Tuesday, June 27, 2017

HealthLandscape Tip: Isolating a Geography

One of our most frequent questions from users is, “Can I make a map that shows only my state or only my county?”  While there is not an official way to isolate an area like this, we did come up with one way you might do it.  Keep in mind that we do not recommend isolating geography in this manner because state and county boundaries, while they have implications for health policy and health spending, are just imaginary lines.  Things occur just on the other side of those lines that may affect the health of the area you are highlighting.  This includes providers from whom people in your area of interest may be seeking health care services, environmental effects, etc.  That said today, I'd like to share a technique for isolating a geography in our Population Health Mapper using the QuickThemes tool.  The same steps can be taken in any HealthLandscape tool.

The QuickThemes tool in the HealthLandscape platform allows users to color in states, counties, ZCTAs, etc. based on data in a table.  Each area that is included in your table will be colored in based on the value in the column you choose to map.  For example, if I had a data table like this that has a list of areas at least one column of data that has a number, percent or qualitative data in it:

Qualitative Information
District of Columbia
West Virginia

With this table, I can add color to the map for these six states based on the information in either the number, percent or qualitative information columns.  The process I describe below takes advantage of this by instructing you to add a column of data with identical values for all the areas you want to mask.  Let’s get started.

The Population Health Mapper first prompts you to select a state; for the purposes of this demonstration I will choose Maryland. When you do this, the Mapper will zoom you to the area you have selected and automatically center the map for you, as it did for me.

Next, I need a list of states with their respective FIPS codes. FIPS stand for Federal Information Processing Series. The US Census Bureau maintains several pages related to FIPS codes, including an explanation of the codes and tables with state FIPS codes and county FIPS codes.


Using the state table provided by the Census Bureau, I will cut and paste the data into an Excel spreadsheet and delete the row for Maryland. By deleting Maryland from the list, it will create a “hole” in my map layer that I am adding.  In order to have the other states colored in, I have to add a column and insert an identical value to each state. For the purposes of this blog, I will name my column "Value for Monochrome" and will insert the numeral 5 for each state, but keep in mind that these values will appear on your map, so you may want to name them things and select values that blend in with the theme of your map better.


Back in the Population Health Mapper, I will start the QuickThemes tool by going to the top right "Tools" button (gray), clicking it and selecting "QuickThemes".  The QuickThemes tool will open and be added to the Tools Accordion on the right. Read the instructions and then go back to the spreadsheet to highlight and copy the rows and columns that have the information you want to include on the map; this will include your header row.  When ready, in the QuickThemes tool click the "Click here to begin" button to get to the screen where you will paste your data.

Click anywhere in the white area and then hit Ctrl+v or Command+v to paste your data.


You will have to give the QuickThemes tool some information.  First , tell it the type of geographic code you are using; I will select "State by FIPS".  You will also have to tell it which column in your dataset has the geographic identifiers in the ID selection box.  In my case, I will select FIPS Code. Last, under Group/Category selection box, select the column that has the single value in it for all states, in my case, "Value for Monochrome". Finally, hit the blue "Load" button.  Hint: If you forget to do the Group/ Category selection before you hit load, you can always do it later.

Using the Zoom Bar at the top left I will zoom to a good level to view states and grab the map with my mouse (left click and hold) and drag the map so that Maryland will be centered. The resulting map has my state of focus, Maryland, uncolored (and ready for the application of points, labels, and other data) and the surrounding states are visible but monochromatic in the background. Note this method can be applied within a state at the county level using the county FIPS codes.

I hope this is helpful in your work. You can contact me for assistance using the chat feature on this tool or the "Contact Us" feature if you are working in the UDS Mapper.

Keith Gardner
User Engagement Specialist


Wednesday, June 7, 2017

The Census Counts: Notes on the critical importance of the Census

Earlier this month, the Director of the Census Bureau abruptly retired after many decades of service to the Bureau.  There is much speculation and concern due to the suddenness of the Director's departure, which, combined with expected shortfalls in funding required for the coming 2020 Census, raises real concerns for the agency’s ability to meet its Constitutionally required obligations.  While it will be some time before we learn the full impact of this sudden leadership upheaval, it would do us well to remind ourselves why we care about the Decennial Census and all the related activities of the Census Bureau.

The Census has been conducted every ten years since 1790, becoming increasingly complex and expensive with each decade.  In 2010, the Census cost a record $13 billion dollars and employed over 600,000 temporary workers.  The Decennial Census is routinely referred to as the largest peacetime activity of the United States Government.  There are shelves full of books explaining the importance and utility of the Census (and the related annual American Community Survey), and I want to focus on what I consider to be the most critical.

Geospatial analysis depends on an accurate Census. HealthLandscape is a web-based platform that allows users to visualize and analyze their data spatially.  To aid in that analysis, we provide a comprehensive library of demographic, social, behavioral, economic, and health-rated data.  Two rich sources of data for the data library are the decennial Census and American Community Survey, to name just a few of the incredible data programs from the Census Bureau.  Geospatial analysis would be far less valuable and provide fewer insights without the ability to compare users’ data to related Census data.

Business and Commerce rely on an accurate Census.  It’s been said that “What gets measured gets managed”.  Census data, and projections derived from these data, are core components of business planning in the United States. Retail outlets use the data to inform decisions for opening (or closing) brick-and-mortar locations.  Transportation planners use the information to plan new roads and highways.  The location of regional and larger airline network hubs is driven by the size of the population and changes that can be anticipated.  School districts use the information to determine whether to build new buildings or combine grades across multiple buildings to best meet the growing (or shrinking) student populations.  

Federal Expenditures require an accurate Census.  The allocation and redistribution of Federal tax revenue is based on Census counts for states, counties, cities, tracts and school districts, to name a few.  As such, Census counts equate to economic power.  Local, state, and national government organizations recognize this and work hard to make sure the decennial Census and related annual surveys are complete and accurate.  A recent Brookings Institution report shows that under-representation in the Census (people not completing the decennial Census form) costs states between $382 (Utah) and $2,564 (Vermont) PER PERSON in annual tax revenue redistribution.  For example, the State of Ohio, with a population of 11.6 million people has an expected $814 potential per capita loss for each person missed in the Census.  A one percent undercount would mean nearly 95 million dollars in lost federal expenditures (money returning to the State) EACH YEAR.  

Democracy and political franchise demands an accurate Census.   At a basic level, completing a Census form and being counted is every bit as important as fulfilling your obligation to vote.  The distributions of representatives at the state and national level are driven by Census counts.  As a result of the 2010 Census, the State of Ohio lost two seats in the US House of Representatives, while Texas gained four, with other states gaining and losing as well.  At the end of all the work, completing the Census form and being counted represents YOUR most basic level of political power.

Since the original 1790 Census directed by Thomas Jefferson, the United States has been fortunate to have a high-quality, apolitical snapshot of the US population.  As we move into the planning and testing phases for the 2020 Census, it's important that we remind ourselves of the critical value of this decennial effort.

Remember: If you aren’t counted, you don’t count.

Mark A. Carrozza, MA