Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Using the CHAS data for demonstrating local housing needs

In December 2009, HUD's Office of Policy Development and Research (PD&R) released updated CHAS data ( One important use for this data is the development of local housing plans, including the Consolidated Plan required for HUD's formula grants such as CDBG. Because many communities are currently in the process of updating their 5 year Consolidated Plan, we rushed to release the data as soon as possible. Unfortunately, we haven't yet completed the development of an online table generating tool similar to the one in the State of the Cities Data System (SOCDS). As a result, using the CHAS data to demonstrate local housing needs may require more original analytical work than in previous years where one could simply plug the numbers into the CPMP tool.

Partially because of this, HUD is <strong>not requiring grantees to use the 2009 CHAS data in the Consolidated Plans currently being developed</strong>.

If you're still reading, you aren't particularly eager to use 9 year old Census 2000 data or to do your own survey of housing needs. This forum exists to help the CHAS data users like yourself share ideas. I will do my best to moderate the forum.

Some suggestions:
<li>First, please consult the resources already available on HUDUser, especially the CHAS basic users guide (</li>
<li>The most useful thing to share would probably be step by step instructions for generating a particular indicator of housing need. For instance, if you've done the leg work to show what percentage of extremely low income renters have a housing problem, describe how you did it (Note: this is the example I did in the CHAS basic users guide).</li>
<li>Feel free to share more general thoughts, ideas, or questions about the CHAS data, but first consider whether they will help other CHAS users. If so, post them here. If not, email me at</li></ul>


The 2009 CHAS tables would be very helpful in providing numbers to support housing policy decisions. In addition, the 2009 CHAS tables are great for updating consolidated plans. The CHAS tables are provided in user friendly Excel files, and I had no trouble downloading and accessing the files. With excellent instructions available, I was able to easily sort through and choose subgroups of households.

There are some new tables in the 2009 CHAS which I found very helpful. Table 10 provides numbers that help identify households at risk of homelessness. Table 13 provides the number of households with potenial for lead-based paint problems.

It is important to understand that the data are not actually 2009 data but are derived from the 2005-2007 American Community Survey (ACS), so it would be helpful to make that clear in the documentation. The 2009 CHAS tables are so much better organized and clearer than the 2000 CHAS, but that means that a user must be careful about updating earlier numbers. Also, some of the data available in 2000 are not provided in the 2009 CHAS tables. In one instance, I used two tables to derive some numbers that weren't available outright in either table. Also, in order to get a level of detail not provided, I estimated some numbers based on 2009 CHAS totals and category percentages from 2000, but many users would not be able to do that.

Policy makers, grant writers, and persons responsible for consolidated plans, should be thrilled that the 2009 CHAS tables are available. Thanks to HUD for making the effort to provide this wonderful updated data resource!
I am trying to create one table for Tenure by Household size by income (very low, low, mod) Does anyone know how to combine these?
jwoods, it's not possible to create that table from the CHAS data currently available. Tables 4 and 7 contain information about tenure and household size, but neither contains income breaks. We will change that with the next CHAS release.

Also, you should not combine tables. If we don't produce a table with the exact combination of characteristics you are interested in, there's no other way to get them through the CHAS.
I think the housing plan made is better than the previous one. It is a way too similar but in some cases it is different and better than he previous one. Anyways, thanks for the information .I enjoyed reading it. But if you could post some extra information then it may be helpful to many others.
I am attempting to determine the number of households by income group (0 to 30%, etc) in Texas using the CHAS 2009 data. When I use a filter on Household_Income to exclude ALL I can no longer select T in the line type. Then I am left with only 117 counties, out of Texas' 254. As I move forward with my data analysis more clarification on the difference between Line type (T, D, S) would be helpful. The data dictionary lists T=total, D=Data, and S=subtotal. But what do these terms actually mean, when is it appropriate to select T versus D, or S?

Suzanne, your question is a very good one. Sorry for the delay in responding.

Line type "T" means total, so it only appears once in each table. The "estimate" has not been restricted by any of the variables in the table. Line type "S" means subtotal. That means some of the variables are active while others are not. If any of the filters is set to "All" then you will get a subtotal. Line type "D" means that none of the variables in the table will be set to "All"; each variable has one of the specific options chosen.

This is important because you should not add across the different line types. That would be like adding the population of Chicago, the population of Cook County, and the population of Illinois state all together; the sum will be far more than the actual population of the state of Illinois.

The subtotals are useful if you are not interested in all of the characteristics in a table. If you only want to know the number of households under 30% AMI, using Table 1, you could use a subtotal rather than adding up all "D" lines.

I am working on a 5-year consolidated plan in Michigan and have looked at filling out the Housing Needs table provided in the CPMP Tool. If what I've read on this forum so far is true, I can't "combine" tables using proportions to infer the number of households in each required category. Looking at the 2009 CHAS Data and the Needs table, if combining tables is not an option, the only stats I can accurately enter are the number of households that are Renter, Elderly Total & w/Any Housing Problem for the three income brackets (<30%, 30%-50%, and 50%-80%). It seems to me there is no way to acurately calculate any other stats in the table since none of the 2009 CHAS tables combine more than 3 of the 4 variables I need to complete the rest of the Needs table. Is that true? Do you have any suggestions for how the rest of this can be completed or calculated? Thanks for your help.


Forum Jump:

Users browsing this thread: 1 Guest(s)