Show Us the Stimulus

Good Jobs First

Every state government now has a website reporting on its role in implementing the American Recovery and Reinvestment Act (ARRA), the $787 billion federal stimulus bill that is shoring up state finances, reinforcing the social safety net and funding job‐creating infrastructure and energy‐efficiency projects. Yet there are wide differences in the quality and quantity of the information the websites offer. Some state ARRA sites reinforce President Obama’s promise that the stimulus plan would be carried out with “an unprecedented level of transparency and accountability.” Others are half‐hearted efforts that provide residents little useful data on the largest federal stimulus since the New Deal, making it harder to measure ARRA’s success.

These are the conclusions of Good Jobs First after a review and evaluation of the states’ ARRA websites. We considered how effective the sites are in conveying information about:

• the categories of stimulus spending;
• the distribution of that spending in different parts of the state; and
• specific projects being carried out by private contractors, including their employment impact.

The study rates each state twice on a scale of 0 to 100: once for its main ARRA website and again for a high‐profile aspect of stimulus spending—highway projects (whether that reporting is done through the main ARRA site or a separate transportation department web page). In each case we evaluate the sites according to about ten factors relating to the quality and quantity of the information presented on the many forms of ARRA spending taking place in every state.

Most states do not score very high. The average of all the state scores for ARRA sites is only 28.1, and the median is 25. Scores for ARRA highway reporting are somewhat better—an average of 37.6 and a median of 35—but still not very impressive.

Only six states score 50 or better for their ARRA website: Maryland (80), Colorado (68), Washington (63), West Virginia (60), New York (53) and Pennsylvania (50). Thirteen do so for their highway reporting, led by Maryland (75), Washington (73), Nebraska (60) and California and New York (each at 58). Only four states (Colorado, Maryland, New York and Washington) score 50 or better on both measures.

In contrast to these, two states score zero for their websites (Illinois and Utah), and two get a goose egg for their highway reporting (Illinois and Kentucky), so that Illinois is the only state to strike out completely in both ratings. Most states that score poorly for their main ARRA website do better with their highway reporting, but apart from Illinois, there are four states that score very low for both: Alabama, District of Columbia, Kentucky and Vermont. Low-scoring states are ones that provide few specifics on how ARRA money is being used in the state. Illinois, for instance, has figures only at a national level and nothing on how much will be spent in the state.

Although this study focuses on state ARRA websites, we also include an evaluation of New York City’s impressive NYCStat Stimulus Tracker, which is comparable in quality to the top state ARRA sites.

Here are highlights of the state scores for specific criteria:

• Most states do a good job of providing information on the categories of ARRA spending. Forty-two states display the data for broad categories (energy, housing, transportation, etc.), and 37 of these also provide details on specific programs. States such as Iowa, West Virginia and Wisconsin use animated bar graphs to highlight the information. On the other hand, some states such as Connecticut and Idaho bury the information in secondary web pages or linked PDF documents.
• Geographic breakdowns are less common than data on program areas. Eighteen states provide the information, and in only three cases (Maine, New Mexico and Virginia) does the website show the information both for each county individually and for all counties side-by-side for comparison purposes. States such as Oregon, Tennessee and Washington use interactive maps to display county breakdowns.
• Few states juxtapose the geographic distribution of stimulus spending with patterns of economic distress, such as county unemployment rates or foreclosure levels. Only Maryland provides this information on its main ARRA website. Five states— California, Florida, Missouri, South Dakota and Washington—do so in their highway project reporting.
• Apart from county dollar totals, state residents may be interested to know where individual ARRA projects such as the repaving of a road or repair of a school building are taking place. Eleven states (California, Colorado, Hawaii, Maine, Maryland, Massachusetts, New York, North Carolina, Pennsylvania, Rhode Island, South Carolina) provide project maps on their main ARRA website, while 30 provide maps as part of their ARRA highway project reporting. Some of these are static PDF files, while states such as California, Colorado, Maryland, Massachusetts, New York and Rhode Island have interactive maps.
• Twenty-one states provide project or contract detail on their main ARRA site, while nearly all (48) do so in their highway reporting—though in some cases the descriptions are minimal. Some states provide project details through interactive maps with pop-up information boxes, though it is not always the case that all types of projects are included.
• Only 10 states (Colorado, Indiana, Maryland, Massachusetts, New Hampshire, New York, Pennsylvania, Rhode Island, Washington and West Virginia) provide contractor names and dollar amounts on their ARRA website. The results are better in highway reporting, where 29 states have both contractor names and dollar figures. Massachusetts is the only state that has indicated it will provide online access to the full text of ARRA contracts, while Delaware and Maine are beginning to do so for transportation projects.
• The paramount objective of the Recovery Act is to address mounting unemployment through job creation and retention. Yet only three states—Colorado, Maryland and Washington—currently provide any employment data for individual projects on their main ARRA site. Eighteen states do so in their highway reporting.
• Despite the importance of knowing how long a project will last and how much of ithas already been completed, only 7 states provide project status information on their ARRA websites; 19 do so in their highway reporting.

Based on our findings, we offer the following recommendations:

• Put a summary of key information about ARRA spending at the top of the home page of the site. A clear bar graph, pie chart or table showing the main spending flows goes a long way in helping the user begin to see what the Recovery Act is all about. There should be clear links to pages with details about the various specific programs. Avoid putting the details in large PDF files that may be difficult to open and tedious to read.
• Provide a map or a table showing how overall ARRA spending and the amounts in key categories are being distributed around the state.
• Along with information on spending streams, provide information on individual projects being funded by those programs, such as a particular transit improvement or weatherization effort. Where possible, display the location of the projects on maps. Interactive displays that allow one to drill down for details are better than static PDF maps.
• For projects carried out by private contractors, be open about the contract award process and the identity of the companies that win bidding competitions. Post the bids and the details, including the full text, of the contract awarded to the winner. Include data on contractor parent companies and use standard identifiers.
• While the federal government’s Council of Economic Advisers is responsible for estimating the overall employment impacts of ARRA and the Recovery.gov website will report jobs data on some (but not all) individual projects, state ARRA sites should also include employment data in their project reporting.
• State ARRA sites should make it clear when individual projects began work (or are expected to start), how long they are expected to run and what percentage was complete as of the latest reporting date. It would also be helpful to indicate which projects are using both ARRA and state funds (and how much of each).
• States should make their ARRA websites as user‐friendly as possible. This means keeping all or most of the program and project information on the main site, rather than requiring users to engage in an online scavenger hunt across agency web pages. The data should also be updated as often as possible and should be available in machine-readable feeds. Sites should have standard web features such as a search engine and a site map.
• ARRA sites should provide readily accessible information about the ways that individuals, organizations and businesses can apply for stimulus grants and contracts. This should include contact names (with phone numbers and e‐mail addresses) and selection criteria.

Most of these recommendations are simply matters of good web design and best practices in government transparency. Yet the use of ARRA websites to inform the public is more than a matter of providing a service to state residents. The way in which the information is presented shapes public attitudes toward the stimulus and could play a significant role in debates over future government interventions in the economy.