When someone asks about using statistics in grant appeals, I recall summer course I took at the U of W between my junior and senior years of high school. The course brought together high school debaters from around the state and gave us the chance to learn more about crafting arguments.

During a critique of a session one instructor said to me "You use statistics like a drunk uses a streetlight, for support rather than illumination."

It took me a few years to really understand what the instructor was trying to tell me. And it is probably the best lesson I ever learned about developing fund raising materials and, in my present role, reviewing them.

Here are two examples. After a suitable introductory paragraph, one proposal letter starts its appeal by stating:

"The home ownership rate in our community is only 60%, which is below our state's average of 64.6%. Therefore it is vital that we start a home ownership center to reverse this trend."

After a similarly appropriate introduction, the second proposal letter starts its appeal with:

"Many families in our community pay more in rent each month than they would have to pay for a single-family home. That higher cost of housing leads to family instability because any little crisis can upset the family budget. One major reason families pay this higher rental rate is that our community has only a few bank branches located in the community and no major home lenders. As a result programs for low-income home buyers are difficult for our families to access. These families often need to take home ownership classes and debt counseling before they can qualify for these good programs. The impact of these factors can be seen in the fact that our community's home ownership rate is only 60%, which is lower than the state average of 64.6%"

While these examples are very bare bones, they do reflect two approaches to grant appeals that I see all the time. Both examples use the same statistic. But which one is the more compelling? While the second example is nearly 100 words longer, those words carry so much value in this context.

First, we know one of the major reasons this organization believes home ownership is below the national average in its community. That tells me that if you were to get a grant, this cause of lower home ownership is the one you would spend most of your time on. More important, you share details. It isn't just that people can't access services from banks or major home lenders, you've told me what services.

Second, the second appeal also tells me why the current situation is bad. I can accept the notion that every dollar is important for a low-income family. And you've added to that by reminding me that higher housing costs strain family budgets.

Third, the second example sets the stage for writing more about how your project will work and what it will produce for the future.

At the end of the first example you have to think about all the possible questions that a grant reviewer might have and then you have to provide some answers.

So the main difference between the two approaches is that the second example focuses the readers mind into the directions and around the questions that you want her or him to think about.

The first example shows exactly what the instructor was telling me; the statistic is used to support the idea of a home ownership center. Its logic model is simply this: If home ownership is lower than average in a community, then the community needs a home ownership center.

Stated that way it is a bit easier to see the jump in logic that the first example asks of the reader.

The second example tells a story. The statistic simply provides documentation to reinforce the story. Indeed, in this example, we may not even be using the right statistic.

The home ownership rate might not really be all that important as compared to another statistic documenting the assertion that "Many families in our community pay more in rent each month than they would have to pay for a single-family home." That statistic, again offered at the end of the story, could reinforce both the difference between rent and mortgage payments in the community as well as the impact on low-income families.

Finally, there is a subtle message here that I want to make explicit. Note how all these statistics, either the one used in the example or the other one proposed, are both grounded in local statistics. For example, one could have used the national average for home ownership as a comparison and created a larger contrast: 60% vs. 66.2%. But comparing the local community to the state makes more sense as the two are more closely linked.

Also, using local statistics reinforces the idea that the organization is on the ground and working in the community. It tells a grant reviewer that the organization has taken the time to study the issue and gathered the best information available. While that does not guarantee the organization will do a good job, it certainly provides a level of assurance over those appeals that cite national information that may or may not apply in the local community.