MIT analyst Jesse Jenkins, among others, has called for a “tax and invest” strategy to combat global warming: A modest tax on carbon emissions would be invested in government research, development, and innovation-promotion programs to commercialize new alternative energy technology.
But there are at least four reasons to question whether typical government technology programs would be as cost-effective an investment as Jenkins and others claim:
- The government role in innovation is neither necessary nor sufficient.
- Government research and innovation activities often follow rather than lead private initiatives.
- The failures of government R&D and innovation efforts are not innocuous or harmless.
- Nothing fails like success.
Better alternatives to fossil fuels are needed to solve climate and other energy-related problems. But the prospects for substantial increases in government funds are poor. Instead, the emphasis going forward should be on getting more and better innovation, faster, out of the resources that are already available. The focus should shift toward open science, open innovation, and greater engagement of non-governmental organizations and individual talent within both rich and poor nations.
INTRODUCTION: Tax and Invest?
In his incisive critique of various schemes to put a price on carbon emissions* to deter global warming, MIT analyst Jesse Jenkins concludes that the most practical arrangement would be what he dubbed “tax and invest.” Under that plan, revenues from a very modest federal tax on emissions — equivalent to a few cents per gallon of gasoline — would be invested in a government program for research, development, and innovation aimed at creating cost-effective alternatives to fossil fuels. (Jenkins estimates that a tax of $5 per ton of carbon would yield about $25 billion in annual revenue.)
Let me emphasize at the outset here that I agree with Jenkins — and others who have presented similar arguments — that accelerating technology innovation is the most effective and politically feasible way to manage not only climate concerns but a thicket of other economic, environmental, and social problems that entangle our existing, largely fossil-fuel-dependent energy systems.
But more caution about the government role in innovation is warranted than what Jenkins and some others suggest.
If perchance a tax and invest program of the sort Jenkins prescribes were enacted — and the prospects seem doubtful — some good probably would come of it. But it also would likely be far more costly, with more confounding side effects, and with less useful results than he and other enthusiasts claim, or at least imply.
Economists broadly have agreed for a long time that the private sector tends to under-invest in research and development (R&D) because a free market does not adequately reward discoverers, inventors, or innovators for the full economic value of their creations. Patents and copyrights were created to help redress some of this imbalance. Public support for pioneering research and exploration similarly has been justified. That’s not in dispute.
But Jenkins and others argue for a more invasive and ambitious government role than that, aimed at large-scale commercialization and dissemination of innovations.
In particular, in a comment on Jenkins’ essay in The Energy Collective, Joris van Dorp expressed doubt about the efficiency of the “tax and invest” approach: “I have some experience with government funded green R&D and am very disappointed. Such programs are riddled with political and ideological contamination and are extremely wasteful.”†
That prompted a fairly detailed rebuttal from Jenkins, including this:
First, you seem skeptical that government funded RD&D [research, development, and demonstration] programs can yield success and assert that “90% of money” is most likely to be wasted. I hope that a careful look at the history of American innovation and the role of government would temper your skepticism. I refer you to my 2010 collection of case studies, “Where Good Technologies Come From,” which documents how everything from jet engines and nuclear power to pharmaceuticals, advances in agrosciences, and of course, the semiconductor, Internet and computers all come more or less directly out of government-funded and, more often than not, government-led innovation initiatives. You may also wish to look at the role of government RD&D programs in developing most of the key technologies underlying the shale gas and oil revolutions now changing the North American energy landscape. In fact, it’s very hard to come up with an example of an important 20th or early 21st century invention that does not trace back at one or more critical junctures to government-funded RD&D programs. So I do not at all share [the] view that government R&D is a wasteful use of carbon revenues. Even if the “success rate” for project funding is 1 in 10 as you assert, that would be a remarkable impact on the energy landscape, and while you characterize that situation as “wasting” 9 out of 10 dollars, you also ignore the much larger societal returns that could be expected from those 1 in 10 successes. How do we quantify the returns on government investment in microchips? Or shale gas? Or jet engines?
FOUR REASONS FOR CAUTION
In what follows, I offer a few counterarguments to the premises of Jenkins’ and others’ confident case for big government R&D and innovation programs. (A more extensive critique can be found in Energy Innovation.)
1. The government role in innovation is neither necessary nor sufficient.
Jenkins’ assertion above makes it sound as if it is almost impossible for useful innovation to occur without government intervention: “…it’s very hard to come up with an example of an important 20th or early 21st century invention that does not trace back at one or more critical junctures to government-funded RD&D programs.”
Similarly, the Information Technology and Innovation Foundation claims, “…it is critical to remember that a consistent government role in supporting high-risk impact, early stage innovation is essential in reaching transformative breakthroughs.”
Some proponents also imply or outright argue that government intervention is sufficient — that is, very likely to assure successful commercialization and market adoption of favored innovations. For instance, this is from the official announcement from President Jimmy Carter of the launch of the Synthetic Fuels Corporation in 1980: “The United States Synthetic Fuels Corporation (SFC) was created by the Energy Security Act. It will be the principal Federal instrument to ensure that the synthetic fuels industry becomes a reality.” The dictionary definition of “ensure” is: “make certain that (something) shall occur….” What Carter and the authors of that initiative did not say was “we’re proposing to spend some billions of taxpayer dollars to see if, maybe, we can help a synfuels industry get off the ground.”
Neither claim, that government sponsorship is necessary or sufficient, is true. The second is easiest to dismiss. Most if not all government technology programs — in the United States certainly — are predicated on the correct assumption that commercialization mainly depends on private enterprise.
There are numerous examples of important innovations that were created and disseminated without government initiative or involvement. To name a few:
- The UNIX operating system was invented at AT&T’s Bell Labs. A variation, Linux, was later created by Linus Torvalds. A derivative of Linux, Android, runs over 70 percent of the world’s smartphones.
- The Salk polio vaccine was developed with money from the March of Dimes (an early example of crowdfunding).
- The Green Revolution grew out of research initiated and financed by the Rockefeller Foundation.
- Robert Goddard’s pioneering development of rocketry was paid for first by himself and then by private philanthropy. It was only after Goddard’s persistent effort that he finally got the War Department interested during World War I.
Of the 20 most important inventions of the 20th century listed by Time magazine, only four came from government initiatives or received significant government support.
The Breakthrough Institute report Jenkins coauthored and cites also understates the critical importance of entrepreneur George Mitchell’s role in making the recent shale gas boom happen. Jim Pierobon offered a fairer assessment:
At George Mitchell’s urging, Messrs. Steward, Steinberger, Bowker and Utley, with help from corporate vice president Mark Whitley, learned a lot about how a relatively simple mixture of sand and water directed by three-dimensional (3D) seismic test data could unleash large quantities of natural gas from shale, which is extremely tight and thus hard to penetrate. Their discovery of the formula was something that even the formidable resources of the U.S. Department of Energy and experts at the Gas Research Institute could not make sense of.
Just because a seemingly successful result followed some government activity, that does not preclude the possibility that the same result could and would have been achieved without any government involvement. Helpful does not necessarily mean essential.
Moreover, whatever government intervention may have accomplished in the past, it may well have less capacity to promote innovation today and tomorrow. Evidence of the diminishing ability of the government to innovate effectively is the progressive shift in Defense Department and other government acquisitions over the past two decades from GOTS, for government off-the-shelf, to COTS, commercial off-the-shelf. That is, it has proven more cost-effective to purchase commercial technology developed in the private sector than the traditional practice of buying products custom-designed to meet government specifications. (Even so, Bloomberg News recently reported that the Pentagon is paying many times the commercial market price for spare parts.)
The declining pace and cost-effectiveness of government-sponsored innovation is suggested by the development of weapon systems. A prototype of the P-51 Mustang, considered the premier fighter aircraft of the Second World War, was flown only 149 days after North American Aviation signed a contract with the British government in 1940. The first production aircraft were delivered to Britain the next year and to the U.S. Army Air Force in 1942. Over 15,000 were ultimately produced. Each airplane cost about $560,000 in today’s dollars.
In contrast, contracts for prototypes of the F-22, today’s best fighter aircraft, were awarded in 1986, to be delivered over four years later. The first production F-22 flew 11 years later. By then the Cold War was over, and changing economic and other conditions made the usefulness of the F-22 questionable. The Air Force originally planned to order 750 aircraft. After 1990, though, the Pentagon continually cut the number until it was recently down to about 180. In 2012, the Government Accountability Office estimated the cost of each aircraft at over $400 million.
The CIA founded venture capital firm In-Q-Tel in 1999 precisely because some officials in the agency realized that the private sector could generate useful technology innovations far faster and better than government-controlled programs could. In-Q-Tel supports the development of commercial technology useful to the intelligence community, but it is independent of government.
2. Government research and innovation activities often follow rather than lead private initiatives.
Proponents of government R&D and labs commonly list notable technology innovations they credit to government initiative (as in the “Good Technologies” report Jenkins mentions above). But it’s important to recognize that government R&D and innovation activities have often followed, were motivated by, reacted to, or copied earlier private initiatives — without which the government projects might not have been possible or ever happened. The Goddard case is just one of multiple examples.
In fact, scientists and inventors commonly had to lobby strenuously to get government interested in important innovations. Pride, turgid bureaucracy, the not-invented-here syndrome, dysfunctional political influences, or sheer failure of imagination often lead government to shut out promising innovations that are offered to it.
Government-sponsored attempts by respected physicist Samuel Langley to develop a viable flying machine did not pan out. Two unknown bicycle mechanics in Ohio, Wilbur and Orville Wright, succeeded where the government had failed. When the Wright brothers’ initial efforts to interest the U.S. government in their invention got a chilly reception, they took it to Europe where other governments paid keener attention.
J. Walter Christie invented a revolutionary tank design in the 1920s. After a series of frustrated attempts to sell his invention to the U.S. War Department, Christie (illegally) marketed it in Europe. The Soviet Union applied Christie’s innovations to the T-34 tank — considered by many analysts the best battle tank of the Second World War. Christie’s concepts were also adopted by the British.
3. The failures of government R&D and innovation efforts are not innocuous or harmless.
Jenkins, in his comment cited above, is right that accounting for all the impacts of government-led R&D efforts is complex and difficult: “How do we quantify the returns on government investment in microchips? Or shale gas? Or jet engines?” One way not to do such an accounting, though, is to tout collateral benefits while ignoring or downplaying costs and collateral damage.
The brief references to “waste” in the dialogue cited above between van Dorp and Jenkins may give a misimpression. The majority of government R&D or innovation projects and activities that fail to yield useful results do not merely represent resources “wasted,” though waste is an important issue. Unsuccessful outcomes of legitimate experiments or trials can yield useful information, and thus are not wasted efforts.
But there still is considerable waste in government-sponsored research, development, and deployment (RD&D) as well as other innovation-promotion activities. Projects are funded, grants disbursed, and so forth more for political or venal reasons than on merit.
Beyond waste, in the standard Washington trinity of government dysfunction there also are fraud and abuse to consider — as well as unintended negative consequences and counterproductive impacts.
Some of the most egregious abuses of government research include:
- Nuclear testing on human guinea pigs: “In 1986, Rep. Edward Markey (D-Mass.) released a report detailing human radiation experiments that [the Atomic Energy Commission] and its successors conducted between the 1940s and the 1970s. Many were designed to measure the effects of radiation on humans, and according to Markey, American citizens thus became nuclear calibration devices for experimenters run amok…. The Markey report, American Nuclear Guinea Pigs, described 31 grisly experiments involving 695 people who were captive audiences or populations that some experimenters frighteningly might have considered ‘expendable.’”
- The CIA’s MKUltra research program: “To observe the effects of the drug on unwitting subjects, they secretly administered LSD to hundreds of mental patients, prisoners, foreign nationals and private American citizens without their consent. … Many of the test subjects experienced adverse psychological reactions, and at least one — a biochemist named Frank Olson — died under mysterious circumstances only days after ingesting the drug.”
- The Tuskegee syphilis study: “For forty years, from 1932 to 1972, 399 African-American males were denied treatment for syphilis and deceived by officials of the United States Public Health Service. As part of a study conducted in Macon County, Alabama, poor sharecroppers were told they were being treated for ‘bad blood.’ In fact, the physicians in charge of the study ensured that these men went untreated.”
More recently, Bloomberg Businessweek reported that a bureaucratic quagmire — including a committee intended to bridge the “valley of death” in the government’s own innovation efforts — retarded the development of ZMapp, the experimental drug that might save people infected with Ebola:
Could a large stockpile of ZMapp have halted the spread of Ebola? No one can say. What’s certain is that the U.S. government hasn’t done a good job taking the idea behind ZMapp and turning it into a treatment. The technology for antibody cocktails such as ZMapp has “been around for a few decades,” says Robert Garry, a professor of microbiology at Tulane University. “This is something that, given the emergency, the government could have moved a little faster on, quite honestly.” He’s more right than he knows. The treatment came into the hands of a little-known Pentagon agency in late 2010, and, Bloomberg Businessweek has learned, ZMapp sat there dormant, waiting for a contract, for two years.
Less morbid than these examples but still serious is the toll of politicization and corruption from government ventures. While citing bankrupt solar power firm Solyndra has become almost a cliché, the Washington Post’s extensive investigation showed that Solyndra’s failure was symptomatic of systemic distortions of government innovation activities by political influences:
Meant to create jobs and cut reliance on foreign oil, Obama’s green-technology program was infused with politics at every level, The Washington Post found in an analysis of thousands of memos, company records and internal e-mails. Political considerations were raised repeatedly by company investors, Energy Department bureaucrats and White House officials.
When I worked on then-new Department of Energy solar energy RD&D programs during the Carter administration, DOE program managers felt political pressure to “demonstrate” useful results. One program pushed for demonstration projects of a solar-thermal hot water technology despite engineers warning that it was not yet commercially viable. When the demonstrated solar-thermal systems failed, they not only didn’t work but imposed onerous costs on the laundries, restaurants, and such that had been willing to try them — including lost revenue and angry customers. Later, when better solar systems finally came along, widely shared memories of those demonstrated failures discouraged adoption and impeded sales. (Decades later, Solyndra followed a similar pattern.)
None of this should be surprising. In their instructive book, Structuring an Energy Technology Revolution, Charles Weiss and William Bonvillian cite the conclusion from case studies by Cohen and Nolte of major government programs aimed at stimulating development of innovative commercial technology: “Once a program has exceeded the level of congressional attention…politics become at least as important as technical validity and the chances of a useful technical result dwindle.”
This finding may be viewed as a corollary to Lord Acton’s axiom, “Power tends to corrupt and absolute power corrupts absolutely.” More to the current point: Government programs will be corrupted and the bigger the program, the bigger the corruption.
Still, as one of my colleagues in a federal research center where I worked often observed, “More can be explained by incompetence than by conspiracy.” The failures of government research and innovation efforts are not simply the result of risk-taking adventures that fall short of objectives. The bungling of the strategically critical government health insurance website was not uncommon but rather symptomatic of recurring, costly failures of government-managed technology projects. As the Los Angeles Times has reported:
…as SAM.gov and other projects demonstrate, technology failures have become the rule in the federal government, not the exception. Websites crash, attempts to modernize systems founder and military systems costing hundreds of millions are abandoned before ever being used.
4. Nothing fails like success.
I take the label of this final counterargument from a comment often made by the late economist Kenneth Boulding. The point is that even the supposed successes of government attempts to foment innovation often come with unintended consequences or side effects that dilute or even overwhelm their benefits.
A notable case is the Eisenhower administration’s Atoms for Peace initiative. For geopolitical reasons, that program rushed to take a light-water nuclear reactor technology that was designed for propelling submarines and scale it up into a commercial, civilian electric power industry. Engineers understood at the time and since that there were several other possible designs for a commercial nuclear power system that were safer, less amenable to weapons production, and that had less onerous waste products. The system that the federal government promulgated — again, mainly for political and military rather than economic reasons — has for six decades stigmatized “nuclear power” with hazards, costs, and occasional disasters that could have been mitigated or avoided had the technology been developed instead by commercial interests for commercial purposes. Just when some analysts expected climate concerns to prompt a nuclear power “renaissance,” the Fukushima disaster not only stopped the movement to expand nuclear development but in some countries reversed it.
By promoting the wrong technology, the government not only diverted innovation but erected barriers to development and adoption of better nuclear power technology. A large, established nuclear power industry wed to the wrong technology has a vested interest in the status quo and has the political clout to oppose change. And the stigma of nuclear power as dirty and dangerous has undermined public support for nuclear power innovation.
The untoward consequences of Atoms for Peace are glaring but common to other government interventions.
Biofuels initiatives have followed a similar pattern. Through mandates and subsidies, the U.S. government (among others) has promoted the production and use of ethanol and other plant-based products as fuel. Various studies have shown that biofuels programs do more harm than good. They promote deforestation, a major global warming contributor, and other forms of pollution. And by diverting food to fuel production, they have increased food costs, doing particular harm to the poor and hungry. Yet the programs have spawned a powerful political lobby of rent-seeking biofuel interests that makes reversing or even just limiting such programs almost impossible.
More recently, federal policy has been effectively forcing doctors and hospitals to replace paper files with electronic health records (EHRs). A principal motive was to reduce healthcare costs, yet so far savings have been hard to detect. In theory, more detailed and consistent digital medical records, combined with powerful “big data” analytics, could lead to better, more effective treatments and services. In the meantime however, most physicians complain — often vehemently — that EHRs are undermining the doctor-patient relationship and thus the quality of care. Medscape’s study found three main gripes:
- “EHRs have made the patient encounter far more annoying and complex than it ever was before.”
- EHRs make doctors “…feel like data entry clerks, with a computer telling them how to practice medicine.”
- And, “…a large number of physicians feel that EHRs erode the doctor-patient relationship by creating a barrier between the two.”
Finally, the favorite poster child touted by enthusiasts for ambitious government technology programs is the Internet. But like light-water nuclear reactors, the Internet’s architecture was initially developed for military purposes, not for the global civilian uses it mushroomed into. Designed to enable communications among a ruling elite to survive nuclear war, the Internet has morphed into a digital venue for cyber crime and cyber war. The same government that gave us email technology more recently produced Stuxnet and myriad other unknown weapons of cyber combat. The boundaries between national security, organized crime, commerce, community, and rogue vandalism are difficult to discern.
CONCLUSION: Time for Plan B
The foremost, most general deficiency of the proposal to tax fossil-fuel emissions to fund government R&D and innovation programs is evident from Jenkins’ own analysis of political economy: The price of carbon that would motivate serious reductions in greenhouse-gas emissions is significantly higher than the price survey research shows the public is willing to pay. So the prospects for the tax-and-invest policy being enacted — even if more politically palatable than other options — still are poor.
The recent 113th Congress passed less useful legislation than any in modern history — far less than the “do nothing” Congress against which President Truman railed. Despite trillions of dollars of repairs needed to the nation’s crumbling roads and bridges, and the imminent exhaustion of the Highway Trust Fund, Congress was able to pass stopgap funding only through the next spring before going on a 5-week vacation at the beginning of August. Despite bipartisan angst about an immigration crisis, it was able to pass nothing to help solve that urgent problem.
Nor is there any likelihood that the Washington gridlock will change in the foreseeable future, given the outcome of the 2014 congressional elections. As Thomas Mann and Norman Ornstein explain in The Broken Branch and elsewhere, it will take deep structural reforms, not just electioneering, to fix what’s wrong with Congress and partisan politics more broadly. But a Pew Research Center study shows that America’s political polarization is intensifying. The needed reform of the country’s political architecture at best is going to happen slowly.
Even if somehow a dedicated funding program for energy technology innovation were to be enacted in the near future, its ability to endure over the decades needed to have a notable impact is doubtful. According to Congressional Budget Office estimates, by 2033 the federal government’s spending on entitlements — Medicare, Medicaid, Social Security, and others — will exceed the government’s total tax revenues. The “discretionary” portion of the federal budget that pays for everything else the government does would either have to disappear or be funded through increasingly costly borrowing. Of course, discretionary spending will not disappear, but will be pressed to shrink. “[U]nder current law,” CBO reports, “spending for all other federal benefits and services would be on track to make up a smaller percentage of GDP by 2024 than at any point in more than 70 years.”
So the proverbial can of fundamental fiscal policy reform that the Congress has continually kicked down the road for over a quarter century will soon become an immovable object. Given that the public is largely unwilling to give up the more costly features of the welfare state, what will follow in the wake of coming reforms is pretty well understood. Taxes will go up, entitlements will be limited, and the capacity of government to undertake costly discretionary projects will be significantly constrained.
Nor is the United States unique in facing a future of limited government ambition. Most of the world’s rich countries are being limited in similar ways. Aging populations strain entitlements and reduce or reverse workforce growth which in turn curtails economic expansion. Demand for entitlements grows even as the basis for tax receipts declines.
None of that is to say that government should stay out of science and technology. In any case, that’s not going to happen. The allure of science and technology to confer economic and strategic advantage has been irresistible to governments going back to the most ancient civilizations.
The world needs breakthrough innovations in energy technology to remedy a mess of serious environmental, economic, and social problems stoked by the conventional systems it relies on today. But the capacity of government to contribute much to meeting that need is going to be increasingly limited for at least a generation to come. And the established institutions and practices that government uses to promote science and technology — contorted by political influences, bureaucracy, gamesmanship, cronyism, and elitism — are increasingly out of sync with the agile way innovation works in the 21st century.
A paradigm shift is needed. Plan A, so to speak, no longer works. It’s time for Plan B.
Details of a Plan B can be found in Energy Innovation. The essential formula is the following.
Plan B begins with the recognition that a “big” energy innovation program does not necessarily need to be big in cost to the public treasury to be big in the scope of its reach, engagement, diversity, and impacts.
Rather, a number of limited-budget, efficient measures to promote significant innovation are available:
- The Information Technology and Innovation Foundation has suggested nine types of low-cost tactical measures that governments with limited budgets still may use to encourage needed energy technology innovations, including: targeting procurement, restructuring tax policies, revising regulations, and using information.
- Properly organized “inducement” prize competitions — such as the X Prize — can generate significant innovations far more cost-effectively than direct government expenditures on R&D, subsidies, and such.
- In-Q-Tel, the venture capital firm launched by the Central Intelligence Agency, is an example of how government can use modest resources to leverage big innovation results.
- Historically, foundations and other private philanthropists have nurtured major, even world-changing breakthroughs in science and technology.
The organization of a breakthrough energy innovation program requires greater decentralization, regionalization, diversification, and broader overall participation than traditional, centralized government technology projects. The “mesh” architecture of modern information technology points the way.
An indicator of what a socially and digitally networked energy-innovation mesh may accomplish is the open-source IT movement that begat products such as the Linux operating system, the World Wide Web, Wikipedia, and such. In fact, leading-edge corporations now are using social networks — such as Innocentive — to open up their innovation efforts to broad participation, demonstrating the potential of “Open Innovation.”
A parallel movement toward “Open Science” shows that crowdsourcing research problems can radically accelerate breakthrough discoveries. One example: Foldit, a massive multiplayer online game engaged a quarter-million players to decipher the structure of an AIDS-related enzyme in a few weeks, solving a problem that had stumped scientists for a decade.
Another key step toward opening up science and innovation is to eliminate the paywalls erected by the journal publishing industry. Conventional scientific and technical journals commonly charge thousands of dollars for annual subscriptions and anywhere from $30 to $100 for access to a single article. Such barriers retard the dissemination of knowledge needed to nurture innovation. The Public Library of Science (PLoS) pioneered breaking the paywall barrier. Instead of charging readers to access articles, PLoS charges authors for publishing them. (The cost of publishing research results can and should be built into grant budgets.)
And a third movement that is opening science and innovation to broad participation is the efflorescence of free Massive Open Online Courses (MOOCs) and other media that enable anyone in the world to acquire valuable knowledge without being hampered by steep tuition charges or excluded by elite university admission committees.
To accelerate the breakthroughs the world needs, energy innovation programs should defocus activity to many diverse nodes at the edge, and nurture bottom-up innovation. Moreover, they need to engage collaboration internationally, not just domestically. That does not mean just collaboration among elite institutions in rich countries. The focus needs to be on networks that link talented individuals in both industrialized and developing countries as full partners in creative collaborations. (See “Global Warming Needs Global Technology”.)
Such a network-oriented program design can achieve broad and diverse participation more effectively and at lower cost than the traditional form of centralized, top-down, hub-and-spoke national programs. The open innovation mesh is the key to untangle the energy policy mess.
Join the discussion of this article at OurEnergyPolicy.org.
*Carbon pricing or tax proposals generally focus on emissions of greenhouse gases such as carbon dioxide (CO2) and methane (C2H4). There are other greenhouse gases, such as nitrous oxide (N2O), that do not include carbon. There also are non-gaseous carbon substances, such as black carbon (soot), that may contribute to global warming by absorbing solar energy.
† Some minor typographical errors in this and other quotations were corrected.
If you liked this article, please recommend ♥ it.
Copyright 2014, Lewis J. Perelman
Picture source: Wikimedia Commons.