imagine that immigrants tend to go to areas where wages are higher and jobs are more plentiful. If this plausible assumption holds true, then there will be a positive correlation in which areas with more immigrants will also tend to have better jobs and wages, but that correlation certainly doesn\’t prove causality. In addition, low-skilled and high-skilled immigrants will be substitutes and complements for different kinds of workers. In some places and jobs, immigrants may even be competing more with earlier immigrants, rather than with native workers.  Given these complexities, on how immigrants affect wages for native workers, the NAS report is necessarily more equivocal:

\”When measured over a period of 10 years or more, the impact of immigration on the wages of natives overall is very small. However, estimates for subgroups span a comparatively wider range, indicating a revised and somewhat more detailed understanding of the wage impact of immigration since the 1990s. To the extent that negative wage effects are found, prior immigrants—who are often the closest substitutes for new immigrants—are most likely to experience them, followed by native-born high school dropouts, who share job qualifications similar to the large share of low-skilled workers among immigrants to the United States. …

\”Until recently, the impact of high-skilled immigrants on native wages and employment received less attention than that of their low-skilled counterparts. Interest in studying high-skill groups has gained momentum as the H1-B and other visa programs have contributed to a rapid rise in the inflow of professional foreign-born workers (about a quarter of a million persons per year during the last decade). Several studies have found a positive impact of skilled immigration on the wages and employment of both college-educated and noncollege educated natives. Such findings are consistent with the view that skilled immigrants are often complementary to native-born workers, especially those who are skilled; that spillovers of wage-enhancing knowledge and skills occur as a result of interactions among workers; and that skilled immigrants innovate sufficiently to raise overall productivity. However, other studies examining the earnings or productivity prevailing in narrowly defined fields find that high-skill immigration can have adverse effects on the wages or productivity of natives working in those fields.\”

The NAS report also notes some other economic effects of immigration:

\”The contributions of immigrants to the labor force reduce the prices of some goods and services, which benefits consumers in a range of sectors including child care, food preparation, house cleaning and repair, and construction. Moreover, new arrivals and their descendants are a source of demand in key sectors such as housing, which benefits residential real estate markets. … Importantly, immigration is integral to the nation’s economic growth. Immigration supplies workers who have helped the United States avoid the problems facing stagnant economies created by unfavorable demographics—in particular, an aging (and, in the case of Japan, a shrinking) workforce. Moreover, the infusion by high-skill immigration of human capital has boosted the nation’s capacity for innovation, entrepreneurship, and technological change.\”

On the issue of how immigration affects government budgets and services, the research takes a variety of perspectives. For example, a single immigrant with a job, no children and law-abiding, tends to pay more in taxes (including sales taxes and income tax withholding) than consumed in government services. A high-skilled immigrant will earn more income and pay higher taxes than a low-skilled immigrant. An immigrant with children in public schools will consume more services. An low-skilled immigration with a lower income level who works long enough to be eligible for Social Security and Medicare will consume more in services. In thinking about how immigration affects government budgets and services, it makes a difference if one takes a short-run perspective of a year, or the typically accumulation of taxes paid and government services over a lifetime.  These lifetime patterns will vary among first-, second-, and third-generation immigrants. Thinking about the costs of government services means that you need to think of immigrants as consuming a share of publicly provide goods like, say, national defense.

With such complexities duly noted, the NAS report lays out some overall conclusions. For example, a standard finding is that over a lifetime, immigration has a positive fiscal effect for the federal government but a negative effect for state and local government.

\”Viewed over a long time horizon (75 years in our estimates), the fiscal impacts of immigrants are generally positive at the federal level and negative at the state and local levels. State and local governments bear the burden of providing education benefits to young immigrants and to the children of immigrants, but their methods of taxation recoup relatively little of the later contributions from the resulting educated taxpayers. Federal benefits, in contrast, are largely provided to the elderly, so the relative youthfulness of arriving immigrants means that they tend to be beneficial to federal finances in the short term. In addition, federal taxes are more strongly progressive, drawing more contributions from the most highly educated. The panel’s historical analysis indicates that inequality between levels of government in the fiscal gains or losses associated with immigration appears to have widened since 1994. The fact that states bear much of the fiscal burden of immigration may incentivize state-level policies to exclude immigrants and raises questions of equity between the federal government and states. …

\”For the 2011-2013 period, the net cost to state and local budgets of first generation adults (including those generated by their dependent children) is, on average, about $1,600 each. In contrast, second and third-plus generation adults (again, with the costs of their dependents rolled in) create a net positive of about $1,700 and $1,300 each, respectively, to state and local budgets. These estimates imply that the total annual fiscal impact of first generation adults and their dependents, averaged across 2011-13, is a cost of $57.4 billion, while second and third-plus generation adults create a benefit of $30.5 billion and $223.8 billion, respectively. By the second generation, descendants of immigrants are a net positive for the states as a whole, in large part because they have fewer children on average than do first generation adults and contribute more in tax revenues than they cost in terms of program expenditures.\”

A different way to slice this data is to look compare first-, second-, and third-generation immigrants at different ages.

\”Cross-sectional data from 1994-2013 reveal that, at any given age, the net fiscal contribution of adults in the first generation (and not including costs or benefits generated by their dependents) was on average consistently less favorable than that of the second and third-plus generations. Relative to the native-born, the foreign-born contributed less in taxes during working ages because they earned less. However, this pattern reverses at around age 60, beyond which the third-plus generation has
consistently been more expensive to government on a per capita basis than either the first or second generation; this is attributable to the third-plus generation’s greater use of social security benefits.\”

Another finding is that because immigrant over the last few decades have become better-educated, their fiscal effect has also improved.

\”Today’s immigrants have more education than earlier immigrants and, as a result, are more positive contributors to government finances. …  Whether this education trend will continue remains uncertain, but the historical record suggests that the total net fiscal impact of immigrants across all levels of government has become more positive over time.\”

My overall sense is that immigration is overall a positive force for the US economy, and I support a allowing steady stream of legal immigration over time–especially high-skilled migrants who have already spent time in the United States being trained at US colleges and universities, or working in US-based firms. But I would also say that the positive economic effects of immigration are not enormous, and can be negative for certain subgroups. Moreover, I suspect that while our social controversies over immigration may often be phrased in economic terms, a lot of the heat and energy in these controversies arises from perceptions about non-economic consequences of immigration.