3 Ways Nonprofits Can Use Demographics to Improve Outcomes

Analyzing Demographics for Better Nonprofit Outcomes

If you run a social services nonprofit, then you probably already collect participant demographic data. Demographics are simply data about the people you serve. Gender, race, income, and family structure are all fairly standard demographic fields we’d find in the US Census form. Demographic data about our participants can also include ‘expanded’ fields such as: has access to reliable transportation, is lactose intolerant, years of work experience, lives safe & affordable housing, and much more.

Nonprofits can use the standard and expanded demographic data they collect to understand the strengths and areas for growth of their programs. This post explains a few ways to do that.

Workforce Development Nonprofit - Example Data

We have a fictitious workforce development nonprofit. We offer resume workshops, soft skills training, interview practice, and office software basics (e.g. email, spreadsheets, word processors) and more. Last year our workforce development nonprofit did the following:

Outputs

• 250 adults assisted with resumes

• 100 adults completed soft skills training

• 125 adults completed practice interviews

• 75 adults completed office software basics

• 500 hours of free daycare

Outcomes

• 200 unduplicated adults obtained full-time employment

• 100 unduplicated adults obtained part-time employment

Total Participants

• 425 unduplicated (i.e. unique) adults received services.

What a great year!

We can tell our board and donors how much great work we did, and they’ll be so impressed. For our annual report, we can throw in some nice charts, a few smiling faces, maybe a success story, and call it a day.

The Big Picture Is Good, And You Can Learn More by Digging Deeper

We collected data about our work all year, and it’s distilled to this little list. What can we learn from this summary of results? Are we doing really well? Are we reaching our intended participants? Are we succeeding with adults who are easiest to support, but rarely succeeding with those who need more?

Our big-picture summary is great, but we need to dig deeper to answer these questions. We can use (you guessed it!) demographic data to learn a lot more about our participants and how well our program is performing.

3 Ways to Learn More With Demographic Analysis

1. Examine outputs and outcomes by standard demographics

By standard demographics, I mean the demographics we think about on a standard Census form - gender, race and ethnicity, educational attainment, income, etc.

2. Examine outputs and outcomes of ‘expanded’ demographics

Expanded demographics are still just demographics (i.e. data about our participant population), but they are outside of what we might immediately think about when we hear the word “demographics.” This might include criminal justice system involvement, health, percent of income spent on housing.

3. Compare our program demographics to the demographics of our target participants.

Let’s work through a few examples to help illustrate the point.

Example 1: Examine Outputs and Outcomes by Standard Demographics

We’ll examine some of our outputs and outcomes using one of the standard demographics we collect. This type of demographic analysis allows us to ask and answer hard questions about our programs using the data we already collect.

Program Priority: There's a huge need for our services among single parents in our community, so people from single parent households are a priority.

Table 1: Outputs and Outcomes Broken Down By Single Parent Status (Counts)

Table 2: Outputs and Outcomes Broken Down By Single Parent Status (Percentages)

Output and Outcome Highlights

• Participants from Single Parent Households are 47% of our total participants served. That’s less than half of all participants, but they are the single largest group (200 Participants out of 425 Total).
  

• Participants from Single Parent Households comprise the largest participants in Resume Assistance (60% of total participants), but they are less likely to receive Office Software Basics training (33%) compared to their percentage of all participants (47%).
  

• Single parent participants are much less likely to get full-time jobs (25%) and a little more likely to receive part time jobs (50%) compared to their percentage of all participants (47%).

So, are we doing a good job serving single parents?

It’s hard to be certain. We can see clear areas of strength and areas for improvement.

Some key questions I would try to answer:

1. Why aren’t single parents participating in Office Software Basics with higher frequency?

2. Why aren’t single parents getting more full-time jobs?

3. Should we serve even more single parents as a percentage of all participants (i.e. more than 47%). If so, how will we recruit more and help them be more successful?

Data Analysis Tip 1: Look at raw counts and percentages to get a better understanding of your data. Raw counts give you a clear understanding of the number of participants served, while percentages can help you understand how well target populations are represented

in your outputs and outcomes.

Example 2: Examine outputs and outcomes of ‘expanded’ demographics.

Program Priority: A key priority for our program is to help people who spend higher percentages of their income on rent get better jobs so that they will be more stable

Examining expanded demographics is very similar to examining standard demographics. We are simply including data that isn’t necessarily the standard 'census form' demographics.

In this example, we look at a continuous value as our “expanded” demographic: the percentage of income spent on rent. Since it’s a continuous value, it can be anything between 0% and 100% or more. We create 3 categories: housing costs under 30% of income (affordable), between 30% and 60% of income (unaffordable), and over 60% of income (extremely unaffordable) to make our analysis easier.

Data Analysis Tip 2: when your data is a continuous value (e.g. income), consider using that data to create categories that represent a range of values.

We include tables and charts to make the data more visual. The tables and charts present the same data, so feel free focus on the presentation you prefer.

Terminology Tip: A Table is a numeric presentation of data that is organized into columns and rows that are labeled. A Chart, Figure, or Graph is a visual presentation of your data such as a bar chart or pie chart.

Table 3: Outputs and Outcomes by Percent of Income Spent on Housing (Percentages)

Figure 1 Description: Bar Chart presenting outputs data (resume assistance, soft skills training, office software basics) from Table 3 in a visual manner along with Total Participants served. Figure 1 presents each output category and the percentage of participants in each housing expenditure category.

Output Highlights

• The vast majority of our participants have unaffordable housing costs. Less than 20% of our total participants spend under 30% of income on housing, so the remaining 80%+ of our program participants spend more than 30% of their income on housing.
  

• Participants with very high housing costs (over 60% of income) are less likely to participant in Soft Skills Training and Office Software Basics.
  

• Participants with high housing costs (30-60% of income) are more likely to participant in Office Software Basics.

Figure 1 Description: Bar Chart presenting outcomes data (Full-Time Job Obtained, Part-Time-Job Obtained) from Table 3 in a visual manner along with Total Participants Served. Figure 1 presents each outcome category and the percentage of participants in each housing expenditure category.

Outcome Highlights

• Participants with high housing costs obtain the vast majority of full-time and part-time jobs.
  

• Participants with lower housing costs (below 30% of income) are somewhat more likely to obtain a full-time job and less likely to obtain a part time job.
  

Are we doing a good job serving people with high housing costs?

We see that we are serving a lot of people with high housing costs and a lot of them are benefitting from our work. But something is missing. We aren't sure if we are serving people with high housing costs as effectively as people with lower housing costs. To understand that, we have to look at the data differently.

Table 4 and Figure 3 help us do that by examining the percent of participants in each housing cost category who obtain jobs. Again, they show the same data in different ways, so focus on the approach you prefer.

Table 4: Percent of Participants Obtaining Employment by Income Spent On Housing

Outcome Highlights

• Participants with affordable housing costs (i.e. spending less than 30% of income) are more likely to obtain a full-time job and any job compared to participants with higher percentages of income dedicated to housing.
  

• Participants with the highest housing costs (over 60% of income) are the least likely to obtain full-time or any job.

• 
  

So, are we doing a good job serving people with high housing costs?

Overall, it appears that the program effectively targets participants with high housing costs. Participants with high housing costs are more likely to participate in the program.

But, participants with high housing costs are also less likely to benefit from the program. More than 80% of participants with lower housing costs obtain some kind of job compared to only 64% of participants with very high housing costs.

Some key questions I would try to answer:

1. Why aren’t participants with very high housing costs (greater than 60% of income) participating in Office Software Basics or Soft Skills Training with higher frequency?

2. Do we need to change our programs to enable people with very high housing costs to participate in Office Software Basics and Soft Skills training at higher rates?

3. Should we be serving more or fewer participants with high (30-60% of income) and very high( 60%+ of income) housing costs?

4. What can we do to improve outcomes for participants with higher housing costs given that they are much less likely to obtain employment compared to participants with lower housing costs (i.e. under 30% of income)?.

Data Analysis Tip 3: It’s important to look at your data from multiple perspectives to get a complete understanding of your data.

Example 3: compare your program demographics to the demographics of your target participants.

Most human services organizations have a target population or target geography. Your target population could be a demographic category or set of categories (single parents, people without a college degree, etc). You might also target a geographic area (e.g. postal code 99999 or a neighborhood) for your services.

In this example, we will look at demographic data for a targeted geographic area compared to our program's demographic data. With this analysis, we can learn who in our target community our program is reaching, and we can also use it to understand our program's effectiveness.

Program Priority: We will people with significant barriers to stable employment.

Table 5: Comparison of Target Geography Demographics to Program Demographics

Demographic Comparison Highlights

• Our program serves a high percentage single parents compared to their prevalence in the target area (47% in the program vs 30% in target area)

• Our program serves households with high housing costs at a lower rate than they exist in the target area (53% in our program vs. 64% in the target area)

• Substantially fewer participants in our program (16%) speak English as a secondary language compared to the target area (33%)

• Before participating in our program, participants have much lower incomes compared to the median income in the target area.

• After participating in our program, participants have similar, but still lower incomes compared to the median income in the target area.

So, are we serving our target area as intended?

Remember, our stated priority is to work with people with the significant barriers to stable employment. It’s up to us how we define “significant barriers,” but we can see that we are under-serving people in our target area with high housing costs as well as those for whom English is not their first language.

Some key questions I would ask:

1. Should we serve more participants with high housing costs? If so, what changes, if any, do we need to make to facilitate that?

2. Should we serve more participants who don’t speak English as their first language? If so, how can we reach out to them and provide effective services?

3. We know that participants with higher housing costs are less likely to get jobs (see above), so how will serving more of them impact our program’s ability to increase incomes? Should we care about that?

Data Analysis Tip 4: When comparing program demographics to target demographics, raw counts are useless in almost all cases. Use percentages or average values.

Getting Started With Demographic Analysis

Whew...we made it...so how do you get started?

The first step is, of course, collecting demographic data on your participants - particularly the data you care about or you think might be informative. You probably already have some of that data if you have any funding from grants or government. Ideally, that data is associated with your participants on an individual level.

Next, think about whether your mission, vision, or values, compel you to serve certain people (e.g. single moms) or communities (e.g. postal code 99999). If so, make sure you capture the demographic data that reflects those priorities.

With that data, you can begin with the relatively simple comparisons outlined above. Even these simple comparisons can tell us a powerful story about who we are serving, how well we are serving them, and they lead us to questions about how we can adjust our services to get closer to our ideal results.

There are plenty more (and more complex) ways to use your demographic data to understand your programs, but following these 3 examples is a great start.

Reporting your impact is hard when you’re juggling spreadsheets. countbubble makes it easy so you can focus on your mission.

countbubble is nonprofit data management software simplified. We help nonprofits track and report participant demographics with ease.

Feel free to email us at contact@countbubble.com to start a conversation, or sign up for email updates.

Ryan Brooks

Founder, CountBubble, LLC

Please connect with us on social media: Facebook and LinkedIn