in a visually appealing and useful way, such as through graphics and organized menus. ...... visitors to view content in
TEXAS TRANSPARENCY: BEYOND RAW DATA
LBJ School of Public Affairs – Policy Research Project
May 5, 2011
Transparency in Texas: Beyond Raw Data
Dedication To the memory of Professor Gary Chapman, our mentor, colleague, and friend. His passion and eagerness to help was truly inspiring.
Project Directors Gary Chapman Sherri Greenberg
Researchers & Contributors Laruen Ames Catherine Bracy Adam Colligan J.T. Harechmak Kimberly Johnson Joshua Levine Kellyann McClain Christina Mechler Chris Rhys Meredith Whipple
Special Thanks To Suzanne Berman Damien Brockmann Talmadge Heflin Jon Liebkowski Matt Stiles Kelly Young
2
Transparency in Texas: Beyond Raw Data
Table of Contents EXECUTIVE SUMMARY .................................................................................................................... 4 INTRODUCTION .............................................................................................................................. 5 Purpose ........................................................................................................................................................ 5 Background .................................................................................................................................................. 5 Research ...................................................................................................................................................... 7 ORIGIN OF DATA ........................................................................................................................... 11 PRESENTATION OF DATA .............................................................................................................. 14 ENGAGING THE PUBLIC ................................................................................................................ 17 RECOMMENDATIONS ................................................................................................................... 19 Origin of Data ............................................................................................................................................. 19 Presentation of Data .................................................................................................................................. 20 Engaging the Public .................................................................................................................................... 22 CONCLUSION ................................................................................................................................ 24 Appendix ...................................................................................................................................... 25 A. TEXAS SCHOOL DISTRICT FINANCE DATA .............................................................................................. 25 B. LEGISLATIVE BUDGET BOARD WEBSITE ................................................................................................. 32 C. CITY OF KYLE .......................................................................................................................................... 38 D. CITY OF MANOR: Transparency on a Budget ......................................................................................... 40 E. WEB TOOLS ............................................................................................................................................ 42 F. LEGISLATIVE RESOURCES ....................................................................................................................... 44 G. QUESTIONNAIRE .................................................................................................................................... 46 H. GLOSSARY .............................................................................................................................................. 52
3
Transparency in Texas: Beyond Raw Data
EXECUTIVE SUMMARY Advances in technology over the past decade have created the expectation that public information be available online. This has never been more important than in the current economic climate, in which all levels of government are facing severe budget shortfalls and the public is demanding accountability. Enhancing transparency in the budget process, as well as in all government functions, requires easy access to information and mechanisms to promote civic engagement. For our research, we gathered information from various stakeholders inside and outside of government on how to increase transparency. We did this through a questionnaire and a statewide conference. This feedback helped us identify strategies and best practices for connecting interested parties to government data and facilitating civic engagement. Our research also includes case studies that further clarify the challenges to achieving transparency in the context of the state government, state agencies, and local governments. We identified barriers to transparency in the areas of data production and content creation, data formatting and presentation, and civic engagement. We included recommendations for three types of stakeholders: government, intermediary organizations, and the public. Intermediary organizations include nonprofit organizations and media outlets. Our recommendations are organized into three sections; “Origin of Data,” Presentation of Data,” and “Engaging the Public.” We recommend that data be comprehensible, machine-‐readable, exportable, and editable. Agencies should assess community needs, appoint a staff person to spearhead transparency, and review and renew data processes. In terms of presentation, we recommend that data be displayed in a visually appealing and useful way, such as through graphics and organized menus. We also encourage agencies to make use of social media and elicit user feedback on their websites. Finally, in addressing civic engagement, we offer recommendations for both news organizations and policy-‐ oriented nonprofits. We suggest that news organizations participate in media literacy programs, form partnerships with civic engagement groups, and offer public events and programming for the community. We recommend that policy-‐oriented non-‐profits utilize social media, provide website interactivity, state clear goals on their websites, and highlight the outcomes of user engagement. Overall, online transparency must foster understanding, provide context, and promote interactivity in order to truly engage the public. The shift we describe here is one from e-‐government—government data on websites— to e-‐governance—truly virtual government that promotes civic engagement.
4
Transparency in Texas: Beyond Raw Data
INTRODUCTION Now, more than any time in recent history, government budgets have become the hot topic of conversation from the nightly news to the office water cooler. People are asking, “How much are we spending? What are we spending it on, and why?” The public wants to know which programs are being cut, which are being saved, whether taxes will be raised and if there might be a better way to govern. Many will turn to the Internet to find answers to these questions and to find out how they can get involved. They will look to government websites, online news outlets and nonprofit public interest and advocacy groups. Under the best circumstances, they will quickly find information in useful, easily understood formats on websites providing avenues for feedback and interaction. In most cases, however, they will encounter websites that are works-‐in-‐progress and are only just beginning to incorporate best practices for online open government.
Governments at all levels are just beginning to understand the power of the Internet for transforming the way they govern. President Obama identified transparency, participation and collaboration as three key components of open government in his Memorandum on Transparency and Open Government, issued on January 21, 2009.i To achieve a government that is more efficient, accountable and responsive to citizens, the government must provide the public with access to useful information that is easily understood and facilitates interaction. The government must engage the public. The availability of online information is key to attaining transparency. Government data, especially financial information, should be online in open formats and easily accessible by the public. Machine readable and exportable data, though, does not in itself stimulate public participation. The data must be both accessible and useable. There must be input from civic groups and media outlets to interpret the data and provide context. Finally, the information must be presented to the public in a comprehensible format with tools for two-‐way communications. Over the years, Texas has been a leader among states in the open government and transparency movement. The Comptroller’s office publishes state spending and contract information online and encourages local entities to do the same. Other state agencies also have financial information online and available to the public; although, the data can be difficult to locate and analyze. This report provides tools and recommendations to overcome the most common challenges that reduce transparency.
Purpose In the past decade, technological advances have created the expectation that public information, including financial data, be available online. This has never been more important than in the current economic climate, in which all levels of government are facing severe budget shortfalls and the public is demanding accountability. Enhancing transparency in the budget process, as well as in all government functions, requires easy access to information and mechanisms for civic engagement. Government agencies, third party organizations that act as intermediaries, and the public can use the analysis and recommendations in this report to make the most of available data and create a more informed and engaged citizenry.
Background Our research builds on work from 2010 by a prior Policy Research Project at the LBJ School of Public Affairs. The 2010 report, Texas Financial Transparency: Open and Online, addresses state level financial transparency websites, local governments’ online presentation of financial information, and openness in
5
Transparency in Texas: Beyond Raw Data the legislative appropriations process.ii The key recommendations for state and local governments are that data should be searchable and exportable, available in raw, real-‐time formats, and consolidated into a “one-‐stop-‐shop.” Data also should be visually appealing and should be put in context to facilitate engagement with users. In the case of local governments, the report recommends the use of a template to help smaller entities publish data, as well as upgrading the requirements for local government to qualify for the Comptroller’s Leadership Circle, which recognizes local government efforts in publishing budget documents, annual financial reports and check registers. The recommendations for the appropriations process are that certain documents that are available to legislators, such as background information and recommendations and changes from subcommittee meetings, should be available online. As a result of this recommendation, the Legislative Budget Board (LBB) began publishing such decision documents this year during the 82nd Texas Legislative Session. Also, the full text of the budget legislation, currently only available to download page-‐by-‐page, should be easy to view and search. Apart from the recommendations for specific entities, the report also identifies four principles for transparency reform: the definition of public now means online, data must be relevant and provide context, data must be available at all levels to allow users to follow the money, and transparency enables participation, encourages collaboration, and increases efficiency. Our research focuses on the fourth principle of transparency, which is the interactive aspect that should stimulate civic engagement. Civic engagement is a key element in a democratic society, so providing mechanisms to enhance engagement supports democracy and the goal of pursuing public interest.iii Studies of local government websites demonstrate that this can be achieved by providing information in accessible formats.iv The transfer of information that is useful and relevant to the public works well to support the concept of public involvement in a representative or pluralist theory, but has been less effective as a tool for direct democracy. All conceptions of democracy rely on some level of civic engagement. As citizens use the Internet to gather information, they show increased interest in political participation.v This can be exhibited through online or personal exchanges and in voting behavior. Governments can also use the online interactions they have with the public to improve trust and accountability.vi A more informed and involved citizenry results in a better functioning government. Civic engagement does not simply occur, however, because a government entity builds a website. There are key elements that increase engagement by providing a way for the public to be more than consumers of data. The most basic characteristic of a website that promotes an active citizenry is some method to solicit public feedback.vii Information must be presented to the public, but also allow for two-‐ way communication, so that people can submit their opinions, which are then used to inform government decision-‐making. Implementing systems to include the public in government processes signifies a shift in e-‐government objectives. In fact, it means moving from e-‐government, which is the online delivery of government services and information, to e-‐governance, which is direct online public engagement in government.viii Rather than being simply a means to streamline services and focus on cost-‐savings and efficiency, developing civic engagement tools focuses on the public. Enhancing civic engagement through online transparency, though, is not always a perfect solution. Implementing new technologies to enhance online governance assumes that the public has easy access to the Internet. Broadband adoption, however, is not universal, and in fact, in some demographics it is as low as 24%.ix Even with Internet access, not all users can navigate complex, data-‐filled websites.x When advocating for more information online, we must be cautious not to widen the digital divide.
6
Transparency in Texas: Beyond Raw Data Universally increasing civic engagement also requires the adoption of policies that reduce the digital divide and enhance computer literacy.
Research Our research identifies three major stakeholder groups affected by government transparency online: government, intermediaries, and the public. We define “government” as all government actors who produce data and publications. “Intermediaries” are all actors outside government who process or give context to government data for consumption by the public. Finally, the “public” is comprised of individuals who use tools and data produced by government and intermediaries. In addressing these groups, this report provides recommendations on how government can best present raw data, how government and intermediaries can best use raw data and provide context for the public, and how the public can find this information and put it to use. Through case studies and interactions with stakeholders we have identified barriers to transparency in data production and formatting, online presentation and civic engagement. We recognize that transparency is a fluid process involving a variety of stakeholders in different roles. Our recommendations and best practices provide a roadmap for governments toward transparency and offers steps for organizations outside of government to encourage civic engagement using government data. The case studies from our research elucidate the transparency challenges in state government, a state agency, and local governments. The Texas Education Agency (TEA) exemplifies an agency that collects and publishes large amounts of data, but has not collected the data in a format that is easy to display online, making it difficult for citizens to understand. The TEA case study provides a visual mapping tool of the TEA data.xi The LBB houses important information that is key to developing an informed citizenry; however, our research found that the LBB’s website is outdated and difficult to navigate. As part of our project, we are creating a prototype for a new LBB website homepage. The presentation of state budget data in unsearchable formats undoubtedly creates challenges for interested citizens.xii On the local level, the City of Kyle has been working towards transparency by soliciting feedback from the City’s residents. We are creating a survey for Kyle officials to learn more about how the public uses the City’s financial information.xiii Our final case study, the City of Manor, is an example of an entity that successfully used emerging technology and open source tools to advance transparency in the face of fiscal constraints. xiv Another important aspect of our research has been to gather feedback and input from stakeholders inside and outside of government on ways for government to become more transparent. We elicited stakeholder feedback by distributing a questionnaire and holding a conference. This feedback informed our research direction and helped us identify best practices to connect interested parties with government data to promote civic engagement. The 2010 Policy Research Team used a questionnaire to obtain feedback on user satisfaction with the accessibility and usability of online government data. The results resoundingly indicated that users were dissatisfied, expressing concern with the availability, exportability, and presentation of data, among other issues.xv Having established that users were largely dissatisfied, our research team distributed a subsequent questionnaire (see Appendix G) to ascertain what would improve user experience in accessing and interacting with online government data.
7
Transparency in Texas: Beyond Raw Data
Government Transparency 2010, “Transparency Questionnaire.”
Seventy-‐five out of 220 responded to the questionnaire. The questionnaire asked people to identify their age, role in accessing government data, and need for accessing government data. Our results indicated that questionnaire respondents were quite diverse as far as age and role, including researchers, nonprofit organizations and businesses. In terms of what participants use government data for, “research” and “tracking policy or trends” received the most responses. A question asked participants to identify what types of information they seek online. We found a relatively even spread among the choices provided with “text of legislation, local ordinances, or rules” receiving the highest number of responses with about 68 percent of respondents selecting this choice. In terms of what format would be helpful in accessing government data, respondents most readily identified exportable data and graphs/charts. The most prevalent barriers to respondents utilizing the data they needed included “can’t locate what you need” and “format”. As far as which resources would improve their access to government data, respondents selected “better organized websites” and “exportable data”. We asked respondents if they had all the tools they needed and how they would use government data. “Research” and “tracking policy and trends” received the most responses though the answers varied widely including more that 50 percent of respondents selecting “inform voting” and “advocacy efforts”. In the open-‐ended question, respondents indicated that a lack of consistency in websites is a major issue. There was a call for a data clearing house or “one stop shop”. To further the conversation with those interested in open government, we worked with a local organization, EFF-‐Austin, on a conference that brought together officials in state and local government, academia, media, advocacy groups and private industry to explore best practices in Open Government.
8
Transparency in Texas: Beyond Raw Data The conference entitled, “Texas Government 2.0 Camp” was held January 28 and 29, 2011 and consisted of one day of panels and one day of “unconference”, or open conference, agenda. The goal of the conference was to discern whether there were better ways to help the public find the needle in the haystack of government information. Through panel discussions and conversations with those in attendance, we were able identify some important issues affecting online transparency. Among the major themes discussed were strengths and weaknesses of government in disseminating data, the role intermediary organizations can play in the effective communication of government data and the importance of collaboration to engage the public in improving government and helping make government information more usable. Several attendees explained that the role of government in producing data sometimes limits the usefulness of the data to the public. Various levels of government often produce data to fulfill specific reporting requirements, so data is tailored to those needs. Often, data is not in a format or layout that is easily accessible. Agencies tend to take an internal view of information needs rather than a holistic approach. Frequently, the information owner posts information on its website, but the public often does not know where to find the data. Additionally, there are numerous potential audiences seeking information. Those audiences include academic researchers, legislative staff, staff of other government agencies, advocacy organizations, and potential contractors and interested members of the public, to name a few. Governments are not well suited to identifying all audiences and putting the information in a format that is usable for each and doing so may be beyond the scope of government responsibility. Some Conference attendees suggested we consider government limitations to determine which concerns government can best address and which can be addressed in other ways. As one attendee said, “Elephants don’t dance well,” and he suggested we should not ask government to go beyond its abilities. Several attendees suggested government could create a centralized location for publishing data from across agencies and departments so people do not need to know the owner to find the data. Government can make information available in machine readable and exportable formats. However, government might not be in the best position to package the information in a “retail” format. Essentially, government might be best suited to bale the hay and stack it, so that looking for the needle is easier. Other attendees suggested that intermediary organizations might be best suited to take the raw data and prepare it for public consumption. Intermediaries include advocacy organizations, watchdog groups, the media and other public interest groups. These organizations know the information that their constituencies want and how it will be most useful. They are in a position to aggregate government data and create applications that display it with context and meaning. Intermediaries play an important role in proactively pursuing information that might be difficult or cumbersome to access and reproducing it in a relatable way. In his Keynote address, Evan Smith, CEO and Editor of the Texas Tribune, commented that much of the Tribune’s data work consists of asking for the information and “standing there tapping our foot;” something many in the general public do not generally have the time or energy to do. In providing context, intermediaries have the opportunity to take the data they gather and connect it to the larger picture. They can help people see more than test scores, but also communities and lives. Hence, intermediaries can organize the variety of needles found inside the haystack and help people understand the purpose of each. Collaboration is another way for government to engage the public and to create useful information. Some attendees suggested using new technologies to give the public a direct role in locating, organizing
9
Transparency in Texas: Beyond Raw Data and translating government data. Several suggested crowdsourcing as an innovative way to get people to tag data, rate its usefulness or even, in some cases, rate its validity. Using collective efforts could be a way to build a map and make sense of data that currently exists in formats or in locations that are not as accessible as possible. Evan Smith also stressed the importance of intermediaries such as the Tribune in collaborating with other intermediaries and with the public. With the correct needle for the job, citizens can work together to sew the quilt of effective government.
i
Barack Obama, "Transparency and Open Government: Memorandum for the Heads of Executive Departments and Agencies ," The White House, last modified January 21, 2009, http://www.whitehouse.gov/the_press_office/ TransparencyandOpenGovernment. ii LBJ School of Public Affairs -‐ Policy Research Project, Texas Financial Transparency: Open and Online, ed. Matt Hartman, Claudia Montelongo, and Jennifer Quereau, May 12, 2010, http://www.utexas.edu/lbj/cpg/docs/prp.pdf. iii Jon Gant and Nicole Turner-‐Lee, Government Transparency: Six Strategies for More Open and Participatory Government (n.p.: The Aspen Institute and the John S. and James L. Knight Foundation, February 2011). iv James K. Scott, "'E' the People: Do U.S. Municipal Government Web Sites Support Public Involvement?" Public Administration Review (May-‐June 2006):341-‐353. v Tony Carrizales, "The Internet Citizenry: Access and Participation," Public Administration Review 69, no. 2 (March-‐ April 2009):350-‐353. vi Caroline J. Tolbert and Karen Mossberger, “The Effects of E-‐Government on Trust and Confidence in Government,” Public Administration Review 66, no. 3 (May-‐June 2006):354-‐369. vii Jon Gant and Nicole Turner-‐Lee, Government Transparency: Six Strategies for More Open and Participatory Government (n.p.: The Aspen Institute and the John S. and James L. Knight Foundation, February 2011): 20. viii Sherri Greenberg and Angela Newell, "Transparency Issues in E-‐governance and Civic Engagement," in Active Citizen Participation in E-‐Government: A Global Perspective (n.p.: IGA Global, Forthcoming January 2012). ix Jon Gant and Nicole Turner-‐Lee, Government Transparency: Six Strategies for More Open and Participatory Government (n.p.: The Aspen Institute and the John S. and James L. Knight Foundation, February 2011): 25. x Ibid. xi Appendix A contains a detailed description of transparency challenges in the TEA and recommendations. xii Appendix B provides a brief literature review of best practices in website design and more information on our version of the LBB website. xiii Appendix C is a more in depth analysis of the City if Kyle. xiv Appendix D is a more in depth analysis of the City of Manor. xv LBJ School of Public Affairs -‐ Policy Research Project, Texas Financial Transparency: Open and Online, ed. Matt Hartman, Claudia Montelongo, and Jennifer Quereau, May 12, 2010, http://www.utexas.edu/lbj/cpg/docs/prp.pdf.
10
Transparency in Texas: Beyond Raw Data
ORIGIN OF DATA There are many political, cultural, and economic forces propelling the demand for online data that is comprehensive and meaningful to the public. Government agencies need to think explicitly about the data they gather, the reports and correspondence they generate, and how they make this available to the public in a useful way. Agencies need an organized policy or system relating to online transparency in terms of collecting, reviewing and publishing useful data. Two reasons warrant special discussion: budget cuts and technological expectations. The global financial crisis and recession that began in 2007 have had a profound effect on revenue management at all government levels. Governments face mounting pressures to make deep funding cuts for public agencies, while minimizing the influence of those cuts on frontline public services. Hence, powerful and detailed financial management information has become, to an even greater extent, a key commodity in running and overseeing public agencies. Simultaneously, the notion persists that nearly any aspect of a government operation can be quantified and displayed electronically, granting users the freedom to decide what they want to know.xvi, xvii Historically, management information has been created for the use of managers, and the number of managers has been restricted to those people directly in an employment or supervisory role in a government department. The 1966 passage of the Freedom of Information Act (FOIA) in the U.S. did not fundamentally alter this system: while the public was given a right to review internal information, in practice, it requires considerable underlying expertise to know what to ask for in the first place.xviii We now confront the question of whether the advent of the Internet and the anti-‐spending political wave will bring about a revolution where FOIA did not. As constituents seek a more comprehensive understanding of government finance, they are finding a system of data generation and reporting that is not geared toward their understanding. The prevailing principle of online transparency is the idea that if data is presented in a raw, machine-‐ readable format, then someone could understand the agency’s work and draw conclusions.xix The problem is that the internal agency data has not been created with constituents in mind. When only a handful of highly invested people are reading about and managing the money in a government program, reports on that program’s budget, accounts, and financial strategy will not naturally contain much context. Users need information on the meaning of terms, the structure of what is and is not reported, and the history of funding for different aspects of the program. In many ways, the push for online transparency has amplified this tendency to publish data without context or explanation by seeking to set data free, stripped of the moorings of its original report, so that it can be rebroadcast to a much wider audience and analyzed more broadly. When governments release raw data sets, it is essential that the fields, labels, and categories be representative of the data being collected. Often, data sets are uploaded to a government website with little or no explanation of what the fields represent, and users are left to draw their own conclusions. In our Texas Education Agency case study (see appendix A), we explore how a myriad of confusing data fields makes the agency’s data unintelligible. Hence, if the data cannot be used meaningfully, then it is not truly transparent.
11
Transparency in Texas: Beyond Raw Data Additionally, data that is unreadable, unusable, unappealing, or unwanted by the public does little to contribute to improving a government’s transparency. One of the major inhibitors to making government data easily accessible and meaningful is the lack of appropriate data formatting. Recently, the amount of data available on government websites has increased, but much of the new data has come in static file formats like PDF. Including valuable data in PDF format rather than exportable formats like Microsoft Excel creates an additional, potentially lengthy, step for those wishing to analyze government data. Government datasets can range in size from just a single measure reported by a few well-‐defined jurisdictions to massive tables combining millions of data points for thousands of entities across dozens or hundreds of metrics. A professional tool created to handle one size of database and to present that data in a way that the public can efficiently use and readily understand may not be a good fit for a database of a different size.
Government agencies’ failure to report and present data well is not always a function of the size or complexity of the database. Rather, it may be an indication that there is simply a mismatch between the approach or technology tool being used and the real needs of the users. Recovery.gov, for instance, is built around a platform that allows users to zoom in to different levels of spending data reports. In other words, the website presents charts and visualizations that are more general at one level, and more specific at a lower level: the site adapts as the user explores. In contrast, the Texas Education Agency employs a fire-‐hose of raw data at the most specific level, which when fired directly at the user, can overwhelm a non-‐expert and impede public understanding (refer to appendix A). Even when datasets are small, a mismatch can cause problems. The City of Manor, Texas, a national star in the field of technology and transparency projects, once received a very low score for its publication of city finance data.xx The sophisticated systems that Manor used to generate public input and advertise e-‐
12
Transparency in Texas: Beyond Raw Data services and community technology initiatives were not well suited to the simple task of posting the city’s budget in an accessible or interesting format. The formulation of an online transparency and open data process does not occur at a single point in time. Instead, agencies must expect their process to evolve as users’ preferred technologies change, and as public expectations about the time, place, and manner of data publications shift. Large, one-‐time outside contracts to deliver a website or a database system may not be cost-‐effective or deliver high-‐ quality service in the long-‐ term, if frequent changes occur in the type of data that the agency gathers or how the agency must publish data to be the most useful to the public. Agencies that do not have a large in-‐house IT operation may find themselves at a disadvantage in meeting the emerging standards of the open data movement. This is true even if they have been provided large grants or contracts at one point to make a great deal of information publicly available. The Texas Legislative Budget Board, discussed in Appendix B, may be one such example. The LBB provides crucial and popular information for state-‐level decision-‐makers, but its lack of internal capacity to update its site’s design and functionality on a regular basis inhibits accessibility. In cases where the government is unable to continuously adopt new technologies, we can rely on intermediary organizations to prepare data for public consumption. If agencies are able to collect data in a way that is meaningful and lends itself to being easily exported, advocacy and media organizations can then use this data to present a more digestible picture, complete with interactivity and context. The key part of this process, though, is that government agencies data must at least be organized in a way that does not need to be decoded.
xvi
Michael Cloud, “The Bounty Hunter Approach: Removing Waste from Government.” The Center for Small Government (website), 2005. http://smallgovernmentnews.com/bounty.htm. Accessed April 29, 2011. xvii “Obtaining Records from Federal Agencies Using the Freedom of Information Act.” The First Amendment Center. http://www.thefirstamendment.org/foia.html#7. Accessed May 1, 2011. xviii “Requester Best Practices: Filing a FOIA Request.” Office of Government Information Services, National Archives (U.S.). http://www.archives.gov/ogis/guidance/best-‐practices-‐foia-‐request.html. Accessed May 1, 2011. xix Mike Ellis, “Pushing MRD Out from Under the Geek Rock.” Electronic Museum (website), July 13, 2009. http://electronicmuseum.org.uk/2009/07/13/pushing-‐mrd-‐out-‐from-‐under-‐the-‐geek-‐rock/. Accessed May 1, 2011 xx Dustin Haisler, Remarks at the TXGov2.0 Conference, January 28, 2011. http://www.ustream.tv/recorded/12305333 (video -‐ cited section begins at 1min45sec.
13
Transparency in Texas: Beyond Raw Data
PRESENTATION OF DATA While access to well-‐formatted raw data is an important element of transparency, it is just as important for governments to provide concise visual representations of data to make it meaningful to users. One of the biggest roadblocks to high quality visual representations of data is the poor design of many government websites. Despite the Obama administration’s Open Government Initiative and other politicians’ ambitious efforts to create more modern and user-‐friendly government websites, they have improved little recently. Many government websites have outdated designs with a look and feel from the late 1990s, and modern government sites like Recovery.gov only make older government sites appear even more obsolete. Users are accustomed to websites with smooth graphics, drop-‐down menus, and other design elements that break up sterile bocks of text. The US Weather Service websitexxi provides a good example of outdated design elements:
However, an aesthetically well designed website by itself is not the only key to presenting visually appealing data. Many government websites lack sufficient images, charts, graphs, and other visual aids to engage users and help them better understand the significance of the data presented. A website full of text-‐based reports and informational materials is certainly helpful, but images communicate the data’s significance more deeply and concisely. Even more helpful than static visuals is an increased emphasis on incorporating reporting software that communicates with government back-‐end databases. This would give users an interface through which they could fill out a form to create custom reports based on information held in databases. Currently, few government websites enable users to view data in visual formats, limiting data analysis to only those with technical knowledge.
14
Transparency in Texas: Beyond Raw Data The State of California’s data sitexxii is a good example of one that combines clean organization and visual presentation with easily searchable databases:
Organization and layout are important elements of a well designed website. Currently, many government websites categorize their content according to classifications that are confusing and unintuitive. Documents pertaining to a particular topic may be located in multiple places, making it difficult to locate all of the pieces related to a topic. The haphazard organization of many government websites can cause confusion and reluctance to use the website. Some government websites contain data that is not easy to find without going through a multi-‐step process. This includes sites that incorporate reporting software, which often requires some knowledge of how the underlying database works to create the desired report. Sites should contain instructions, including “how-‐to” guides to assist users in successfully navigating websites and locating the appropriate content. Many documents on government websites lack any description or summary of contents, presenting a barrier for many users with limited time. Without structures that make it easy to find information and quickly discover the relevance of documents, it may be too burdensome for the public to use these sites.
xxi
http://www.weather.gov (accessed May 2, 2011)
15
Transparency in Texas: Beyond Raw Data xxii
http://data.ca.gov (accessed May 2, 2011)
16
Transparency in Texas: Beyond Raw Data
ENGAGING THE PUBLIC
Even when data and government websites achieve a high level of usability by presenting meaningful data in a way that is visually attractive and easy to use, their effect is limited by the extent to which the public actually engages with the information. Any assessment of the quality of government data transparency must include an evaluation of how the data will be used. This presents a challenge for open data advocates, as citizen participation has continued to decline across several metrics over the past several decades.xxiii The Internet promises to promote active civic engagementxxiv, but many government agencies and policy-‐focused nonprofits have shown a lack of understanding of how to use networked technologies to engage their constituencies effectively. For example, a content analysis of civic websites aimed at youth in the United Kingdom showed that many of these sites fell far short of providing the basic elements needed to promote engagement.xxv Because of their top-‐down, hierarchical structure and lack of interactivity, these sites have failed to promote efficacy in their users. This lack of understanding of the Internet’s potential power to engage signals a disconnect between those promoting best practices for government data and those who are in the best positions to transmit that information into action. As our questionnaire responses show, many of those working for civic-‐ based groups in Texas use government data for internal purposes (research, trend tracking, etc.) but are less likely to use it as the basis of their communication with constituents (advocacy, creating media products). In contrast to these policy-‐oriented nonprofits, many news organizations and transparency groups around the country have already taken up this task. Most notably, ProPublica, a nonprofit investigative newsroom, and the Sunlight Foundation, a nonprofit focused on government transparency, have made open government data central to their operations.
ProPublica’s presentation of the Recovery.gov spending data is a leading example of the potential intermediary organizations have to translate open government data into usable products. Its Recovery Tracker tool is just one example of these products. xxvi The data and tools section of its site shares equal prominence with its investigative reports, many of which are based on government data sets. The Sunlight Foundation’s cutting edge Sunlight Labs initiative is another product that makes accessing government data easier. While its tools are mostly targeted at developers, many of its applications have been used to further its own information dissemination mission, through its blog, social media channels, and other publications. The Texas Tribune has applied many of the same techniques employed by ProPublica to report on policy in the state of Texas. Also working from the nonprofit, policy-‐focused model, the Texas Tribune offers a data library, which includes more than sixty data sets and tools.xxvii Two thirds of the Tribune’s web traffic is directed at its data library.xxviii While nonprofit, public interest, and journalism organizations are creating products to make government data accessible, their audiences are still very small.xxix In many cases, they reach audiences that are already highly politically and technically sophisticated. To grow their audience, and hence, have
17
Transparency in Texas: Beyond Raw Data a larger effect, these organizations must envision a role beyond information dissemination. To be sustainable in this media environment, new and existing journalism organizations must see themselves as advocates for civic engagement, media literacy, and open access to information. Because of their ability, as a whole, to reach wider and more diverse audiences, policy-‐oriented nonprofits can increase the number of people who are aware of the value of open government data and the effect it has on issues they care about. Our questionnaire, answered by representatives of these nonprofits (see Appendix G) showed that while many of them were accessing government data, they were mostly using it for internal purposes such as drafting recommendations, doing research, or tracking trends. Many fewer respondents said they used government data as part of their advocacy or media strategy. We recommend that policy-‐oriented nonprofits also make interaction with and dissemination of government data central to their advocacy and communication strategies. These organizations have built-‐in audiences that have a demonstrated interest in policy issues. By demonstrating the value of government data in lobbying for their constituents’ causes, these groups can create a cycle whereby successful lobbying proves the case for open government data, which encourages the public to advocate for improved data standards, which will ensure more effective lobbying efforts. Policy-‐oriented nonprofits also have expertise, access to human networks and money that individuals do not have. These organizations are able to use those resources to make meaning of government data and aggregate feedback into a collective response.
xxiii
Robert Putnam, "Tuning In, Tuning Out: The Strange Disappearance of Social Capital in America," in Controversies in Voting Behavior, 4th ed., ed. Richard G. Niemi and Herbert F. Weisberg (Washington DC: Congressional Quarterly, 2001). xxiv Yochai Benkler, The Wealth of Networks: How Social Production Transforms Markets and Freedom (n.p.: Yale University Press, 2006). xxv R. Gerodimos, "Mobilising Young Citizens in the UK: A content analysis of youth and issue websites," Information, Communication & Society 11, no. 7 (2008):964-‐988. xxvi Jennifer LaFleur, Dan Nguyen, and Sydney Lupkin, "How Much Stimulus Funding is Going to Your County?: ProPublica Recovery Tracker," ProPublica, accessed April 27, 2011, http://projects.propublica.org/recovery/. xxvii "Data Pages," The Texas Tribune, http://www.texastribune.org/library/data/. xxviii Mallory Jean Tenore, "Texas Tribune databases drive majority of site's traffic, help citizens make sense of government data," Poynter., last modified March 2, 2011, www.poynter.org/latest-‐news/top-‐stories/121281/ texas-‐tribune-‐databases-‐drive-‐majority-‐of-‐sites-‐traffic-‐help-‐citizens-‐make-‐sense-‐of-‐government-‐data/. xxix Dotinga, Randy. "Nonprofit Journalism on the Rise."Christian Science MonitorFebruary 12, 2008. http://www.csmonitor.com/USA/2008/0212/p03s01-‐usgn.html (accessed April 30, 2011)
18
Transparency in Texas: Beyond Raw Data
RECOMMENDATIONS
In the previous sections of this report we outlined the reasons for and main challenges to improving transparency of financial data. We have divided the online transparency process into three steps: how the data is collected and formatted, its online presentation, and using that data to engage with the public. In an ideal setting government agencies would be able to collect and produce all data with a public audience in mind, rather than strictly for internal reporting demands. This data would be presented to the public by both government entities and intermediaries using creative online technologies that are clear and easy to use. Finally, the public would be drawn to these websites and provide meaningful feedback that is used by the government agencies in decision-‐making. Reaching this ideal state of transparency should be a goal for long-‐term transparency strategies, in which presenting information to the public is the primary objective. We recognize that most agencies and organizations are not currently in a position to revamp their entire data collection systems and invest in the newest technologies. The recommendations presented in this section can be incorporated into current transparency efforts, often without incurring new costs.
Origin of Data In the previous sections of this report we outlined the reasons for and main challenges to improving transparency of financial data. We have divided the online transparency process into three steps: how the data is collected and formatted, its online presentation, and using that data to engage with the public. In an ideal setting government agencies would be able to collect and produce all data with a public audience in mind, rather than strictly for internal reporting demands. This data would be presented to the public by both government entities and intermediaries using creative online technologies that are clear and easy to use. Finally, the public would be drawn to these websites and provide meaningful feedback that government agencies can use in decision-‐making. As a long-‐term goal, agencies should move toward automatic online publication of all public financial data and supporting documents not subject to privacy and security constraints. Agency staff should work to incorporate public accessibility of data used in their jobs while adhering to privacy and security standards. Short of restructuring agency data collection procedures at every level, which currently is neither feasible nor affordable, we provide recommendations to use current resources and move toward a more transparent system by enhancing data publication standards and nurturing a culture of openness. • Make data comprehensible to users outside the agency. This can be accomplished by an automated process translating internal agency files into reports or datasets that are useful to the public or by having agency staff translate the codes, jargon, and other quirks in internal agency files into reports, tools, and datasets that are useful to the public. Automating this translation process could potentially be a one-‐time expense and may avoid using large amounts of staff time. (Appendix A discusses the benefits and shortcoming of the LONESTAR reports system, a project of the Texas Education Agency.) •
Provide datasets in a format that is: § Machine-‐readable (such as an excel or comma-‐separated value-‐.csv file, not PDF) § Exportable
19
Transparency in Texas: Beyond Raw Data § §
Editable by the user in commonly available program Friendly to the creation of informative charts and graphs
If an agency is serious about providing data directly to the public in a format that most citizens can understand, without the need for an intermediary interpreter, then those tools for visualizing (chart and graph creation) become especially important. This is one success of Recovery.gov, which generates easy-‐to-‐interpret visualizations of most pages of stimulus spending data automatically without the need for significant user technical knowledge. •
Reach out to communities, both inside government and intermediary groups, to assess existing capacity to interpret the agency’s raw numbers. We recognize that not every public entity is well suited to building or running a sophisticated data website directed at public consumers. Hence, there is an important opportunity for intermediaries to take on much of this work, using the raw data provided by the originating agency. Agencies can then use that awareness to determine whether or not it is vital for the agency to invest in a sophisticated publication system that ordinary users can view in detail and customize directly on the agency’s website. Armed with information about existing capacity, a publishing agency also can make informed requests to support agencies (such as the Texas Department of Information Resources) for assistance.
•
Identify an individual currently on staff to champion public usability and open publication.1 One reason transparency efforts may fall short is that transparency is simply forgotten. Designating someone to ask questions such as, “Is this data in a machine-‐readable format?” or “Will a member of the general public understand these categories?” will help the agency remember to make adjustments and enhance accessibility.
•
Make a commitment to review and renew processes at regular intervals to respond to new demands and opportunities for public understanding of agency activities. Even the most transparent agency that combines an enthusiastic agency culture, a high quality public website, and effective relationships with partners inside and outside government will stagnate over time if it does not make the effort to stay up to date with constituent needs.
Presentation of Data The following recommendations describe website presentation elements that can help governments and intermediary organizations overcome challenges.
•
Pay attention to the visual appeal and layout of online resources, so that people can easily and quickly find their way around the website. Currently, the online community has very high expectations. Simply dumping information on a plain-‐looking website is no longer acceptable for the modern, tech-‐savvy consumer of online data. With the development of sleeker and more streamlined websites throughout the public (recovery.gov, data.gov) and private (espn.com, facebook.com) sectors, people have come to expect that not only will any information they want be online, but that they will be able to find it easily. In a world where many people have easy and free access to the Internet, the cost of
1
We of course recognize the need to protect privacy rights for individuals, and so in all cases “public financial information” refers to that information which is subject to FOIA legislation.
20
Transparency in Texas: Beyond Raw Data obtaining online information is a person’s time. For most people, the amount of time they are willing to spend sifting through confusing website configurations is extremely small. The assumption that information will be easy to find means that people will quickly navigate away from a website that does not meet their visual presentation expectation. Specific tools for achieving efficient navigation include:
§
Visually direct users to important site elements by using images or headings. Like a newspaper, a website can adopt a layout that draws the attention of visitors to content through headlines and pictures. They serve as a guide to the page, inviting visitors to view content in the order desired by the page designer. If a lead picture, headline, or graphic is effective, its prominence will draw attention to a single element on the page.
§
§
§
§
Use sub-‐headings, pictures, and graphics to provide visual cues. These secondary visual cues direct visitors to additional areas of the site once they are finished absorbing information from the primary news story or data set. Overall, a website that incorporates visual cues effectively both attracts and retains the attention of visitors. Divide content into several broad categories using a main menu. Menus contribute added value when they “drop down” to show visitors exactly where they can find information on the site while also providing a clean interface that is not cluttered with confusing links. Our re-‐design of the LBB website took this concept into account, since the current website contains a large number of poorly categorized hyperlinks, making site navigation extremely difficult. Ease of navigation is another best practice for website design. Ensure menus are consistent on all parts of the website. Visitors should see the same main menu no matter what part of the site they are visiting. This consistency makes it easy to navigate away from the home page and still find information on other parts of the site. Have a search bar to provide an additional mechanism for navigation. Users seeking specific information that is not contained in the main menu or its subcategories need an easy tool to search. This helps reduce the number of people that cannot find the information they need.
•
Design websites in a way that attracts and holds the attention of users. To engage with the typical, impatient citizen, governments need to design their websites as if they were selling a physical product in exchange for the time of site visitors. Therefore, website design is almost as important as content quality when it comes to strengthening the online relationship between governments, intermediary organizations, and the public.
§ §
Post short videos that convey main ideas and summarize information. Make content accessible through social media sites such as Facebook, Twitter, and Reddit. This will also allow viewers to share information with their networks.
21
Transparency in Texas: Beyond Raw Data §
Facilitate feedback from visitors to improve government performance. One purpose of pursuing online financial transparency is to make government more accountable to constituents. This accountability cannot occur if citizens do not have a way to communicate with elected officials and government agencies. Comments sections are good ways to facilitate feedback, but may not always solicit intelligent analysis of government activity. Another means of soliciting feedback could be to allow visitors to sort data and create their own charts to analyze information. These charts can then be shared between visitors using comments sections or social media. This allows citizens to find the data they need and also informs governments of what data is most in demand.
Manor, Texas excelled at making sure citizens felt that their comments were taken seriously by creating a public forum where comments could be viewed, implementing some citizens suggestions, and rewarding citizens for their involvement (see Appendix D).
Engaging the Public
Based upon our evaluation of the challenge of engaging the public, we are essentially recommending that news organizations see themselves more in the role of advocates and that policy-‐oriented nonprofits and other advocacy groups recognize their ability to create new communication channels that supplement news organizations. The following recommendations address specific ways in which these roles can be achieved. News Organizations •
Participate in the creation of media literacy programs. The future business model of news organizations—especially those such as the Texas Tribune and ProPublica that are specifically focused on policy reporting—depend upon consistently growing audiences. One way to create that demand is to demonstrate to the public the importance of information and to give people the skills to find it. Many schools include media literacy programs in their curricula, and those programs could benefit greatly from the active participation of news organizations.
•
Form partnerships with groups that promote civic engagement. News organizations depend upon an engaged public for their survival, and so developing partnerships with groups that promote civic engagement will help drive public demand for policy information.
•
Assert an active role in the community through public events and programming. News organizations have the ability to harness public attention in much the same way that libraries and museums do—through developing issue-‐focused programs and hosting live events. These activities complement the primary role of news organizations—as information providers—as the output from these events and programs can be turned into content and news products.
Policy-‐Oriented Nonprofits
22
Transparency in Texas: Beyond Raw Data •
Understand the power of social media. While social media channels can be effective in broadcasting information, the added potential for the audience to provide feedback based on that information-‐-‐and engage in conversation with organizations-‐-‐is the true value of these new communication tools. Sites like Twitter and Facebook allow organizations to communicate with their constituencies on a more personal and reciprocal level, which can be powerful in promoting self-‐efficacy and continued engagement.
•
Provide interactivity. Websites are not harnessing the full power of the Internet if they are simply text-‐based, static products. Many tools exist to promote active engagement with information (see Appendix E), letting users do something while they are on a website. Interactive tools promote learning while keeping users’ attention for longer periods of time.
•
State clear goals. Many civic websites fail to clearly state the purpose of the campaign for which they are trying to galvanize their audience. To effectively engage a constituency, the urgency of the mission and the metrics by which success can be measured must be presented clearly and succinctly.
•
Show the effects. The most effective online issue-‐based campaigns communicate to users the outcomes of their actions by highlighting the successes, made possible by user engagement, of past campaigns. Connecting an individual’s contribution to a concrete change promotes efficacy and encourages citizens to continue their engagement. For these recommendations to be effective, they must be fully integrated into the operating mission and philosophy of the organization. These new methods of communicating with an audience reflect a new social paradigm, which assumes interactivity and two-‐way engagement. News organizations and issue-‐oriented nonprofits that continue to operate as closed boxes, broadcasting information to citizens as they see fit, will slowly lose relevance as new kinds of institutions emerge to fill the gap between citizens and government.
23
Transparency in Texas: Beyond Raw Data
CONCLUSION In light of the financial crisis and constraints on public resources, the impetus to make public financial data available online has never been greater. In this era of deficits and budget cuts, the public is deeply concerned with the efficient spending of their taxpayer dollars. The Internet and online tools provide unprecedented potential for online transparency to connect the public to financial data in a meaningful way. While online transparency could once be described as the availability of government data online in open and easily accessible formats, machine-‐readable and exportable data alone do not stimulate public understanding and participation. Online transparency must foster understanding, provide context, and promote interactivity in order to truly engage the public. The shift we are describing here is one from e-‐ government—government data on websites— to e-‐governance—truly virtual government that promotes civic engagement. This report offers a variety of recommendations and best practices that not only provide a roadmap for governments to achieve greater transparency, but also clear steps for organizations outside of government to take a role in encouraging civic engagement through use of government data. As governments gather and share financial information online, they should consider how comprehensible it will be to the public. Government agencies should continue to provide accessible, machine-‐readable, and exportable data, but they should also provide context, explanations, or visuals to help users make sense of the information. By prioritizing online transparency and being adaptable to changing technologies and public expectations, government agencies can promote civic engagement with limited resources. We must stress that government is not alone in this process. Intermediaries like news organizations, nonprofits, and advocacy organizations play an important role in building upon government efforts to improve online transparency and engage the public. By using social media and other interactive web tools, intermediaries have an opportunity to educate, engage, and partner with their constituencies and other organizations. Intermediaries are vital to an effective feedback loop that provides the public with government data, enables mechanisms for communication, and ensures that the government understands public needs and priorities. This new social paradigm, a two-‐way interactivity between the government and public, is the objective of online transparency. While the scope of this report is on short-‐term, cost-‐neutral recommendations to improve online transparency of financial data, we recognize the potential for further study. For example, more research is needed on long-‐term recommendations to improve online transparency—not only for financial data, but all public information. The evolving nature of technology ensures that the study of transparency will continue to change.
24
Transparency in Texas: Beyond Raw Data
Appendix A. TEXAS SCHOOL DISTRICT FINANCE DATA Summary Various entities, including the Texas Education Agency (TEA), collect and publish large volumes of detailed data on Texas school districts’ finances. Measured against basic standards of the open government movement (such as data being machine-‐readable), these publications would appear to represent a significant achievement in online transparency. However, the complexity, lack of standardization, tendency to duplicate or leave out information, and incoherent structure of publication across multiple websites means that even a savvy member of the public is hard pressed to extract any meaning from the data. If this information is to be useful to stakeholders in public education, the TEA must reconsider its system of managing data publication, and must begin promoting public consumption as a key consideration in creating data sets. Purpose This report addresses the thorny issues of translation, interpretation, and publication that hamper the practical use of raw data. As both political culture and technology culture adapt to the 21st century, many people who are not traditional “insiders” at government agencies – such as legislative staff, advocacy groups, and concerned voters – are taking on a more direct role in analyzing those agencies’ activities, priorities, and finances. Our goal is to understand the gap between potential new ways to publish finance data and the usefulness of current publications. There are significant barriers, however, to understanding just how big this gap may be. These barriers result from a limited understanding of factors such as: • citizen apathy; • the resources and priorities of media organizations that act as interpreters of government information releases; • the availability of experts who develop and use software and statistical programs that parse data; and • differences in how government data is used by the public at the local government level, as opposed to at the federal and state levels. These difficulties are not all new: bond-‐rating agencies have a long history of attempting to boil down large sets of financial information into a simple, universal credit rating of an entity. That process, however, has been narrowly focused on a specific audience and traditionally closed off from general public view and understanding. The needs of data users are constantly changing as larger groups of people begin to access government data. Determining just how much transparency efforts fall short remains a moving target. Background of Texas School District Finance Data For the purpose of this report we sought a case study that is characterized by: • plenty of raw numbers; • widespread citizen concern about the subject; • publishing agencies that have the capacity to produce complex datasets; and • data that describes entities easily compared to one another in an apples-‐to-‐apples fashion.
25
Transparency in Texas: Beyond Raw Data To focus on data publication management challenges we researched the data published regarding Texas independent school districts’ (ISDs) finances. With more than 1,000 entities, school district level financial data meets the criteria outlined above. In Texas, school districts are almost all independent local entities, carrying out comparable work. Because both direct state funding for school districts and inter-‐district fund transfers are governed by complex, state legislated formulas, the Texas Education Agency gathers volumes of raw information about districts’ budgets and spending and aggregates this data in a standardized way. Additionally, school finance and investment in education interests numerous constituents who do not otherwise consider themselves to be highly political or enthusiastic about government data. This is particularly true for school district bond referendums, which draw the public into important decisions about a government entity’s finances. Voters’ rejection of several school district bond proposals in 2010 highlights stark differences of opinion regarding district finances between a districts’ managers and the public. This case study examines 33 distinct data products that are available from official government websites. Rather than analyzing how data is generated in the first place in local school districts, this report draws lessons from the centralized locations that compile disparate local information, which they provide to the public in a standard form. However, we recognize that most stakeholders are interested in only one or a few districts and will seek information at the local level first – by checking a school district’ website. Therefore, we also make recommendations about how local data publishers can find new ways to present information from a central source like the TEA, such as a comparison of local debt burden with that of other districts. Analysis of Technical and Organizational Challenges While the existence of dozens of finance-‐related data publications presents a very complex picture to an outside observer, much of the raw information that feeds these publications is compiled under a single umbrella at the TEA known as the Public Education Information Management System (PEIMS). This includes every piece of demographic, performance and management information collected by the state about districts, in addition to budget and accounts figures, which is all translated into numeric codes for central storage. PEIMS is a very powerful tool for the TEA, as it develops many different reports, but there is no easy way for an outsider to navigate the encoded information. Users looking for specific information, such as physical education spending per pupil in a specific district, could spend hours sifting through data without success. Even if a member of the public knew, in technical language, exactly what to search for (“data standards” for PEIMS data dictionaries and code-‐translation tables), there are more barriers to using information in the PEIMS system. Currently, guides on how to translate PEIMS data exist, but users who search for PEIMS using a search engine are not directed to the guides.xxx Essentially, users of PEIMS raw data require a map just to find the map. If a user can find the TEA guides to PEIMS interpretation, she is confronted with two separate websites for code translation. One of these contains numerous, lengthy MS Word documents full of code translations that provide no real understanding of the database information.xxxi The other presents similar information in html form rather than Word. These guides are so full of jargon that they fail to serve any useful purpose.xxxii The original target audience for the PEIMS system was, of course, professionals who administer and submit items into the system directly. However, the purpose of the information is to meet state funding formula requirements and to inform district officials and state lawmakers about school funding and demographics. Even within these fairly narrow professional confines, the system is too convoluted to add value for all but the most sophisticated users.
26
Transparency in Texas: Beyond Raw Data Instead of making PEIMS into a one-‐stop shop for easy data analysis and reporting, the TEA produces a many individually tailored reporting systems to answer specific questions, usually matched to a statutory requirement in the Texas Education Code.xxxiii Along the way, certain pieces of the raw PEIMS file are separated, recombined, and offered as a direct download in various formats. Some of these that are relevant to district finances are listed in one somewhat obscure page in the TEA website; however, there is no consistent pattern on the TEA website of linking to this central page when users are seeking district finance data.xxxiv The tables accompanying this section give a survey of some of the tools, a few of which are addressed in more depth below. The PEIMS Financial Data Downloads The TEA provides reports on school district finances that advertise the ability to make comparisons between districts on various measures, but the reports are generally difficult for an outsider to use and lack sophisticated visual renderings of the results for comparison. The open government movement’s answer to such situations has generally been simple: “give the raw figures,” thus enabling a person with basic spreadsheet skills to locate a row, column, or cell and create a chart or basis for comparison. The case of downloads for PEIMS Financial data illustrates the limits of this philosophy.xxxv The financial data for ISDs for even a single budget for a single year contains over a million rows of numbers, which cannot even be opened in MS Excel. Using a program such as MS Access, a user only finds a sea of numeric cells with no obvious way to re-‐convert them into knowledge that a human can interpret. Numeric values are assigned to the names of the districts themselves, making it impossible to identify the district the data describes. State Aid Reports The most widespread method the TEA uses to publish ISD finance data is custom html pages. In this format, the TEA server (often reading a database through a statistical program such as SAS, as shown at the bottom of the readouts) pulls a specific number from the state’s files and displays it as text on the reader’s Internet browser. This restricts user ability to compare two pieces of statistical information drawn from different geographic areas, different budget codes, or different measures of financial health. Instead, these narrow reports fulfill mandated state collection of information that is used to plug into pre-‐existing funding formulae and other decision-‐making structures. The question of how to separate out pieces of the larger PEIMS database to accomplish a new, creative task is critical. If outside observers are interested in comparing among districts’ expenditures or debt per student or per dollar of assessed property value or per deviation from the mean standardized test score, for instance, they are out of luck unless the publication system administrators have already built in this particular calculation. The publication of TEA State Aid Reports illustrates this problem. Users interested in understanding how much money districts pay into and/or get from the state may find a page in which the menu of choices for looking up data are not intuitive (“Chapter 41 Cost of Recapture Report…IFA Payment Report…EDA Eligible Debt Service”) and are not accompanied by a plain-‐English explanation of what each field means.xxxvi Many of the data lookup tools on this page require every query to be made using a single school district’s identifier code, rather than its name or a format that would allow cross-‐district comparison. Others tools, inexplicably, allow selection from a list of districts. And, once a number or set of numbers is retrieved from the PEIMS system, the output page generated for a user does not contain links to information that could help a user interpret what she is seeing.
27
Transparency in Texas: Beyond Raw Data In the end, the questions that drive the public to the database – “Is this district getting a fair shake from the state funding system?” “Is this district in too much debt for its size or wealth?” “Is this district at risk of financial exigency?” – are not answered. xxxvii Snapshotxxxviii and LONESTAR The publicly available database downloads and data readouts from the TEA mostly seek to fulfill a minimum technical standard for the possibility of outsider use while failing to make this workable in practice. There are two examples of TEA-‐driven data projects, though, which appear to be more focused on this “retail” distribution of more polished information to stakeholders. Both, however, suffer from serious shortcomings in terms of prominence, usability, and formatting. Snapshot is a TEA website that operates under the apparent assumption that presenting less information in a more direct format, with fewer technical details, will create a better understanding for less sophisticated users. It allows a user to search for a district by name and provides, for a particular year, a text readout of some basic statistics about student numbers, demographics, and overall spending per student. It also divides expenditures into very broad categories such as “instructional” and “central administrative” to address questions of overhead. There is no simple way to compare between districts other than to manually compare them to the readout, although a user can find a statewide report of average figures. Snapshot also sacrifices the ability of the PEIMS system to give a detailed look at a district’s spending in a certain area, which in theory should allow for time series comparisons and cross-‐ sectional comparisons. Regardless, Snapshot is located away from most of the finance data reports published by the TEA, and there is no obvious route to finding it for a new user who is simply looking to find data about a district to answer commonsense questions. LONESTAR, a joint project between the TEA and the Texas Higher Education Coordinating Board, has similar challenges to those of the Snapshot system. Hosted at its own website that is poorly linked by other agencies and divisions of the TEA, LONESTAR is an attempt to create a more user-‐friendly interface for certain parts of the PEIMS dataset.xxxix There is more emphasis on visually appealing formatting and presentation. The result, again, is that the ability of the user to truly “explore” information about a district is severely limited, and the ability of a user to draw comparisons between districts is practically non-‐existent. LONESTAR does produce summary reports for five-‐year periods for each district. A diligent user could copy out key variables one at a time from various reports to build a comparison. Unfortunately, the server appears incapable of rendering these summaries in html, forcing the user to download a PDF of each one. This limitation hampers the ability of outsiders to creatively use the information, although the PDFs do contain visually striking charts and graphs, which are absent elsewhere in school district finance data publications. FIRST Several TEA publications seek to categorize the districts’ financial performance using criteria that go beyond the bare numbers. The most prominent is the Financial Integrity Rating System of Texas (FIRST).xl In FIRST, districts’ “integrity” is aggregated into a very simple tier of superior, above standard, standard or substandard performance. Unfortunately, FIRST suffers from several major shortcomings that inhibit its usefulness to the public and also having the potential to cause misunderstandings: • The worksheet used to calculate FIRST performance combines different types of measures, which may cause confusion about what is actually being described in the rankings.xli The district’s accounting and monitoring integrity (whether their numbers match up) is mixed with their financial health in terms of assets and cash (whether they are running a deficit). On top of
28
Transparency in Texas: Beyond Raw Data
• •
•
•
these financial numbers, the calculation also includes student outcomes. Therefore, a district in financial trouble but which tracks its woes effectively may appear similar to a district in financial health but poorly managed accounting reports. The public-‐access FIRST database makes it difficult to find a district, as it is not searchable, and the user must click through many pages of districts sorted alphabetically or by ID code. The pictorial icons used to denote the status of a district are not accompanied by obvious explanations. Moreover, the icon granted to a district that is failing in terms of FIRST integrity is very similar to the one conferred by the Texas Comptroller to a district successfully meeting a certain threshold of financial transparency. Unlike many other aspects of TEA-‐published data, many ISDs are keen to tout their FIRST certification, so this confusing standard of districts’ financial health is often the reference point for a user interested in a local district’s status. The FIRST system is a poor gateway for a user into the broader universe of finance data publications, since it does not link to other data tools sufficiently and does not provide a guide for understanding what is behind its ratings.
General Formatting Inconsistencies We have discussed a few examples of TEA data publications within a narrow view. There are substantial formatting inconsistencies in the way TEA presents data to a user. Apparently, TEA has produced website data tools by multiple people, at various times, with different website designs, different underlying database software and different context for different audiences. Hence, it is clear that putting machine-‐readable data in the public realm does not inherently facilitate stakeholder’ ability to use or understand that data. Statutory Authorizations as a Driving Force The confusing network of datasets described above has grown from years of legislation requiring the collection of data related to ISDs’ finances, which is used in funding formulas and other management decisions. Texas mandates have followed a national pattern of demanding agencies create more narrow reports that answer specific questions related to a specific provision of a funding formula or federal report. Lack of Intra-‐Agency Coordination It appears that there is limited communication within the TEA, especially regarding making online, public databases accessible. When initial contacts were made with several divisions, no TEA employee, whether in data generation, data analysis, data reporting, or data support software, identified her job with the task of addressing how the agency informs the public about school districts’ financial health. There were also cases in which employees performing very similar functions on similar data were unaware of each other. A consistent theme among transparency advocates and open government professionals has been an emphasis on the importance of having an internal champion at any agency who wants to develop a stronger base of informed stakeholders. This champion could develop or commission products that other agency staff could use to communicate better. She also could serve as a leader in developing a work environment that encourages employees to consider the public when collecting and publishing data. Data Tools Mapping Project
29
Transparency in Texas: Beyond Raw Data To illustrate the numerous reports and datasets related to school finances in Texas, we created an interactive map. The approach is to categorize web-‐based datasets and lookup tools that allow the public to view school district finance data. We have mapped the datasets based on which office is listed on the website as responsible for the tool. These categories do not always match with the present TEA structure. However, they illustrate the confused picture that outsiders encounter when attempting to find information about their school districts. Our tool serves the first need, and we hope to expand it with an updated map of actual responsible officials to help with the second need. Uncertainty about budget cuts to TEA administrative staff may delay this process.
Recommendations • Designation of a public information manager The Texas Education Agency should assign one individual currently on staff oversight responsibility for how the agency provides information to the public. The agency should seek an internal champion who can exercise leadership in ensuring that public use is considered as data tools are inaugurated and updated. • ISD Widget Development Certain key comparative statistics generated from the TEA site should be made available as a site widget, which is a small, custom piece of HTML code that will display when a user visits a non-‐TEA website that has installed the tool. Hence, some of the information on ISD websites could come directly from TEA, so that citizens could get updated information about their district’s finances without having to rely on local officials to update it. • Demolition of old website architectures
30
Transparency in Texas: Beyond Raw Data
•
•
•
The TEA website has too many sub-‐servers listed in its web addresses (URLs) and too many different frames (borders around a website that contain standard links and items that are meant to appear on all pages of a site). The design inconsistencies make it very difficult to navigate from one area of data to another, or from one area of the agency to another, even when they contain complimentary pieces of data. There should be a single architecture to enhance the user experience. Data.gov-‐style full exploration The TEA and U.S. Department of Education borrow from sites like Data.gov and Recovery.gov where possible. These sites are able to adapt very large databases so they are searchable and can be put into graphic format. Use SAS engine to compare variables Even if the TEA cannot adopt the federal spending-‐track capability to make sophisticated charts in the style of data.gov, the existing SAS engine under State Aid Reports should be able to make comparisons between districts. This could be accomplished through simple measures such as actually listing the names of the districts, rather than just their ID codes. Provide useful guides for PEIMS Data Downloads Publish MS Access Data dictionaries that work for outside users, with step-‐by-‐step instructions.
xxx
"PEIMS Data Standards," Texas Education Agency, http://www.tea.state.tx.us/index4.aspx?id=3014. “PEIMS Data Standards – 2011-‐2012.” Texas Education Agency. http://www.tea.state.tx.us/index4.aspx?id=2147490362. xxxii “Public Education Information Management System PEIMS Data Satndards 2011-‐2012,” Texas Education Agency, http://ritter.tea.state.tx.us/peims/standards/wedspre/index.html. xxxiii Such as www.data.gov or www.recovery.gov xxxiv “School Finance Reports and Data,” Texas Education Agency, http://www.tea.state.tx.us/index2.aspx?id=6741&menu_id=680. xxxv “Forecasting and Fiscal Analysis,” Texas Education Agency, http://ritter.tea.state.tx.us/school.finance/forecasting/downloads/downloads.html. xxxvi “School District State Aid Reports,” Texas Education Agency, http://ritter.tea.state.tx.us/school.finance/funding/sofweb7.html. xxxvii See the SoF for Austin ISD at http://ritter.tea.state.tx.us/cgi/sas/broker?_service=marykay&_program=sfin.sof2011.sas&district=227901. xxxviii “Snapshot: School District Profiles,” Texas Education Agency, http://ritter.tea.state.tx.us/perfreport/snapshot/. xxxix “Data About Texas Public Schools,” Texas Education Agency LONESTAR, http://loving1.tea.state.tx.us/lonestar/Home.aspx. xl “Financial Integrity Rating System of Texas (FIRST),” Texas Education Agency, http://www.tea.state.tx.us/index4.aspx?id=3864. xli “Figure 19 TAC §109.1002(d),” Texas Education Agency, http://ritter.tea.state.tx.us/rules/tac/chapter109/19_0109_1002-‐3.pdf. xxxi
31
Transparency in Texas: Beyond Raw Data
B. LEGISLATIVE BUDGET BOARD WEBSITE The Legislative Budget Board (LBB) is “a permanent joint committee that establishes budgetary recommendations for the Legislature regarding state agencies and estimates the resulting costs in proposed legislation.”xlii The LBB plays an important role in the Texas budgeting process by aggregating financial and performance data for the state, and making that information available to legislators and the public for use in the budgeting process. The LBB was one of the earliest entrants into government transparency through the implementation of its website in the late 1990s. While the LBB provides a wealth of information about the performance of state government, there have been persistent calls for the organization to increase transparency efforts.xliii Last year, an LBJ School research team identified a number of weaknesses in the LBB’s site and a series of general recommendations for improving the site’s functionality. The LBB has expressed interest in revamping its website: however, this is not possible in the current budget and legislative cycle. Our goal for this case study, therefore, was to take the existing information from the LBB website and organize it in a more useful and intuitive format to improve accessibility. Background In assessing the Legislative Budget Board website, we consulted recent publications studying factors influencing the success of e-‐government websites. These studies (Zhang and von Dran, 2002; Tat-‐Kei Ho, 2002; Lupia and Philpot, 2005; Scott, 2006; Lau, 2007) not only analyze how government websites can better serve their users, but also analyze the qualities that make a successful website in general. We have compiled the best practices from this literature as a roadmap for government entities facing problems similar to those faced by the LBB. The 2002 Zhang and von Dran study used a model described by Kano, a Japanese management consultant and researcher, to identify what customers expected from government websites versus other types of domains. Kano identified three levels of customer expectation that a website should meet in order to succeed: (1) basic, (2) performance, and (3), exciting.xliv The “basic” elements are those that the customer takes for granted and cause complaints when not included, the “performance” elements address the customer’s “consciously stated needs,” and the “exciting” elements are “those [unexpected] features that delight customers and inspire loyalty.”xlv In addition to identifying the levels of customer expectation defined by Kano, Zhang and von Dran administered a questionnaire to experienced Internet users (primarily highly educated females) that ranked the quality factors specific to different website domains. These domains were selected based on their different user purposes: “financial (e.g. CNNfn.com, quote.yahoo.com), e-‐commerce (e.g. Amazon.com, e-‐Bay.com, or other e-‐tailer Web sites), entertainment (e.g. a cartoon or a game Web site), education (e.g. National Geographic or a university’s Web site), government (e.g., U.S. Department of Labor, and the White House Web site), and medical or health information Web sites (e.g., WebMD.com and Dr. Koop).”xlvi For all six website domains, the top five ranked quality features included “easy to navigate,” “up-‐to-‐date information,” the existence of a “search tool,” and “accuracy of information.” Other important quality features included “completeness/comprehensiveness of information,” “site technical features,” and “currency/timeliness/update.” These features fit the “basic” and/or “performance” categories of the Kano model.
32
Transparency in Texas: Beyond Raw Data Specifically for the government websites, this study identified the top five must-‐have features as: (1) easy to navigate, (2), clear layout of information, (3) up-‐to-‐date information, (4) search tool, and (5) accuracy of information.xlvii Even though this study is older, we believe the standards it describes are still relevant, especially as the Kano model helps explain the evolution of the LBB website. That is, the Kano model argues that over time, exciting elements become performance expectations and performance expectations become basic. The Legislative Budget Board had the foresight to post its information online in the 1990s, and at that time the website met standard customer performance expectations. Over time, the website’s features and structure have become outdated, as newer e-‐government websites adapted more “exciting” features, such as interactive graphics and user-‐specific portals. The quality standards identified in the Zhang and von Dran study are useful, but subjective. For example, the term “easy to navigate” holds different meaning depending on the user. A website designed for internal use will be easy to navigate for an administrator, but may be harder to navigate for an outside user, and a website with all the information gathered on the home page may be useful to some users, but others may prefer a user-‐specific interface. Alfred Tat-‐Kei Ho addressed this issue in Reinventing Local Governments and the E-‐Government Initiative, in which he argues that e-‐government has transitioned over time from a bureaucratic paradigm, in which agencies designed their websites according to an administrator’s ease of use, to a customer-‐oriented paradigm, in which the website aims to help the citizen interact with government. This shift is in line with President Obama’s 2008 memo on transparency and Director Orszag’s 2009 Directive on Open Government, which stressed not only that agencies should make their information available online, but should also allow for citizen feedback and ease of use.xlviii The best practices Tat Kei-‐Ho identified for agencies making this shift include interdepartmental collaboration, and creating a one-‐stop customer resource. He defines a “one-‐stop customer service” as going beyond offering a lot of content on the homepage (the LBB status quo) to a user-‐oriented portal design, which categorizes information and services according to the needs of different user groups.xlix Texas.gov is a good example of a user-‐oriented portal design, with the information on the homepage categorized into intuitive sections: Do, Discover, Connect and Ask which then have examples of what users have found in each category highlighted below the links. Tat Kei-‐Ho’s study argues that not only will the user-‐oriented portal structure encourage citizen participation, but also will improve the government-‐citizen relationship through efficient provision of information and “customer-‐responsive processes.”l However, the study also notes that this shift cannot occur without an advocate within government who can prioritize making these changes. Edwin Lau, in the chapter “Electronic Government and the Drive for Growth and Equity” in Governance and Information Technology, argues that government websites can boost user satisfaction by collecting “statistics on the number of visitors and page views, frequency with which pages are clicked (or not) and the most common search terms in order to better understand who is using the portal and for what purpose.”li Lau also notes that www.usa.gov ensures user satisfaction by conducting usability and focus group testing to “verify the effectiveness of the information and services to which it is providing access.”lii
33
Transparency in Texas: Beyond Raw Data Lau agrees with Tat Kei-‐Ho that collaboration between agencies can help them identify the most important information requested by users. He also argues that collaboration can help agencies “detect deficiencies in the offering of online services.”liii Lupia and Philpot identified yet more website-‐building best practices, arguing “the viewer’s perception of a site’s effectiveness and efficiency is critical.” (Italics in original). They conducted a survey to collect data on how viewers assess a website’s effectiveness, concluding that “a certain site can increase a viewer’s political interest during a viewing session only if she: ● is aware of the site or visits a site that makes her aware of it; ● views the site; ● perceives the site as providing interesting information effectively and efficiently; ● stays on it long enough to elaborate on the site’s content; and ● the elaboration changes her beliefs about some phenomena, which, in turn, changes her interest in politics.” The LBB’s website does not necessarily aim to increase a viewer’s political interest, but the second and third points are intuitive requirements for a successful website: that the viewer perceives the site as providing information she wants and that the viewer stays on the site long enough to take in what the site offers. Lastly, James K. Scott’s 2006 study “E” the People: Do U.S. Municipal Government Web Sites Support Public Involvement? offers recommendations for successful public involvement on e-‐government websites (not just municipal web sites, as the title suggests). The study makes two salient points: (1) government websites that cater to diverse constituent groups and their needs and that make information-‐gathering and transactions more convenient will “make government agencies more effective and competitive and foster greater citizen interest and involvement in public issues.”liv (2) Making government information easier to find will make government more open and accountable to constituents.lv In sum, the literature suggests the following best practices for successful government websites and websites in general: ● ● ● ● ● ● ● ● ● ● ● ● ●
exceed users’ expectations by providing “performance” and “exciting” elements make the website easy to navigate by using a user-‐specific portal design provide up-‐to-‐date information provide a search tool ensure accuracy of information ensure completeness/comprehensiveness of information provide up-‐to-‐date technical features keep website content current allow for citizen feedback make information-‐gathering and transactions convenient collaborate with other agencies to determine users’ information needs conduct usability testing to ensure the portal design meets users’ needs collect statistics to determine who is using the website and for what purpose
34
Transparency in Texas: Beyond Raw Data LBB Site Structure We examined the LBB’s current website’s structure to identify major weaknesses and how to best address them. Poor Visual Presentation The LBB’s website currently consists of a series of web links organized under the following headings: About the LBB, LBB Staff, Budget Bills and Reports, Performance Reports and Reviews, Agency Reference Documents, Other Publications and Resources, Agency Data Entry, and Agency Instructions. The site is fully text-‐based, with no images except for two logos at the top of each page. Almost all of the links on the website are to PDF documents housed on the LBB’s site, many of them containing important qualitative and quantitative data about the state. Today’s Internet users are accustomed to much more visually appealing sites, and websites such as www.recovery.gov have now become a standard for government website graphic design. Without images to help illustrate the LBB’s work, the large amount of text may be overwhelming to the user and discourage further use of the site. Outdated Design The current LBB website resembles a design style reminiscent of sites created in the late 1990s, and some of the content on the site appears equally outdated. The predominance of text and lack of images makes the site visually unappealing and out of sync with common design styles to which users are accustomed. The lack of drop-‐down menus, tabs, or other means of organizing content force the user to adjust to how the website functions, rather than being able to intuitively understand how to navigate. The presence of content that appears to be remnants of the original launch of the LBB’s website leads the user to believe proper site maintenance is not being undertaken, making the user less likely to visit the site frequently. Links to download Adobe Reader 9 and Crystal Reports appear to be randomly placed, and the login pages for state agencies and staff use appear to be based on early Internet designs. Confusing Categorizations Understanding which documents are classified under each of these headings is not an intuitive process, and there is significant overlap between some categories, contributing to user difficulty. For example, the Comptroller’s request for agencies to submit revised budgets reduced by five percent (known as the “Five Percent Reduction”) includes documents listed under four separate headings. This makes it difficult for even legislative staff to easily find the specific documents they need to participate in the process. The “Links” category is the most confusing, with a wide spectrum of information ranging from approved reduced budgets to links to staff email access or to download Adobe Reader 9. Furthermore, the LBB does not include descriptions of the documents included on the website, so users must dig through documents to find their specific information. Our Vision for the LBB To envision how the LBB could become more transparent solely by better utilizing the resources already available on its website, we decided to develop a re-‐design of the LBB’s website. The re-‐design better organizes the existing resources and presents them in a manner meaningful to users following the recommendations for presentation discussed in this report. Our goal is to reinforce our main point, that a site is not transparent unless it is easily accessible and used in a meaningful way.
35
Transparency in Texas: Beyond Raw Data A New Site Design The first major change we recommend for the LBB’s site is to modernize the “look and feel” so that users will feel comfortable navigating it as easily as any other website. As noted in the recommendations section, the price people pay for using a website is time and most are unwilling to spend much of it trying to decipher a confusing website. With a freely available website template, we were able to create a more colorful design for the site that is in sync with the standard set by other government sites. The re-‐design did not require any extraordinary level of web design expertise and we added a minimal number of images to the site, to reflect our belief that improvements may be made to the LBB’s site simply by re-‐branding and organizing the information already available. Document Classification System We recommend website designers should divide content into broad categories using a main menu. An important way we addressed the lack of organization on the LBB’s website was through development of a system of classifications to make the categorization of documents more logical. In the first step, we identified broad common themes within existing information on the LBB’s current website. We then included these classifications as tabs at the top of our version of the site – About LBB, Budget, State Government, Federal Funds, How To, and Reporting. Under each of these main categories, we created a series of sub-‐categories that help to further delineate the differences among documents available on the site. Through using a document hierarchy that better reflects what is available on the website, our goal is to make LBB’s work more transparent by making data accessible. “How-‐To” Guides for Site Use Because of the complexity of some of the LBB’s database resources (particularly the budget reporting and contract searches), we developed a series of guides to assist users in making the best possible use of those resources. These guides aid Legislative members, agencies, media, or even members of the public in accessing information though step-‐by-‐step instructions. For example, certain databases contain search terms or options whose meanings are not readily apparent from their labels, so we developed a guide to clarify those terms so people better understand how the database functions. The goal of this component is remove barriers and increase usage of the LBB’s resources. Descriptions of Document Contents Additionally, we created a series of descriptions of document contents to improve organization of the site. Some documents, such as the Texas Factbook, contain extremely valuable information that is not apparent in the title of the document. We took a look at the individual documents, wrote a one-‐ sentence summary, and included that with each document. Our goal was to make data more accessible by making it easier for users to understand what they can extract from each of the documents on the website.
xlii
“Tribpedia: Legislative Budget Board,” The Texas Tribune, accessed April 27, 2011, http://www.texastribune.org/texas-‐taxes/legislative-‐budget-‐board/about/.
36
Transparency in Texas: Beyond Raw Data xliii
82R 4580 KJM-‐D: Relating to the maintenance by the Legislative Budget Board of a searchable database containing certain budget information. , S. SB 700, 82d Leg., 82(R) (Tex. ), accessed February 23, 2011, http://www.capitol.state.tx.us/BillLookup/History.aspx?LegSess=82R&Bill=SB700. xliv Ping Zhang and Gisela M. von Dran, "User Expectations and Rankings of Quality Factors in Different Web Site Domains," International Journal of Electronic Commerce 6, no. 2 (Winter 2001/2002): 9-‐33. http://www.jstor.org/stable/27751011. xlv Ibid., p. 12. xlvi Ping Zhang and Gisela M. von Dran,"User Expectations and Rankings of Quality Factors in Different Web Site Domains," International Journal of Electronic Commerce 6, no. 2 (Winter 2001/2002): 27. http://www.jstor.org/stable/27751011. xlvii Ibid., p. 25. xlviii Peter R. Orszag, Open Government Directive, http://www.whitehouse.gov/open/documents/open-‐ government-‐directive. xlix Alfred Tat Kei Ho, "Reinventing Local Governments and the E-‐Government Initiative," Public Administration Review 62, no. 4 (July-‐August 2002): 437. http://jstor.org/stable/3110358. l Alfred Tat Kei Ho, "Reinventing Local Governments and the E-‐Government Initiative," Public Administration Review 62, no. 4 (July-‐August 2002): 436. http://jstor.org/stable/3110358. li Edwin Lau, Governance and Information Technology: From Electronic Government to Information Government, ed. Viktor Mayer-‐Schonberger and David Lazer (Cambridge: MIT Press, 2007), 46. lii Ibid. liii Edwin Lau, Governance and Information Technology: From Electronic Government to Information Government, ed. Viktor Mayer-‐Schonberger and David Lazer (Cambridge: MIT Press, 2007) 36. liv James K. Scott, "'E' the People: Do U.S. Municipal Government Web Sites Support Public Involvement?" Public Administration Review (May-‐June 2006): 345. lv Ibid,, p. 349.
37
Transparency in Texas: Beyond Raw Data
C. CITY OF KYLE Members of the 2010 State Finances and Online Transparency Policy Research Project conducted research with the City of Kyle to gain perspective on the challenges facing local governments regarding improving financial transparency online. City Officials in Kyle identified issues such as technical capabilities, including lack of standardized databases and software, privacy concerns about certain types of information and uncertainty about how the public might interpret data without context.lvi Despite these challenges, Kyle has made some important achievements in online transparency and City officials continue to move forward. Now, City officials are seeking input from the public for a website redesign aimed at making City information more useful and accessible. In the last year, Kyle officials remained committed to attaining the Gold Circle Level of the Texas Comptroller’s Leadership Circle, which recognizes local governments for their efforts in online financial transparency. Gold reflects the highest level of transparency and goes to local governments that publish the budget, audit and check register online. The award criteria also include ease of access, contact information for local officials and public information requests, among others.lvii With guidance from the prior research project, Kyle received the Gold Circle recognition in December 2010. Because of Kyle’s ongoing commitment to improving online transparency, we decided to continue the dialogue begun last year. We opted to approach Kyle focusing on the civic engagement perspective. With Kyle’s recent award, we saw an opportunity to see how the public has reacted, what changes the city has seen in engagement and what plans Kyle officials have for further improving transparency and engaging the public. With City official’s willingness to seek public input, we also saw an opportunity to study Kyle’s efforts to encourage civic engagement.
The Road to Gold We met with an official in the Finance Department in Kyle who played an integral role in helping the City achieve Gold. She indicated that City leaders are looking for new ways to make financial information available online beyond the criteria of the Leadership Circle. She participated in the Texas Government 2.0 conference that we held in January where she gained insight on displaying financial data in user-‐ friendly formats. Kyle’s final step to reach the Gold Circle level was publication of the City’s check register online. In that process, Kyle officials encountered many of the anticipated barriers. The most difficult part of that process was getting the software to report the data needed for publication without investing additional financial resources. City officials worked with the vendor and were able to create a customized report within the existing software package. A challenge in customizing the report was how to display check register data while maintaining an appropriate level of individual privacy. The City’s website was also a limitation, requiring the department to “make due with what you have.”lviii Since receiving the award, Kyle officials have not received much feedback from the public. Some were expecting potentially negative feedback for publishing the check register, but that has not occurred. Kyle also has not had much positive feedback. Local news media covered the award, and the City publicized it in print and electronic newsletters. To continue efforts toward improved transparency, City officials feel they need public input.
38
Transparency in Texas: Beyond Raw Data Next Steps Though Kyle officials would like to see some of the recommendations from the previous year implemented, such as a template or opensource software that cities can use to publish financial data in a standard format, they are moving forward with a website redesign. The City’s goal for the redesign is to improve the ability of all departments within the City to make information publicly available. Members of the Finance Department know that they need input from the citizens of Kyle so they can make the best use of the new website. Officials would like to hear from a broad sample of residents, not just the few who are highly involved in City government. The new web platform will make it easier for individuals in each department to update information and make changes to the content and display. The Finance Department produces charts, graphs and summary information for City Council members and the official we met with felt the web redesign would improve their ability to publish that kind of information to the public, as well. The site will include feedback mechanisms for the public to communicate with City officials and input on the site. Engaging the public in this way is a critical step in moving beyond the e-‐government model of providing forms and documents online to an e-‐governance model that encourages residents to interact with City leaders and participate in guiding the City’s direction.lix When government agencies, such as the City of Kyle, take action to improve transparency, they must attempt to make those efforts known if they want to move beyond a one-‐way information portal. Previous visitors, who did not find the information they needed, may not return because of past experience and new visitors may not be aware of efforts to improve transparency. Dialogue with the public is necessary for governments to gauge the success of their efforts, to understand the constituencies they serve and to come together to improve government and the community. The first step involves publicizing transparency efforts and then seeking public input. Working together with the community, governments at all level can begin to practice e-‐governance. We assisted Kyle officials in developing and distributing a questionnaire aimed not only at those well versed in city government, but also casual users of the City’s website. With consultation from several City departments, we developed a brief electronic questionnaire to gauge familiarity with the City’s website, reasons for use and to solicit feedback on ways to improve the site. Preliminary results were not available by the final publication of this report. Input from this questionnaire will provide guidance to City officials in fine-‐tuning the current redesign and create ideas for future improvements, helping the City of Kyle exceed current local government transparency standards and truly engage city residents.
lvi
LBJ School of Public Affairs -‐ Policy Research Project, Texas Financial Transparency: Open and Online, ed. Matt Hartman, Claudia Montelongo, and Jennifer Quereau, May 12, 2010, http://www.utexas.edu/lbj/cpg/docs/prp.pdf. lvii "Texas Comptroller Leadership Circle," Texas Transparency: Window on State Government, accessed April 27, 2011, http://www.texastransparency.org/local/leadership.php. lviii Personal Interview, March 3, 2011 lix Ibid
39
Transparency in Texas: Beyond Raw Data
D. CITY OF MANOR: Transparency on a Budget
The Internet era has improved the public’s access to information and raised expectations regarding government openness. Now, governments at all levels can make large amounts of information available to the public online with ease. A concern of many local governments is that they simply do not have the budget to change existing systems. However, the expense of implementing more transparent methods of government can be minimal. Consider the case study of Manor, Texas. Manor is a suburb of Austin with a population of about 6,500. The City has had its share of financial constraints, particularly with the recession. Dustin Haisler, CIO for Manor until November 2010, needed to be resourceful and stretch his $100,000 budget thin. In 2008, he began allocating this budget towards small efforts that would improve transparency, including accounts with social networking sites and records retention for these accounts.lx This was followed by rebuilding the website to include collaborative components, using the free and simple Wordpress publishing platform. Eventually, Haisler started a crowdsourcing suggestion website called Manor Labs and installed QR codes around the city. Manor Labs is a website where residents make suggestions and then vote on the best ideas. Ideas earn point, called Innobucks, which are redeemable for prizes. From Manor Labs came the idea for SeeClickFix, an app that enables residents to report maintenance problems around the city, costing Manor only $1200 a year.lxi Manor Labs was built by Spigit, a social networking software company in California. Spigit waived the typically $5000 a year fee for the project, as Manor was one of its first customers. Manor has similar arrangements with over a dozen other technology companies, and has found other deals through testing products and using open source software.lxii More than 2,000 residents of Manor have visited Manor Labs, which is much higher than their usual city council meeting attendance.lxiii Quick Response (QR) codes are two-‐dimensional bar codes made up of printed squares in various sizes, as large as four by four.lxiv These squares hold more information than typical barcodes.lxv Residents can easily download free software onto their mobile phone, take a photo of the barcode with their camera, and be directed to a website with more information about the place or item. The City uses codes to label construction projects, City buildings, and City vehicles, which they can scan to get real-‐time work updates.lxvi Manor used an online, downloadable code generator to encode specific URLs on to each QR-‐code. There have been some improvements over the three years, like limiting the codes to 80 characters to better enable camera phones to read them. This can be done with URL compression technologies like Tiny URL.lxvii The codes are reusable so beyond the initial investment, the only cost to Manor is the paper and printing for the QR codes, and the minimal time to update and change the websites affiliated with each QR code.lxviii (See Appendix E) These tools have saved the City time, money and resources, while tremendously boosting civic engagement. The QR program has reduced the number of calls about public projects, and 70 percent of the population has signed up for the online bill payment resources implemented in 2008.lxix Although Haisler left the City of Manor after accepting a position at Spigit, the City continues to develop these programs through public education and working with local cell phone merchants to install the
40
Transparency in Texas: Beyond Raw Data software in phones at purchase.lxx To prove that this case study is not an anomaly, Manor is also currently working with the small, rural City of De Leon (population 2,500) to implement the same technologies.lxxi Manor received the Most Innovative Use of Technology award in Texas in 2008 from the Center for Digital Government.lxxii
lx
Max Chafkin, "Why the High-‐Tech Industry Loves Manor, Texas," Inc. , August 24, 2010, accessed March 1, 2011, http://www.inc.com/magazine/20100901/why-‐the-‐high-‐tech-‐industry-‐loves-‐manor-‐texas.html. lxi Ibid. lxii Michael Jeffers, “Cheap ‘QR’ Codes Are a Budget-‐Friendly Project for Manor, Texas,” Government Technology’s Digital Communities, February 4, 2009, accessed March 1, 2011, http://www.digitalcommunities.com/articles/Cheap-‐QR-‐Codes-‐Are-‐a-‐Budget-‐Friendly.html.. lxiii Ibid. lxiv Susan Gonzales, “23-‐year-‐old techie puts Manor on map,” Austin American-‐Statesman, January 3, 2010, accessed March 1, 2011, http://www.statesman.com/news/texas/23-‐year-‐old-‐techie-‐puts-‐manor-‐on-‐map-‐ 160878.html. lxv Michael Jeffers, “Cheap ‘QR’ Codes Are a Budget-‐Friendly Project for Manor, Texas,” Government Technology’s Digital Communities, February 4, 2009, accessed March 1, 2011, http://www.digitalcommunities.com/articles/Cheap-‐QR-‐Codes-‐Are-‐a-‐Budget-‐Friendly.html. lxvi Dustin Haisler,“White Paper: Redefining Government Communication with QR-‐Codes,” City of Manor, September 2009, accessed March 1, 2011, http://www.cityofmanor.org/comwhitepaper.pdf. lxvii Ibid. lxviii Max Chafkin, "Why the High-‐Tech Industry Loves Manor, Texas," Inc., August 24, 2010, accessed March 1, 2011, http://www.inc.com/magazine/20100901/why-‐the-‐high-‐tech-‐industry-‐loves-‐manor-‐texas.html. lxix Nick Judd, “In Texas, a Small Town Hopes for a Gov 2.0 Makeover Miracle,” techPresident, September 1, 2010, accessed March 1, 2011, http://techpresident.com/blog-‐entry/texas-‐small-‐town-‐hopes-‐gov-‐20-‐makeover-‐miracle. lxx Susan Gonzales, “23-‐year-‐old techie puts Manor on map,” Austin American-‐Statesman, January 3, 2010, accessed March 1, 2011, http://www.statesman.com/news/texas/23-‐year-‐old-‐techie-‐puts-‐manor-‐on-‐map-‐ 160878.html. lxxi Nick Judd, “In Texas, a Small Town Hopes for a Gov 2.0 Makeover Miracle,” techPresident, September 1, 2010, accessed March 1, 2011, http://techpresident.com/blog-‐entry/texas-‐small-‐town-‐hopes-‐gov-‐20-‐makeover-‐miracle. lxxii Michael Jeffers, “Cheap ‘QR’ Codes Are a Budget-‐Friendly Project for Manor, Texas,” Government Technology’s Digital Communities, February 4, 2009, accessed March 1, 2011, http://www.digitalcommunities.com/articles/Cheap-‐QR-‐Codes-‐Are-‐a-‐Budget-‐Friendly.html.
41
Transparency in Texas: Beyond Raw Data
E. WEB TOOLS Web applications can assist interested parties in improving the interaction between citizens and government. The following is a list of privately developed tools that can be employed to improve transparency efforts or used as examples of best practices of resources that promote open government. • Socrata (http://www.socrata.com): Socrata helps organizations save time and money by streamlining the data publishing process and automating maintenance and updates. Negating the need for additional trained IT staff, this tool allows an organization to publish data simply and easily to comply with transparency mandates. Socrata also manages and measures open data projects to organize the sheer volume of information and increase constituents’ ability to consume and utilize data. Socrata allows users to interact with data in a variety of ways (visually, through API’s, over the phone) as well as distribute data online. Users can create their own charts and share them with others. Data sets are searchable, sort-‐able, and tag-‐able. The filter function allows users to sift through data and limit observations to specific information needed to make valuable comparisons. Data sets are exportable to PDF and excel formats. Socrata promotes civic engagement by allowing users to subscribe to data set so that they are notified of changes. Users can also post comments on specific data points. This could allow citizens and government officials to comment on data of interest. • Spigit (http://www.spigit.com): Spigit utilizes open innovation technology that enables governments to engage citizens, elicit feedback and discussion, and facilitate collaboration. Citizens can communicate ideas with government officials through the use of an idea forum. Ideas can be posted and scored, with the most popular ideas featured on a leader board. Crowdsourcing technology allows governments to not only examine an idea, but also answer questions and determine who or what would be affected by the idea. Citizens can watch their idea advance through levels of government so that they can engage in the governing process. This platform promotes civic engagement with limited resources using idea management and innovation discovery. • Firmstep (http://www.firmstep.com/): Firmstep provides applications that connect citizens with the information and government services they need. The AchieveService application allows citizens to identify available government services, check their eligibility, and receive updates. The AchieveForms application allows citizens to easily apply on the web for government programs. These eforms are user-‐friendly and cost effective and reduce administrative burdens for government and red tape for constituents. Both private corporations such as Siemens and small local government like Corby Borough (UK) have utilized Firmstep’s application to connect with constituents. • Quick Response codes: QR codes are two-‐dimensional matrix barcodes that can be read by camera phones. Once these square bar codes are scanned by an Internet-‐capable phone, they can connect to a URL on the phone’s browser with more information about the object or site that is attached to the code. These URLS can be compacted online and stored in the barcodes. URL’s can be compacted on a number of free online websites, including http://www.tinyurl.com. The phones that scan them must have a reader application downloaded. A list of readers and compatible phones can be found at www.qrme.co.uk/qr-‐code-‐resources/qr-‐code-‐readers.html. Free Mobile QR-‐code applications can
42
Transparency in Texas: Beyond Raw Data be downloaded at the following websites: • Google ZXing(Android) www.zxing.org/w • 3GVision's I-‐Nigma (Variety of Devices) . http://www.inigma.mobi • Block5 (iPhone) http://www.block5.com/iphone • Listed As "QR App" In The iPhone App Store
For more information on QR codes, visit http://www.qrme.co.uk or http://2dcode.co.uk/. • Open Plans is a nonprofit organization that provides technology tools for open government. They provide open source technologies to local governments, including OpenGeo, a user-‐ friendly mapping tool for community members, and Civic Commons, which allows cities and states to adapt and share code effectively. To find out more, visit www.openplans.org.
43
Transparency in Texas: Beyond Raw Data
F. LEGISLATIVE RESOURCES A key element of our democracy is the formation of new laws through a representative legislative system. For this system to work, democratically elected representatives must act on the interests of the public and be held accountable by the people they represent. This requires more than just a vague interest on the part of the public, but a strong motivation to follow what is happening and voice opinions. Historically, public involvement in the legislative process has been difficult and relied heavily on media coverage. If individuals sought more information they had to personally attend hearings and committee meetings. The availability of information online has dramatically increased the ease and convenience of accessing legislative updates. Available information, though, must be coupled with the means to act on the information to create a truly engaged polis. During the research portion of this project the 82nd Legislative Session convened in Texas, providing us with the opportunity to examine the existing mechanisms Texans can use to access legislative information and engage with legislators. We found that the state government website, Texas Legislature Online, provides a large amount of information and tools to track legislation. Users can search by bill number or key words and view the bill text in several formats, as well as the text of amendments and fiscal notes. Users can also see where the bill is in the legislative process, set up custom tracking, and find the contact information for all legislators. What is lacking on the state site is the ability to voice an opinion in the public sphere and initiate a dialogue. Several third party websites try to fill the gap by supplying a forum for online public discourse of the legislation being debated at the Capitol. These sites (listed below) often connect users to the original text on the state site, but on their own site have links to social media networks and news articles and allow users to express support or opposition for specific bills. These features are vital to progress from simply the online provision of bill text to the active participation of the public. Below we have complied a list of legislative resources individuals can use to follow—and participate in— the 82nd Legislative Session in Texas. Online resources and descriptions Texas Legislature Online http://www.capitol.state.tx.us/Home.aspx This is the official site of the State Legislature and allows users to search by bill number, subject, or text within the current and past legislative sessions. Searching for a bill will lead to a page that shows the bill’s history, versions of the text in PDF, Word and HTML, fiscal notes, amendments, and actions. It also provides a diagram of bill stages and where the particular bill if interest is in the legislative process. The My TLO feature allows users to create a custom bill list to track legislation. Alerts can be sent directly to the user when changes or movement occurs on bills on the list, or if bills are assigned specific subjects. Bill Hop
44
Transparency in Texas: Beyond Raw Data http://tx.billhop.com/ Bill Hop allows users to search for bills filed in the 82nd Legislative Session and displays all versions of the text. Users can register and create a username to leave comments, track pages, and rank pages. The home page displays bills that have been ranked favorable, unfavorable, conservative, or liberal by other users. Texas Tribune http://www.texastribune.org/session/82R/bills/ The Tribune provides interactive treemaps that show bills filed in the 82nd Legislative Session by subject and legislator, distinguishing between the House and Senate. There is a how-‐to video to help users navigate the maps. Clicking on a box (either subject or legislator) provides a list of legislation that can be further filtered. Each bill then links to a page that provides information on actions taking on the bill and quick links to the history and text on the Texas Legislature Online website. Open Government http://tx.opengovernment.org/ Open Government displays all bills filed in the 82nd Session and can be filtered by chamber, recent actions, date introduced, most viewed and most in the news. It is also possible to search for a bill by key word. Each bill page shows its status, sponsors, recent actions and links to the official text. It also provides links to social media mentions and related news and blog coverage. Users can make comments and quickly share the page using social media links and a quick link to email your representative (input zip code). BackMic.com http://backmic.com/ BackMic users can browse bills by subject or author or view the BackMic scorecard that shows all bills. Users who create and account can vote thumbs up or down on legislation and make comments. Each bill’s page displays comments from the bill author, BackMic contributor, and the mood on the bill based on votes. Users are directed to Texas Legislature Online to read text and track action on the bill. There is a how-‐to video to help users. MyGov365 http://www.mygov365.com/state/texas MyGov365 houses legislative information at both the federal and state level. Users can browse bills and view text, status, sponsors and comments. Users who register on the site are able to track specific bills, make comments and vote yea or nay.
45
Transparency in Texas: Beyond Raw Data
G. QUESTIONNAIRE
46
Transparency in Texas: Beyond Raw Data
47
Transparency in Texas: Beyond Raw Data
48
Transparency in Texas: Beyond Raw Data
49
Transparency in Texas: Beyond Raw Data
50
Transparency in Texas: Beyond Raw Data
51
Transparency in Texas: Beyond Raw Data
H. GLOSSARY
Civic Engagement -‐ Active citizenship in which people are easily able to have direct input into governmental processes and policy formation to identify and address issues of public concern. Crowdsourcing -‐The act of outsourcing tasks on the web to an undefined, large group of people or community to harness the power of collective knowledge to achieve the given task. Data Production -‐ The process of gathering, compiling, and presenting public information. Digital Divide -‐ The knowledge and access gap between people to digital and information technology. It can refer to both a disparity in resources and skills needed to effectively utilize web based technologies and the physical access to technology. Drop-‐Down Menu -‐ User interface widget that enables users to choose an item from a list when activated. Exportable -‐ When data on a web page is able to be sent to a program that allows the user to view the data in an organized way and to edit for their own purposes Intermediary Organizations -‐ This refers to non-‐governmental organizations that actively engage citizens including advocacy organizations, watchdog groups, the media, or interest groups. Machine Readable -‐ A data format that can be fed into and read by any computer or mobile device. Raw Data -‐ Source data that has not been processed or manipulated. Reporting Software -‐ Generates human-‐readable reports from data sources. Social Media -‐ Web-‐based media used for social interaction. Transparency -‐ The ease of accessibility, exportability, and usability of accurate online data. Widget -‐ An application that can be installed in a web page by a user.
52