The map above shows spending on health care per person in each U.S. state, with darker colors indicating higher amounts. Various factors in each state influence the composition of these expenditures: the age structure of the population, income level, level of competition among health care providers, and local policies and regulations. Thus, everyone can develop an interpretation of why some states spend more on health care based on, for example, older populations, higher incomes, greater market power of health care providers, and policies that lead to more spending.
As it turns out, the story hasn’t changed much over the past 20 years. The map below shows that expenditures in 1997 don’t look much different from expenditures in 2017. Relatively speaking, of course: Expenditures have at least doubled since then, but the fundamental forces that drive health care costs across states seem persistent. For example, New York, Pennsylvania, and New Jersey still spend more than Virginia, Kentucky, and Tennessee, which still spend more than Utah, Nevada, and Arizona.
How these maps were created: The original post referenced interactive maps from our now discontinued GeoFRED site. The revised post provides replacement maps from FRED’s new mapping tool. To create FRED maps, go to the data series page in question and look for the green “VIEW MAP” button at the top right of the graph. See this post for instructions to edit a FRED map. Only series with a green map button can be mapped.
The Bureau of Labor Statistics (BLS) measures payroll employment with their establishment survey—or, more formally, their Current Employment Statistics survey. The establishment survey records the number of jobs on company payrolls on the 12th of each month. The recent partial U.S. government shutdown presents a good opportunity to look at how shifts in government payrolls might affect payrolls overall.
Because the establishment survey does not consider furloughed workers to be off the payroll, any decline in payroll employment as a result of a government shutdown will arise from either quits or formal layoffs in the government sector. Of course, contagion can also occur in the private sector: For example, demand for hotel rooms in Washington, D.C., may be lower until the government reopens or airlines may see a drop in demand if airport security lines become too long. The BLS estimates that the recent partial government shutdown (the longest in history, from December 22 to January 25, 2019) has not substantially affected federal employment. For now, let’s look at three previous government shutdowns highlighted in the FRED graph.
The dark blue line in the graph shows the monthly change in overall nonfarm payroll employment. Keep in mind that changes in overall payroll employment may not necessarily be caused by government shutdowns. The light blue line shows the monthly change in federal government payroll employment, and the red vertical lines show three long-term government shutdowns: October 1978, November-December 1995, and October 2013.
During these shutdowns, federal employment as measured by the establishment survey didn’t change substantially: +7,000, -21,000, and -14,000, respectively. But shutdowns may have an effect on overall payrolls. The list below shows the changes in payroll employment in the month of, the month after, and the second month after these three previous government shutdowns:
October 1978 shutdown: October +335,000, November +435,000, December +280,000
November-December 1995 shutdown: November +144,000, December +146,000, January -5,000
October 2013 shutdown: October +225,000, November +267,000, December +67,000
As we can see from the data, in previous shutdowns, private payroll employment growth has tended to decrease in the second month after the shutdown, before rebounding a few months later.
During the most recent shutdown, the BLS remained funded and continued to conduct their surveys. In fact, they recently released their January numbers: +1,000 jobs on government payrolls and +304,000 on payrolls overall. If the current experience plays out in the same way as it did in the past, we can expect payroll growth to decline temporarily in March 2019.
How this graph was created: Search for “All Employees Total Nonfarm Payrolls” in FRED and select the series. Next, select the “Add Line” option in the “Edit Graph” menu. Search for “All Employees: Government: Federal” and click “Add Data Series.” From the “Format” tab, move the federal government employees series to the right axis. Using the “Edit lines” menu, click on each line and set the units to “Monthly change, thousands of persons.” Last, go back to the “Add Line” menu and select the link to create a user-defined line. Enter the dates of the three shutdowns in both the start and end categories. For example, for the October 2013 shutdown enter “2013-10-01” in both the start and end boxes. Then set the value start/end to be the max and min values of the left axis. Repeat this for all three shutdowns. Change the colors of the added lines to red on the “Format” menu.
In October 2009, in the wake of the Great Recession, the U.S. unemployment rate peaked at 10%. This economy-wide number is useful but masks important regional patterns. To reveal a more detailed picture, we use GeoFRED to look back at county-level unemployment in October 2009.
In this map, counties are divided into three types, according to their unemployment rates:
a rate above the economy-wide peak of 10%
a rate between 4% (the current rate as of January 2019) and 10%
and a rate below 4%.
Counties in the first group, with rates above 10%, were concentrated on the West Coast and in the Midwest and South Atlantic regions. Counties in the second group, with below-average rates, include other parts of the West and a significant portion of counties in the Northeast (e.g., Wyoming, New York, and Massachusetts).
Counties in the third group, with unemployment below 4%, are concentrated in a column that includes North Dakota, South Dakota, Nebraska, and Kansas. This is no mean feat. In October 2009, the U.S. economy was still reeling from the recession and the financial crisis. But even amidst these poor economic conditions and a national unemployment rate of 10% (the highest since April 1983), these counties managed to maintain extremely low rates of unemployment—lower, in fact, than the current economy-wide rate of 4%, which is exceptionally low by historical standards and has been aided by 10 years of economic expansion.
How this map was created: The original post referenced an interactive map from our now discontinued GeoFRED site. The revised post provides a replacement map from FRED’s new mapping tool. To create FRED maps, go to the data series page in question and look for the green “VIEW MAP” button at the top right of the graph. See this post for instructions to edit a FRED map. Only series with a green map button can be mapped.
