Shrinking city budgets have created more competition for limited dollars, even among worthy projects. At the same time, councils and citizens are demanding more transparency in how their money is being spent, and with what result. Understanding how to evaluate and use data to show results is becoming an increasingly critical skill set for local managers.

Dr. Maureen Berner, associate professor of the MPA program at the UNC School of Government, and author of Statistics for Public Administration: Practical Uses for Better Decision Making, shared strategies for utilizing data to guide decision-making, build support, and demonstrate ROI, during a May 20 webconference, Statistics for Everyone: How to Translate the Facts to Make Better Decisions in Local Government.

Find out what your audience wants, then give it to them

Step one, Dr. Berner advised, is understand who your decision-makers are and what information they value.  For example, if your decision-making body is your city council, you may need to determine if there is one member of the group who is more influential than the others.  In that case, your strategy may be to get that person on board with you by tailoring your presentation to him.  If your decision-maker is an internal manager, you may find that person is already familiar with the need your project is filling, and you will need to focus your data on why you have chosen that particular method or product.

Once you have identified your project gatekeepers, the next step is to define what data they need, and in what form they want it. If, for example, you are championing an after-school program designed to prevent youth from getting into trouble in the evening hours, you may include data that shows a decrease in crime rates among communities with similar programs. That data may be demonstrated in the form of facts, statistics, or personal stories.  A council member voting on an after-school program for example, may want to hear testimonials from children who have benefitted from the program, while another may prefer to see comparative crime statistics. It is imperative to know your audience and their preferences.

One pitfall to avoid, Berner cautions, is to focus too much on cost. Often times, she explained, a manager may be tempted to simply calculate the cost of a project and then make the case that the government can afford it. This leaves out the answer to the central question: Is this project a good investment? While cost will eventually become a deciding factor, by making it central to your argument, you will fail to demonstrate why the program is important at all.

Understanding Surveys:  How much information do you need?

When it comes to surveys, more information is better, Berner said. Unless it is not. It is always best to survey your full population for the most accurate information. However, that is frequently impossible or impractical, depending on the size of the population and the difficulty in collecting data. In those cases, it is best to survey only a sample size of your population and extrapolate from there.  

One you have determined what a representative group of your survey population should look like, Berner suggests you aim for a survey of 350 people. This number, she said, works well for populations of any size, as long as it includes a reasonable representation of the total population. This number may be adjusted depending on how precise you need to be with your statistic, how diverse your population is, and how confident you need to be in your results.

To guide this decision, Berner suggests you apply what she calls the “golden ratio” of surveys: What is the importance/work ratio you need to achieve optimal results? A project for which the outcome has largely been determined, or for which the overall impact is small may not require a large and expensive survey. For something that is of critical importance to the community, it may be worth doing a full survey to generate better results.

Quick Tips for Success

Berner concluded the presentation by offering these quick tips for maximizing your survey data:

• When possible, survey your entire population.
• If you do use a sample population, a small sample works just as well as a larger one, provided the sample is representative of the total population.
• Digital surveys, via email or online, are better than phone or mail surveys for receiving feedback and calculating the results.
• Avoid “nice to know” extraneous questions on your surveys. Too many questions will negatively affect response rates.
• Think about ways in which you are already gathering information before you use a survey.
• Consider focus groups, especially if you need in-depth information.
• Involve your staff. They may have insights as to what information is helpful or realistic to gather.
• MPA students are a great resource for gathering data. They are often willing to provide their labor in return for a chance to learn more about local government.
• Manage expectations. If your council insists on a survey, make sure they understand the limitations, and know that response rates are historically low.