What is “Benchmarking” of Bureau of Labor Statistics Employment Data?

BLS has released the “preliminary benchmark” information for the Current Employment Statistics (CES) survey, the source of monthly information on jobs.

You know what a bench is

Image of a park bench

and you know what a mark is,

Image of a checkmark

but what pray tell is a benchmark? And what does this preliminary benchmark tell us?

So as not to bury the lead, I’ll let you know that this year’s preliminary estimate of the benchmark revision is a bit bigger than it has been in the last few years. Our preliminary estimate indicates a downward adjustment to March 2019 total nonfarm employment of 501,000. Still, that estimated revision is only -0.3 percent of nonfarm employment. In most years our monthly employment survey has done a good job at estimating the total number of payroll jobs. More details on that below. This year our survey estimates are off more than we would like. Our goal is to provide estimates that are excellent and not just good or pretty good, and that’s why we benchmark the survey data each year.

What is benchmarking and why do we do it?

The CES is a monthly survey of approximately 142,000 businesses and government agencies composed of approximately 689,000 individual worksites. As with all sample-based surveys, CES estimates are subject to sampling error. This means that while we work hard to ensure those 689,000 worksites represent all 10 million worksites in the country, sometimes our sample may not perfectly reflect all worksites. So the monthly CES estimates aren’t exactly the same as if we had counted employment from all 10 million worksites each month. To fix this problem, we “benchmark” the CES data to an actual count of all employees, information that’s only available several months after the initial CES data are published.

In essence, we produce employment information really quickly from a sample of employers, then anchor that information to a complete count of employment once a year.

The primary source of the CES sample is the BLS Quarterly Census of Employment and Wages (QCEW) program, which collects employment and wage data from states’ unemployment insurance tax systems. This is also the main source of the complete count of employment used in the benchmark process. QCEW data are typically available about 5 months after the end of each quarter.

Each year, we re-anchor the sample-based employment estimates to these full population counts for March of the prior year. This process—which we call benchmarking—improves the accuracy of the CES data. That’s because the population counts are not subject to the sampling and modeling errors that may occur with the CES monthly estimates. Since the CES data are re-anchored to March of the last year, CES estimates are typically revised from April of the year prior up to the March benchmark. Then estimates from the benchmark forward to December are revised to reflect the new March employment level.

We will publish the final benchmark revision in February 2020 and will incorporate revisions to data from April 2018 to December 2019. (Thus, we’re not showing a 2019 number in graph and table below). On August 21, BLS released a first look at what this revision will be—what we call the “preliminary benchmark.” This preliminary benchmark gives us an idea of what the revised nonfarm employment estimates for March 2019 will be.

The size of the national benchmark revision is a measure of the accuracy of the CES estimates, and we take pride that these revisions are typically small.

Chart showing differences in nonfarm employment after benchmarking, 2009–18

For total employment nationwide, the absolute annual benchmark revision has averaged about 0.2 percent over the past decade, with a range from −0.7 percent to +0.3 percent.

The following table shows the total payroll employment estimated from the CES before and after the benchmark over the past 10 years. For example, pre-benchmark employment for 2018 was 147.4 million; post-benchmark employment was also 147.4 million.

Nonfarm employment estimates before and after benchmarking, March 2009–March 2018
Year Level before benchmark Level after benchmark Difference Percent difference
2009 132,077,000 131,175,000 -902,000 –0.7
2010 128,958,000 128,584,000 -374,000 –0.3
2011 129,899,000 130,061,000 162,000 0.1
2012 132,081,000 132,505,000 424,000 0.3
2013 134,570,000 134,917,000 347,000 0.3
2014 137,147,000 137,214,000 67,000 <0.05
2015 140,298,000 140,099,000 -199,000 –0.1
2016 142,895,000 142,814,000 -81,000 –0.1
2017 144,940,000 145,078,000 138,000 0.1
2018 147,384,000 147,368,000 -16,000 <-0.05

The 2019 preliminary benchmark revision is following the same pattern, with an estimated difference of -0.3 percent. We provide this first look at the benchmark revision to give data users a sense of what we are seeing in the data. The final benchmark may be a little different—could be higher, could be lower. But based on recent experience, we are confident the benchmark released next February will show only a moderate difference from what we’ve been publishing each month and will validate the accuracy of our monthly CES estimates.

Want to know more? See our Current Employment Statistics webpage, send us an email, or call (202) 691-6555.

Building a Business? Start Here

You have an idea.

It’s time to get serious about it.

Entrepreneurial drive got you to this point, but now it’s time to chart a plan. For that you need a reliable overview of the factors that can lead to a flourishing business — or work against it.

The U.S. Census Bureau’s Business Builder application is designed to provide small business owners with key data to give them a clear-eyed view of their potential market. This data-mapping tool combines data from the Census Bureau’s American Community Survey, Economic Census, and County Business Patterns, and the U.S. Department of Agriculture’s National Agricultural Statistics Service.

For version 2.6 of the tool, released this month, the Bureau of Labor Statistics has collaborated with the Census Bureau to include data from our Quarterly Census of Employment and Wages (QCEW). QCEW is based on quarterly mandatory reports to the Unemployment Insurance systems in each state, covering more than 95 percent of the jobs in the U.S. economy. It is the most complete and current source of data on employment and wages at a detailed geographic and industry level.

To help illustrate why this tool is so useful, and why the data from the QCEW broadens that usefulness, I’ll make up an example.

Ever since you can remember, your grandmother, who was born and raised just outside of Naples, has fed you a type of pizza full of unusual flavors that has never been equaled in all your travels. As you grew and came into your own as a cook, she entrusted you with her secret knowledge, like a magician passing along her repertoire to a favored protégé.

Ever since, you’ve dreamed of sharing the pleasures of that delicacy with the world, and you’re going to start with a pizzeria somewhere near your home in Olympia, Washington. You may ask yourself: What exactly does the restaurant market look like in Olympia? Who are my potential customers? What kind of wages do they earn?

The Census Business Builder is a good place to start.

Census Business Builder home screen

Here, you can enter the type of establishment you’d like to research, as well as the area where you intend to do business. You find that data are not available for Olympia, but knowing that Olympia is the county seat, you are able to search in Thurston County.

The resulting map provides data on income, education, wages, and perhaps most importantly for you, the number of similar establishments in the area – also known as your competition.

Map of Thurston County, Washington, showing Census Business Builder search results

With the new QCEW data, another crucial batch of information is at your fingertips: more up-to-date establishment counts, employment numbers, and wages. It also provides an important metric known as the location quotient. This measure lets you compare an industry’s employment concentration or wages in your search area with the country as a whole. Will you be able to hire enough staff? What might you need to pay them if you want the best in the business?

Map of Thurston County, Washington, showing Census Business Builder search results with QCEW location quotient

The possibilities advance from this example as far as your entrepreneurial mind wants to take them. It is you, after all, who will transform these numbers into the real-world business that fulfills your vision. Our job as public servants is to give you the most relevant tools to realize that transformation. We’re grateful for the opportunity to collaborate with the Census Bureau to bring you this vital information in this user-friendly format.

The Census Business Builder is updated twice per year using feedback that comes from customers and stakeholders, including small business owners, trade associations and other government agencies. The update also adds QCEW data into the Regional Analyst version of the tool, which is designed for chambers of commerce and regional planning staff who need a broad portrait of the people and businesses in their area. The December release, for example, will add more QCEW features to the Regional Analyst version.

BLS publishes data from the QCEW program every quarter in the County Employment and Wages news release. QCEW data are available through our Open Data Access and the QCEW Databases.

Why This Counts: Measuring Occupational Requirements

You probably know that BLS publishes data and analysis about employment, unemployment, job openings, earnings, productivity, occupational safety and health, and more. But did you know we also publish information about how often workers have to lift objects; the maximum weight they lift or carry; whether they work in extreme heat or cold; and how much training and experience they need for a job? We call these characteristics “occupational requirements.”

What are occupational requirements?

The Occupational Requirements Survey provides information about the requirements of jobs:

  • Physical demands of work, such as keyboarding, reaching overhead, lifting or carrying
  • Environmental conditions, such as extreme heat, exposure to outdoors, proximity to moving parts
  • Education, training, and experience requirements, such as prior work experience, on-the-job training, and license requirements
  • Cognitive and mental requirements, such as interaction with other people, independence of work, and the amount of review

How did BLS get into doing this survey?

This survey is one of our newest statistical programs; we first published data on December 1, 2016.

The Social Security Administration asked us to help them obtain accurate and current data to use in their disability programs. They are developing an Occupational Information System, which will use data from the Occupational Requirements Survey. That means the survey is crucial for Social Security to manage their disability programs fairly and efficiently.

How can I use occupational requirements information?

Users of Occupational Requirements Survey data include:

  • Researchers exploring occupational change
  • Jobseekers and students
  • Government agencies evaluating skill gaps
  • People with disabilities and their advocates

Let’s discuss a couple of examples to show you what I mean.

Educational requirements

You may want to know the minimum formal education requirements for jobs. The survey has a stat for that! In 2018, a high school diploma was required for jobs covering 40.7 percent of workers, while 17.9 percent had a bachelor’s degree requirement. The chart below shows the percent of jobs by minimum education requirement.

Percent of jobs with a minimum education requirement, 2018

Editor’s note: Data for this chart are available in the table below.

We have more information on education, training, and experience. The 2018 news release showed that on-the-job training was required for about 77 percent of workers, and the average duration was 34 days.

We also have information on preparation time, which includes minimum formal education, training, and work experience a typical worker needs to perform a job. Preparation time between 4 hours and 1 month was required for 31.5 percent of workers.

Environmental Conditions

Is the noise level at your workplace closer to a library (quiet) or a rock concert (very loud)? For some jobseekers, understanding the noise level and other environmental conditions might be extremely important as they evaluate job options. The chart below provides examples of the noise intensity in different occupations.

Percent of jobs with noise intensity level requirements, selected occupations, 2018

Editor’s note: Data for this chart are available in the table below.

Examples of work environments with different noise intensity levels include:

  • Quiet: private office, a golf course, or art museum
  • Moderate: department stores, business office, or fast food restaurant
  • Loud: manufacturing plant, atop large earth moving equipment, or jobs next to the highway
  • Very loud: rock concert venues, working with jack hammers, or rocket testing areas

How do we collect job requirement data?

To collect job requirement data, our field economists ask business owners, human resource professionals, worker safety officers, and supervisors to collect requirements of work. Field economists do not use paper or online questionnaires to collect these data; instead, they rely on a conversational interviews and descriptive documents, such as task lists, to collect information on occupational requirements.

How are we improving the survey?

Survey scope: Since it began, we have continued to refine the survey to improve its accuracy. In the third year of collection, we redefined the survey scope to focus on critical job functions—that is, the reason the job exists.

Survey content: Beginning with the current sample in collection, we added questions about cognitive and mental requirements. The Social Security Administration asked for this change so we can provide information on the requirements for workers to adapt to changes in the pace of work, solve problems, and interact with others.

Sample: The survey sample is collected over a 5-year period. That will provide the large amount of data necessary to publish information about detailed occupations. We have revised the sampling process to ensure we collect information about less common occupations.

Website: We recently improved the web layout to make it easier for users to find the data they want.

Where is more information?

We have data for occupational groups and occupations through the Occupational Profiles. All data are available through the public data tools. For concepts, methods, and history of the survey see the Handbook of Methods or visit our homepage.

Let us know if you have questions or comments about occupational requirements:

  • Email
  • Phone: (202) 691-6199

Use these gold-standard data to learn more about your job requirements or to find out about new ones. Whatever your occupational requirements question, “We have a stat for that!”

Percent of jobs with a minimum education requirement, 2018
Education requirement Percent
No minimum education requirement 31.5%
High school diploma 40.7
Associate’s degree 3.8
Associate’s vocational degree 2.1
Bachelor’s degree 17.9
Master’s degree 2.3
Professional degree 0.9
Doctorate degree 0.5
Percent of jobs with noise intensity level requirements, selected occupations, 2018
Occupation Quiet Moderate Loud
Bus and truck mechanics and diesel engine specialists 49.0% 51.0%
Computer programmers 60.1
Construction laborers 48.6 51.4
Electricians 49.0 51.0
Highway maintenance workers 46.2 53.8
Home health aides 54.1 45.9
Library technicians 56.0
Medical transcriptionists 68.7
Paralegals and legal assistants 66.5 33.5
Welders, cutters, and welder fitters 48.2 50.9

Baseball, Hot Dogs, and Statistics

Summer is in full swing, which means that when I’m not talking about BLS data, I’m talking about baseball, something I could do full time. Luckily these two topics have a lot in common; nothing quite says statistics in the summertime like baseball does. We fans have followed baseball statistics for nearly as long as the game has been played. Teams today increasingly use statistics—or “analytics”—to decide which players to add to their rosters, who to play in any game situation, and even where to position fielders for certain batters. That’s a lot like the innovations BLS and the other federal statistical agencies focus on to help people make informed decisions for their families, businesses, and the broader economy.

I grew up a St. Louis Cardinals fan until the mid-1960s, when the lamentable Kansas City Athletics moved to Oakland and the American League expansion team, Kansas City Royals, brought winning baseball to Kansas. I had a second home, as it were, in the seats of Kauffman Stadium, or “The K” to us Kansas City Royals fans. Today I cheer for the Washington Nationals, having lived in the D.C. area for the past 25 years. Regardless of your favorite team, if you’re a baseball fan you know statistics are a huge part of how your team performs. While I could talk about George Brett’s batting average, home runs, or wins above replacement, let’s instead look at where America’s agency on labor data meets America’s favorite pastime.

Spectator sports employed over 144,000 workers in 2018. The map below shows the metropolitan areas with the most jobs in spectator sports. The New York-Newark-Jersey City, NY-NJ-PA metro area, home of the Yankees and Mets, employs the most people at 8,674. Spectator sports include all professional and semi-professional sport teams.

Employment in spectator sports by metropolitan area, 2018

Editor’s note: Data for this map are available in the table below.

Also consider the Occupational Employment Statistics program. In 2018, there were 27,780 radio and television announcers and 7,480 public address announcers. There were also 236,970 coaches and scouts and 19,090 umpires and referees in 2018.

Here’s a look at some of the occupations within the spectator sports industry:

Employment in selected occupations in the spectator sports industry, May 2018

Editor’s note: Data for this chart are available in the table below.

If you’re going to the game, the Consumer Price Index can tell you how prices have changed over time. For example, ticket prices for sporting events decreased 0.8 percent from June 2018 to June 2019. Over the same period, parking fees and tolls increased 3.2 percent and prices for food away from home increased 3.1 percent. At the low end of the increases is beer away from home—a modest 0.8 increase over the year.

Or maybe you just decide to enjoy the game in front of your new 80-inch flat screen TV. You aren’t alone. According to the American Time Use Survey, Americans spend, on average, 2.84 hours a day watching television in 2017. That’s almost enough time to watch your typical 9-inning baseball game.

Can we squeeze any more baseball out of BLS statistics? There’s wage data for ushers, occupational injuries for umpires, and productivity in certain recreation industries. But we’ll save that for another day. For now, sit back and enjoy the game. Play Ball!

Employment in spectator sports by metropolitan area, 2018
Metropolitan area Employment
Abilene, TX 14
Akron, OH 215
Albany-Schenectady-Troy, NY 537
Albuquerque, NM 241
Anchorage, AK 204
Ann Arbor, MI 84
Atlanta-Sandy Springs-Roswell, GA 1,948
Augusta-Richmond County, GA-SC 314
Austin-Round Rock, TX 1,071
Bakersfield, CA 229
Baltimore-Columbia-Towson, MD 1,872
Bangor, ME 32
Barnstable Town, MA 9
Baton Rouge, LA 15
Beaumont-Port Arthur, TX 5
Billings, MT 78
Birmingham-Hoover, AL 372
Bloomington, IL 54
Boulder, CO 22
Bridgeport-Stamford-Norwalk, CT 83
Burlington-South Burlington, VT 92
Canton-Massillon, OH 13
Cape Coral-Fort Myers, FL 718
Charleston, WV 118
Charleston-North Charleston, SC 256
Charlotte-Concord-Gastonia, NC-SC 6,081
Chicago-Naperville-Elgin, IL-IN-WI 4,763
Cincinnati, OH-KY-IN 1,473
Cleveland-Elyria, OH 2,019
Coeur d’Alene, ID 36
College Station-Bryan, TX 20
Colorado Springs, CO 214
Columbia, SC 383
Columbus, OH 813
Corpus Christi, TX 153
Dallas-Fort Worth-Arlington, TX 3,392
Deltona-Daytona Beach-Ormond Beach, FL 1,294
Denver-Aurora-Lakewood, CO 999
Des Moines-West Des Moines, IA 375
Detroit-Warren-Dearborn, MI 1,682
Dover, DE 222
Dubuque, IA 72
Duluth, MN-WI 75
El Paso, TX 269
Erie, PA 105
Eugene, OR 67
Evansville, IN-KY 209
Fairbanks, AK 29
Flint, MI 6
Florence, SC 43
Fort Collins, CO 4
Fresno, CA 542
Grand Junction, CO 67
Grand Rapids-Wyoming, MI 234
Great Falls, MT 72
Greeley, CO 96
Greenville, NC 54
Greenville-Anderson-Mauldin, SC 234
Gulfport-Biloxi-Pascagoula, MS 121
Hagerstown-Martinsburg, MD-WV 21
Harrisburg-Carlisle, PA 316
Hartford-West Hartford-East Hartford, CT 233
Hickory-Lenoir-Morganton, NC 111
Hot Springs, AR 665
Houston-The Woodlands-Sugar Land, TX 2,968
Huntington-Ashland, WV-KY-OH 27
Huntsville, AL 47
Indianapolis-Carmel-Anderson, IN 2,786
Jacksonville, FL 1,397
Janesville-Beloit, WI 52
Joplin, MO 11
Kahului-Wailuku-Lahaina, HI 3
Kansas City, MO-KS 1,737
Kennewick-Richland, WA 84
Lake Charles, LA 11
Lakeland-Winter Haven, FL 387
Lancaster, PA 160
Las Cruces, NM 18
Las Vegas-Henderson-Paradise, NV 835
Lebanon, PA 16
Lexington-Fayette, KY 1,556
Lincoln, NE 172
Little Rock-North Little Rock-Conway, AR 148
Los Angeles-Long Beach-Anaheim, CA 7,100
Louisville-Jefferson County, KY-IN 1,337
Lubbock, TX 12
Manchester-Nashua, NH 144
Medford, OR 23
Memphis, TN-MS-AR 1,349
Miami-Fort Lauderdale-West Palm Beach, FL 6,476
Midland, TX 99
Milwaukee-Waukesha-West Allis, WI 2,218
Minneapolis-St. Paul-Bloomington, MN-WI 3,337
Missoula, MT 20
Mobile, AL 79
Myrtle Beach-Conway-North Myrtle Beach, SC-NC 171
Naples-Immokalee-Marco Island, FL 23
Nashville-Davidson–Murfreesboro–Franklin, TN 721
New Orleans-Metairie, LA 1,522
New York-Newark-Jersey City, NY-NJ-PA 8,674
North Port-Sarasota-Bradenton, FL 549
Norwich-New London, CT 85
Ocala, FL 490
Ogden-Clearfield, UT 120
Oklahoma City, OK 1,227
Oxnard-Thousand Oaks-Ventura, CA 82
Palm Bay-Melbourne-Titusville, FL 301
Phoenix-Mesa-Scottsdale, AZ 4,398
Pittsburgh, PA 1,603
Pittsfield, MA 21
Portland-South Portland, ME 317
Portland-Vancouver-Hillsboro, OR-WA 1,042
Providence-Warwick, RI-MA 348
Raleigh, NC 1,339
Reading, PA 198
Richmond, VA 381
Riverside-San Bernardino-Ontario, CA 688
Roanoke, VA 98
Rochester, NY 676
Sacramento–Roseville–Arden-Arcade, CA 957
Salem, OR 56
Salisbury, MD-DE 138
San Antonio-New Braunfels, TX 757
San Diego-Carlsbad, CA 1,720
Santa Rosa, CA 223
Seattle-Tacoma-Bellevue, WA 3,176
Sherman-Denison, TX 6
Sioux Falls, SD 138
Spartanburg, SC 15
Spokane-Spokane Valley, WA 209
St. George, UT 20
St. Joseph, MO-KS 13
St. Louis, MO-IL 1,608
State College, PA 75
Stockton-Lodi, CA 130
Tampa-St. Petersburg-Clearwater, FL 3,302
Trenton, NJ 159
Tucson, AZ 83
Tulsa, OK 281
Virginia Beach-Norfolk-Newport News, VA-NC 148
Washington-Arlington-Alexandria, DC-VA-MD-WV 2,931
Waterloo-Cedar Falls, IA 65
Watertown-Fort Drum, NY 15
Wausau, WI 34
Weirton-Steubenville, WV-OH 18
Wheeling, WV-OH 47
Winston-Salem, NC 414
Worcester, MA-CT 97
Yakima, WA 21
York-Hanover, PA 156
Youngstown-Warren-Boardman, OH-PA 84
Yuba City, CA 34
Employment in selected occupations in the spectator sports industry, May 2018
Occupation Employment
Security guards 8,490
Ushers 7,610
Food and beverage servers 6,980
Groundskeepers 2,350
Parking lot attendants 1,850

New State Data on Labor Productivity and Job Openings and Labor Turnover

While international trade has become increasingly important to our economy over the past 60 years, U.S. households and businesses continue to rely primarily on local markets for most goods and services. The products we create come from all over our country. Workers, businesses, and policymakers care deeply about the economy in our own backyards. That’s why BLS recently began publishing new data on labor productivity by state and, separately, on job openings and labor turnover by state.

State labor productivity

Our measures of labor productivity for states are still experimental, meaning we’re still assessing them and considering ways to improve them. These measures cover the private nonfarm sector for all 50 states and the District of Columbia from 2007 to 2017. They show that labor productivity growth varies a lot from state to state. From 2007 to 2017, labor productivity changes ranged from a gain of 3.1 percent per year in North Dakota to a loss of 0.7 percent per year in Louisiana. In 2017, labor productivity grew fastest in Montana (2.0 percent), West Virginia (1.9 percent), California (1.8 percent), and Hawaii (1.7 percent). You can get the complete dataset from our state labor productivity page.

U.S. map showing productivity growth in the private nonfarm sector in each state from 2007 to 2017

Editor’s note: Data for this map are available in the table below.

We construct these state measures from data published by several BLS programs and by our colleagues at the Bureau of Economic Analysis. A recent Monthly Labor Review article, “BLS publishes experimental state-level labor productivity measures,” explains the data and the methods for putting them all together. The article also highlights how you might use these new state data. We’re happy to have your feedback on these new measures. Just send us an email.

State job openings and labor turnover

We also have new data on job openings, hiring, and separations by state. Data from the Job Openings and Labor Turnover Survey are widely used by economic policymakers and others who want to understand the job flows that lead to net changes in employment. We have these data back to December 2000 and update them every month for the nation and the four broad census regions. Now we have them for all states and the District of Columbia too. These state estimates are available from February 2001 through December 2018 for the total nonfarm sector.

Many of you have told us you want more geographic details about job openings and turnover. To make sense of data on job openings, for example, it helps to know where the jobs are. The survey sample size is designed to estimate job openings and turnover for major industries only at the national and regional levels. For several years we have researched ways to produce model-assisted estimates for states. As with the state productivity data, these estimates are experimental. We plan to update the state estimates each quarter while we assess your feedback on the models and the usefulness of the data. We encourage you to send us your comments.

But wait, there’s more! We’ve updated the BLS Local Data App!

In previous blog posts, we’ve told you about our mobile app for customers who want to know more about local labor markets. This app now includes employment and wage data for detailed industries and occupations. (It doesn’t yet have the new data on state productivity, job openings, and turnover.)

Interested in local data for a particular industry or occupation? The latest version allows you to quickly search or use the built-in industry and occupational lists. Want to know which industry employs the most workers in your area or which occupation pays the highest? The updated app allows you to sort the employment and wage data across groups of industries and occupations. You can still find data on unemployment rates and total employment. You also can find your state, metro area, or county by searching for a zip code or using your device’s current location.

These new data and features result from the continued partnership between BLS and the U.S. Department of Labor’s Office of the Chief Information Officer. Be on the lookout for more new features to be added in future releases.

Download the BLS Local Data app from the App Store or Google Play today!

Annual percent change in labor productivity in the private nonfarm sector, 2007–17
State Annual percent change
North Dakota 3.1
California 1.7
Oregon 1.7
Washington 1.7
Colorado 1.6
Oklahoma 1.6
Maryland 1.5
Montana 1.5
Pennsylvania 1.5
Massachusetts 1.4
New Mexico 1.4
Vermont 1.4
Idaho 1.3
Kansas 1.3
Nebraska 1.1
New Hampshire 1.1
South Carolina 1.1
Tennessee 1.1
Texas 1.1
West Virginia 1.1
Alabama 1.0
Hawaii 1.0
Kentucky 1.0
Minnesota 1.0
New York 1.0
Rhode Island 1.0
South Dakota 1.0
Virginia 1.0
Georgia 0.9
Arkansas 0.8
Missouri 0.8
Ohio 0.8
Utah 0.8
Illinois 0.7
North Carolina 0.7
Delaware 0.6
Florida 0.6
Iowa 0.6
Indiana 0.5
Mississippi 0.5
New Jersey 0.5
Wisconsin 0.5
Alaska 0.4
Arizona 0.4
District of Columbia 0.4
Michigan 0.4
Maine 0.3
Nevada 0.3
Wyoming 0.1
Connecticut -0.5
Louisiana -0.7