Could the Big Dig Have Cost Less?

As the parks it created finally fill with activity and the project fades from newspaper headlines, Boston’s Big Dig is subtly slipping into the city’s history.

Officially known as the Central Artery Bridge/Tunnel Project, the Big Dig buried an elevated freeway in downtown Boston and added a new freeway tunnel under Boston Harbor connecting Logan Airport to the city’s urban core.

Most of all, the project is known for its delays (over 15 years of construction) and huge price tag: over $14.6 billion, a total of roughly $22 billion with interest. The cost overruns were so severe it was responsible for temporarily cutting off the state’s Federal highway funding, dozens of lawsuits, and was even featured multiple times on NBC Nightly News’ regular feature “The Fleecing of America.”

Commonwealth politicians are content to put it the project behind them, trying to distance themselves from any involvement. Urbanists, meanwhile, are eager to point out the project’s many accomplishments. Downtown property values have increased, new parks have been created, and developers are beginning the work of re-knitting urban neighborhoods long divided by the unsightly highway. Reduced gridlock has improved urban air quality, and the harbor tunnel has improved access to the airport for transit riders (through the Silver Line) and motorists alike.

However, one unanswered question lurks in the air: could the Big Dig have cost less?

Like any historical counterfactual, a definitive answer is impossible. However, much could be revealed through a detailed case study. What delays were avoidable? What complications did the engineers overlook? Is there any evidence cost estimates were deliberately manipulated? Should the partnership between the Turnpike Authority and Bechtel/Parsons Brinkerhoff been differently designed to improve accountability? Could the project’s 144 separate construction contracts have been written differently? Were Federal highway standards or environmental laws responsible for excessive costs? In short, could changes to the underlying public policy structure have resulted in different costs and construction time? To my knowledge, these questions have not been deeply examined by scholars. If any readers are aware of studies, please post them below.

The result may not be nefarious scheming by greedy contractors. In fact, scholar Bent Flyvbjerg has argued major causes of cost overruns are systematic under-estimation of costs due to political pressure, and a failure to properly account for the inevitable risks that occur as a result of the complexity of megaprojects.

Image via Wikipedia

Far from a matter of historical interest, the high cost of transportation infrastructure is a pressing policy issue in a state currently spending $3 billion re-build bridges and planning an extension of the city’s Green Line, among other proposed projects. The high cost of transportation projects was raised recently on an email list operated by the Somerville Transportation Equity Partnership, in a thread grousing about the $600,000 price for a bike cage and high cost estimates for the Green Line extension. The issue provoked a rare response from the Commonwealth’s Secretary of Transportation Jeffrey Mullan himself, who denied the costs were inflated or the bidding process is faulty. Although the message concedes “things cost a lot; unacceptably so sometimes,” the post omits any speculation as to why.

Perhaps one day the question will be subject to the scrutiny it deserves.

Who Needs ‘Centralized City Planning’? Everyone.

The always-interesting Witold Rybczynski has a provocative piece up on Slate arguing that the failure of government-led urban planning means that “in a democracy, a vision of the future city will best emerge from the marketplace.” I don’t disagree with his observation that private organizations and real estate developers have taken the lead in shaping our cities, however I don’t believe it follows that the government has no role whatsoever.

Such an argument erases the many ways governments are deeply involved in planning urban spatial structure: designing and operating streets and other infrastructure, regulating urban land markets through enforcement of property rights and zoning, shaping the location and character of development through wetlands and other environmental regulations, subsidizing and shaping the housing finance system, and establishing and enforcing building codes and standards, to just name a few.

Although Rybczynski is right the government has largely withdrawn from the business of directly engaging in architecture and urban design (and that’s a good thing), the lesson isn’t that government should (or will) withdraw completely. The stark contrasts of quality of life between well- and poorly-governed cities illustrates just how important these more subtle processes of planning remain. His argument reminds me of Peter Montgomery’s thoughtful analysis of Jane Jacob’s The Death and Life of Great American Cities. Although her critique of Robert Moses and heavy-handed modernist city planning is important, Montgomery argues her celebration of the urbanity of her neighborhood omits the government processes that establish the framework of urban life (zoning, subway system, urban services, etc). In this way it can be read as a neoconservative tract, writing out the role of government. (In addition, Montgomery argues she ignores corporations, class and race divisions, and metropolitan equity).

To be fair, Rybczynski does stress the importance of government for “management” and “little plans,” and to a degree I’m just rejecting his definition of planning. But the point I hope to make is the “urban visions” created by real estate developers aren’t a pure product of the market, but derivative from government-determined transportation systems, zoning, and metropolitan spatial structure.

The more interesting and accurate conclusion to draw from the failures of modernist city planning is to consider which forms of government planning are still active and desirable. In this sense, Rybczynski’s article is a bit behind the times. The tremendous interest in high speed rail, urban transit, green building codes, the government’s role in wind power and broadband, and housing finance regulation has reminded us of the central role of government in shaping our cities. Hopefully this will be the legacy of the Obama era: that the choice between government and the market is a false dichotomy. Because the two are mutually dependent, addressing public problems (such as city planning, and yes, health care) requires attention to the design of each.

> Witold Rybczynski – Don’t Plan On It: Centralized city planning is not the answer to the problems facing America’s cities.

RSVP Now for Government 2.0 Camp New England

Registration just opened for Gov 2.0 Camp New England, a one-day unconference I’m helping to plan. It will be held Saturday, March 6th at the Harvard Kennedy School of Government in Harvard Square. The others involved in planning are Yasmin Fodil (KSG), Laurel Ruma (O’Reilly Media) and Sarah Bourne and Jessica Weiss (Commonwealth of MA). Thanks to this great group we’ve already got an interesting list of attendees registered, and a number of topics percolating on the wiki. In true unconference style the sessions won’t be finalized until the day of the event, but we are encouraging collaboration on the wiki.

What is government 2.0? I attempted to define the topic earlier this month, but I’m not hung up on definitions. If you’re interested in applying Internet technologies to the business of government, we hope you’ll attend.

> See Conference Wiki or Registration

What is Government 2.0?

With last year’s Gov 2.0 Summit and the explosion of social networking service GovLoop, “government 2.0″ has become a buzzword in technology and government circles. What does government 2.0 refer to? And what exactly was the government 1.0 that we’re improving on? This article attempts to define the term and unearth some of the hidden assumptions and implications that result from applying concepts developed in Silicon Valley technology startups to the complex and age-old problem of governance.

The term government 2.0 is a deliberate reference to the term “web 2.0,” coined by publisher Tim O’Reilly to refer to interactive, social websites like Wikipedia and Facebook, which have revolutionized how people use the web. Before delving into the meaning of government 2.0, we should consider government 1.0, the government analogue to web 1.0. Although less common now, the term most often used for this initial approach to technology in government is e-government.

The Center for Technology in Government defined e-government as having three components: e-management, e-services, and e-democracy. The first two have been largely realized. Governments have adopted, to varying degrees of sophistication, internal information technology systems such as networks, databases, and intranets. As we will see, government 2.0 practices often rely on these underlying systems. Governments have long provided e-services to constituents through websites, email, or APIs, including tax payments, service requests, and digital applications and paperwork. The last component, e-democracy, has been more elusive. In the web 1.0 world, this has most often meant emailing elected officials or signing petitions on topics. These activities have grown, although in the U.S. context exist mainly outside of government websites or structures.

At a lecture hosted by the Kennedy School Government 2.0 Professional Interest Council this fall, Nicco Mele suggested we adopt Tim O’Reilly’s web 2.0 principles as a starting point for government 2.0. My essay builds on his interesting lecture.

1. Government as Platform

O’Reilly’s first principle is “the web as platform,” adjusted for our purposes to be “government as platform.” The most obvious examples of this are where government agencies provide data or host competitions to encourage creative ideas that serve the public interest. The “apps” competitions in Washington, D.C. and New York and sponsored by the Massachusetts Department of Transportation, are a start to this trend. In these competitions, government provides the data, and an ecosystem of third party developers and tools helps unleash the value for the public, creating new tools, resources, and analyses.

Another example where government acts as platform is the phenomenon of participatory budgeting, pioneered by cities in Brazil and now has spread to a number of cities around the world. This approach puts budgetary decision-making, or some part of it, directly in the hands of citizens, bypassing existing representative models of decision-making. The technical dimensions of this are only now being explored, and in the Brazilian case above deliberation and voting online complemented conventional public meetings.

When it comes to service delivery, it is less clear what “government as platform” means. It may echo a broader political agenda that has sought to re-define the role of government through systematic privatization of formerly government functions, such as education or public services. After all, when governments provide educational or housing vouchers, aren’t they acting as the intermediary, or a platform? The political implications of shifting government from a service provider role to a facilitating role deserves consideration. This issue is connected to a host of issues surrounding contracting and public private partnerships. Governments may want to retain some types of service delivery if the good cannot be contracted for, or the public wants to enforce certain service standards.

2. Harnessing Collective Intelligence

The second principle is “harnessing collective intelligence.” Obama’s Memorandum on Transparency and Open Government identified collaboration as a policy goal for the federal government. In fact, Obama’s Deputy Chief Technology Officer for Open Government Beth Noveck experimented with collaboration tools to create an open government policy last summer. In other areas there are limited successes of citizen-government collaboration. The Peer to Patent program pools expert opinion to speed the patent process. The Next Stop Design project in Salt Lake City, Utah used crowdsourcing to select the design for new bus shelters. One of the people involved in the project, Daren Brabham, is writing a PhD dissertation on the application of crowdsourcing to public problems. In Melbourne, the consulting firm Collabforge ran a wiki as a component of a conventional planning process to generate the new city plan.

Fundamentally, this trend will face several types of powerful resistance.

First, it can run counter to traditional concepts of representative democracy, where elected officials work “down” through an expert bureaucracy to create and implement policy. Archon Fung has proposed “empowered participation” can be deployed as a governance method for specific issues, such as Chicago’s school committees or neighborhood policing committees. However, creating these structures depends on modifying existing forms of governance. Existing projects have avoided this in several ways. The apps competitions aren’t about creating policy, and the government hosts can always disavow responsibility. Idea-generation contests usually reserve final decisions to designated juries. Policy-creation projects retain the final decision-making power with conventional authorities. However, pushing this further into what Beth Noveck calls “wiki government” will require addressing this tension with existing practices.

Second, a host of public problems require technical expertise to analyze or solve. The question of how to integrate technical forms of knowledge with citizens is far from resolved. The cutting edge involves putting modeling tools in the hands of citizens, who use them as “decision support tools,” but this runs counter to existing models of professional practice and the very real need for significant expertise to complete complex analyses.

Lastly classified data and national security, a major governmental function, may never be opened to the public. Interestingly, Department of Defense has been interested in the collaborative potential of internal communication across their vast bureaucracy through wikis, for example launching a wiki to improve the Army Field Manual.

3. Open Data Standards

The third principle is the use of data standards. Expanding access to government data is a major trend, with initiatives underway at the federal, state, and local level to create data portals. The concept of linked data, emerging out of the Wikipedia project, seems poised to move into government datasets. In fact, greater linking and cross-comparison among the expanding amount of available government data will create a positive pressure to ensure cross-compatibility. Within Massachusetts state government, for example, town-level data has become a standard for comparison and analysis. With the federal government in setting metadata and other standards already, this may happen slowly but some signs are already in place. Using this to evaluate government may be misleading: the primary purpose of government isn’t to create data, although it is an important one. The technological viewpoint threatens to be reductionist, viewing the government as primarily engaged in collecting and hosting data. In reality, most money and effort in government is spent on delivering healthcare, education, national defense, grant programs, and regulatory actions, where data can play a supporting role (perhaps as indicators) but is not even always a mandatory input to governance.

In Boston, the author of a recent major report studying the city’s transit agency said in November he wouldn’t ride the busy Red Line due to serious maintenance issues that threaten to cause a train derailment. At roughly the same time, data enthusiasts were demanding real-time data about bus and train arrivals at the MassDOT developers conference. When our transit systems are in real danger of catastrophic failure, shouldn’t we spend all available funds preventing disaster for the existing riders, rather than inventing technology to make use more convenient? How can these important goals be balanced properly?

4. Customer Service

The last principle discussed by Nicco is customer service, based on O’Reilly’s “rich user experience.” An emphasis on customer services is undeniable at all levels of government. Cities have launched successful 311 systems for managing citizen requests, and governments have been subscribing to the “plain language” movement make government information more understandable and usable to citizens. However, just like “government as platform,” this principle too often reduces government to a consumer-producer relationship where the government provides services just like private firms might in the marketplace. Customer service is important, but so is engaging with citizens to generate ideas and implement solutions. In exchange for expecting service, citizens have the responsibility to understand the resource and legal limitations of government.

5. Incremental Policy

O’Reilly has several additional principles: end of the software release cycle, lightweight programming models, and software above the level of the single device. Of these, I think the principle for government is the advent of more iterative forms of policy making. The field of planning has developed theories of incrementalism or “muddling through,” to reflect the real-world pace of change. The web supports both short bursts of activity but also long-term archiving, and professionals are only now learning how to use the tools to develop sustained interest and engagement through ongoing conversations and communications.

Conclusion

What do we learn from this exercise? First, I’m not sure government 2.0 is yet a new type of government, instead a collection of promising trends. The adoption of new social and technical approaches of idea creation and governance don’t resolving age-old questions about what government should be doing, and how it should approach principles of equity and justice. In fact, what could emerge is a new, technically-enabled model of in the tradition of the “developmental state,” the concept that the state itself is engaged in economic and community development. This is perhaps the most important lesson of these trends: existing government processes should be examined and where they are not working be re-invented to take advantage of the ability of technology to expand the activity of governance beyond the institutions of government.

Highlights From Princeton Planning and the Internet Summit

I recently returned from a conference on “City Planning, Civic Engagement and the Internet” held in Princeton, New Jersey co-sponsored by Princeton’s Woodrow Wilson School of Public and International Affairs and Center for Information Technology Policy. The conference was planned largely by Christian Peralta, the former editor of Planetizen, who did a great job assembling a fascinating group and making sure everything ran like clockwork. For the benefit of those who couldn’t attend I thought I would write a short description of some of the highlights..

Best Practices in Local Government

An employee of an independent government agency, the Metropolitan Area Planning Council, I took particular interest in the representatives from local governments. Representing the City of Toronto’s Public Consultation Unit were Mike Logan and Robert Davis. Their unit has evolved since its creation in the late 1980s into the city government’s go-to resource for public involvement. I think this is a model that could be replicated elsewhere: one office maintains the expertise about all the approaches to involve the public, and works with the project sponsors to create and implement an appropriate and resource-efficient approaches. It also creates one central place at the city for citizens to approach with questions. They presented on some of their work to use Facebook to reach communities (it required special permission from the IT department), and discussed the unusual challenge of working in Toronto’s highly multicultural environment, which requires extensive translation. Public consultation coordinator Mike Logan even handed me a business card with the information imprinted in braille on it, which itself was a statement to their commitment to excellence in accessibility.

Another particularly noteworthy presenter was Mark Elliott, whose consulting firm Collabforge set up a wiki for a recent planning process in the City of Melbourne, Australia. As you might expect from someone who earned a PhD with a dissertation on “A Theoretical Framework for Mass Collaboration,” Mark impressed me with his thoughtful approach to integrating collaborative technologies to planning. In general I think advocates of wikis underestimate the technical complexity of the technology, as well as the limitations to a radically flattening technology. Mark’s work on FutureMelbourne was apparently successful and he’s definitely someone to watch.

Also attending was Seattle’s Chief Technology Officer, Bill Schrier, who blogs about technology and government at his blog Chief Seattle Geek. Mark Bosworth, a GIS expert from Portland, Oregon’s regional planning agency Metro gave a whimsical presentation on the history of GIS and highlighting some of their many customized web applications including a bicycle trip planner (of course), and a “build your own” transit system tool.

Private Sector Innovation

Several consultants attended, presenting on a wide range of topics. Edward Andersson, from the UK consulting firm Involve, gave a thoughtful presentation on the history of participation in the UK and their firm’s approach. The company’s website PeopleandParticipation.net is a rich resource on the topic. Rhiza Lab’s Josh Knauer and Jeff Christensen presented on their firm’s powerful online data and mapping tools. It was a wonder they made it since Josh explained they’ve been working nearly around the clock on their FluTracker website. Lastly Jocelyn Hittle and Jason Lally from PlaceMatters displayed some amazing touch and light-sensitive technology made using two Wii remotes and a lot of ingenuity.

View from the Academy

The academic speakers provided interesting perspective and a glimpse of their latest research. Ohio State’s Jennifer Evans Cowley presented on her research analyzing the use of social networking in urban planning, and has even created a Facebook group dedicated to the topic. Hunter College’s Laxmi Ramasubramanian presented on the theoretical context for public participation, and Iowa State’s Chris Seeger presented on his extensive background in participatory GIS.

Out of the Box

Of course, some of the presenters fit none of these categories. Adrian Holovaty, founder of the totally unique Everyblock.com, presented on his work harnessing the web’s geographic data to create a hyperlocal news source. Although I missed the presentation, Matthew Golas from PlanPhilly.com described that website’s civic mission to foster dialogue on planning in Philadelphia. Also presenting were John Geraci, from DIYCity, a project to imagine a new interactive “DIY” urbanism, and Nick Grossman from the invaluable Open Planning Project, the folks behind Streetsblog. (Aside: We need a Boston Streetsblog) The Sunlight Foundation’s John Wonderlich and Ali Felski are working hard in D.C. to improve government websites and access to data. (My friend Tom Lee is also with their lab).

For much more see the #ccisummit Twitter tag. The sessions were also recorded, and they will be eventually posted to the conference website. Attendees: what did I miss?

Tolls More Equitable Than Sales Tax For Funding Freeways

You’ve heard the buzz about “Lexus Lanes,” a new trend where tolls are adjusted in order to keep some freeway lanes flowing smoothly. They’re related to the idea of charging higher prices for parking, or even a congestion charge such as the one considered for New York City. It’s widely thought the lanes are unfair, since they allow wealthy drivers to zip past congestion. There’s only one problem with that view: a new study disproved it, arguing instead toll lanes are more just than the usual method for funding highways, sales taxes.

Two California professors considered the issue in a new article titled, “Just Pricing: The Distributional Effects of Congestion Pricing and Sales Taxes.” The study found that the lanes were disproportionately used by middle and upper-middle income people, and that the tolls were regressive. So what’s the rub? It turns out the usual means for paying for transportation infrastructure, such as sales and gas taxes, are even more regressive than tolls. In fact, the study concludes that:

… if [sales tax] funds had been used to finance the express lanes, the study found, the poor and wealthy would have paid more. Middle- and upper-middle-income taxpayers would have paid $26 million less each year than they paid under the current cost-distribution system, and the very poorest residents would have paid over $3 million more than they actually did under the current toll system.

They conclude that “Using sales taxes to fund roadways creates substantial savings to drivers by shifting some of the costs of driving from drivers to consumers at large, and in the process disproportionately favors the more affluent at the expense of the impoverished.” The authors propose two policies to overcome the remaining regressive character of tolls: giving out free travel credits to low income commuters, or using the funds to invest in public transit. The comparison is between tolls and general sales taxes, not gas taxes, but I suspect gas taxes would have been only slightly less regressive than sales taxes. (Because the poor own fewer cars and drive less)

Previously I also suggested we should consider other benefits of congestion pricing in the equation – greater transportation choice for all (including low-income commuters), less pollution, and perhaps a shift in behavior towards transit, carpooling, or other more efficient modes. I also discussed before some of the implications for another form of congestion pricing — raising parking meter rates.

What most frustrates me with congestion pricing critics is not their concern — not enough research has been done on equity, and it is a valid point to discuss — but how misplaced it seems given our skewed policies. Our society is riddled with deeply regressive policies. Sales taxes, gas taxes, and lotteries are all known to be regressive. We spend more than twice as much money subsidizing housing for the rich than we do for the poor. The poor disproportionately live near sources of pollution, and consequently have higher rates of asthma, heart disease, and other diseases caused by environmental factors. Meanwhile, our public transit systems, critical lifelines to opportunity for the very poor, are crumbling. In that light, adopting less-regressive congestion pricing and spending some of the revenue on transit service seems like a good decision.

> UCLA: “Joint UCLA–USC study shows that toll roads are more fair than taxes”
> LATimes Blog: “Study finds congestion pricing doesn’t hurt the poor”

The Equity of Housing Tax Benefits

The wide-ranging housing bill recently passed by Congress includes a program to help homeowners avoid foreclosure, money for community development, and other measures. One of its important provisions is a one-time tax benefit of $7,500 (or 10% of the home’s purchase price, whichever is less). Unlike many of the existing tax benefits of home ownership, there’s an income limit and the benefit must be repaid, although without interest. The income limit is $75,000 for singles and $150,000 for couples.

The National Association of Realtors explains the the pay-back requirements:

It is not a full credit because you do have to pay back this amount over a 15 year time period from the second year. The payback provisions also have many conditions, which we are further researching. But in the worst case, you would need to pay back the $7,500 over a 15 year time span from 2010. So in your 2010 tax filing, you would need to pay $500. Even in the worst case scenario of paying back the tax credit fully over a 15 year time span, the tax credit is still a huge benefit to homebuyers. First, money today is worth more than money tomorrow – far more than money 15 years from now. Money loses value over time due to inflation and from the interest income one would receive on that money before fully paying it back.

Let’s take a look at the effects of the existing homeowner tax benefits.

This 2004 report from the National Low Income Housing Coalition has some interesting analysis. This graph breaks down the estimated size of the various housing tax benefits, including the mortgage interest and property tax deduction and housing capital gains exemption. They point out these three benefits were more than double HUD’s total outlays in 2004.

The particularly interesting things about these benefits is how deeply regressive these benefits are. When housing programs and tax benefits are factored, we spend 2.2 times more money on housing for the richest two-fifths of the population than the poorest two-fifths. The richest fifth of society — with incomes above $86,585 in this 2004 report — get over $50 billion in benefits.

Since the benefit is something of a sacred cow, I don’t expect it to go away anytime soon. We should just keep them in mind when debating funds for affordable housing or the equity of congestion charges.

> National Low Income Housing Coalition: Changing Priorities: The Federal Budget and Housing Assistance 1976 – 2005
> USAToday: Housing rescue bill may fall short; Who benefits?
> The Tax Foundation: Who Benefits From the Mortgage Interest Deduction?