Improving How We Measure Prices for New Vehicles

We have a guest blogger for this edition of Commissioner’s Corner. Brendan Williams is an economist in the Office of Prices and Living Conditions at the U.S. Bureau of Labor Statistics.

For nearly as long as cars and trucks have been sold, the BLS Consumer Price Index (CPI) has tracked changes in the prices consumers pay for new vehicles. Our traditional method of determining the change in vehicle prices is to survey dealers and collect estimated prices for models with a specific set of features. For example, a Brand X 8-cylinder two-door sports coupe with a sunroof. We recently debuted a research index for new vehicles based on a large dataset of prices actually paid, which we call “transaction” prices. This is just one of many efforts currently underway in the CPI (and throughout BLS) to identify and introduce new sources of data into our statistical measures. As you are about to learn, a lot goes into introducing these new measures.

We purchased the new data for new vehicles from J.D. Power. The new dataset includes records of the prices paid during hundreds of thousands of transactions every month—far more than the roughly 2,000 vehicle prices in the CPI sample. The larger dataset provides more precise measures of price change.

But it’s not as simple as plugging the new data into the monthly CPI. We found that applying current CPI methods to the transaction data produced a biased index. So we had to make some changes. We combined an estimate of the long-run trend in new vehicle prices with a measure of high-frequency fluctuations in the market. The long-run trend is based on the year-over-year price change between a vehicle in the current month and the same vehicle in the prior model year 12 months ago; we get these values from the J.D. Power data. The high-frequency fluctuation is extracted from a monthly index based on current methods used in the CPI.

The research index includes all types of new vehicles—cars, SUVs, and trucks. And since the data reflect actual transactions, the shift in consumer preference from cars to other types of vehicles is reflected in the data. This differs from the currently published CPI, which has maintained a roughly equal weight between cars and trucks.

The new vehicles research index performs very similarly to the published index. From December 2007 to March 2020, the research index (untaxed) increased 8.2 percent, while the official new vehicles index (which is taxed) increased 7.7 percent. Looking under the hood, the research truck index is also similar to its published index. The difference in the car indexes is larger, with the official index showing a 5.2-percent increase, while the research index shows only a 1.5-percent increase.

Chart showing trends in research and official price indexes for new vehicles, 2007 to 2020

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

While the new vehicle indexes look similar, the research index has a much lower standard error, which means there is less variation in the data. The research index had a 12-month standard error of 0.11, compared to the 0.43 standard error in the new vehicles index.

This research index is just one of many ways BLS is innovating the CPI and all our measures. For more information on BLS efforts to use new sources of data in the CPI, see “Big Data in the U.S. Consumer Price Index: Experiences & Plans.” Details of the methods and other aspects of research are in, “A New Vehicles Transaction Price Index: Offsetting the Effects of Price Discrimination and Product Cycle Bias with a Year-Over-Year Index.”

We are asking for your feedback about whether to use this research index or the current index. We specifically want to know whether you think this proposal improves our methods and data sources. Please tell us what you think about the research new vehicles data by emailing cpixnv@bls.gov. You can send other CPI-related questions to cpi_info@bls.gov.

Research and official price indexes for new vehicles
MonthResearch index, trucks untaxedOfficial index, trucks untaxedResearch index, all vehicles untaxedOfficial index, all vehicles untaxedResearch index, cars untaxedOfficial index, cars untaxed

Dec 2007

100.0100.0100.0100.0100.0100.0

Jan 2008

99.9100.299.6100.199.2100.0

Feb 2008

100.199.999.899.799.599.7

Mar 2008

100.899.3100.299.399.699.5

Apr 2008

99.998.799.698.999.399.2

May 2008

99.698.199.698.599.699.1

Jun 2008

100.197.7100.898.4101.599.2

Jul 2008

98.797.1100.098.3101.499.6

Aug 2008

96.395.898.397.6100.799.3

Sep 2008

95.794.797.996.9100.599.0

Oct 2008

95.894.797.896.8100.398.9

Nov 2008

95.294.797.296.999.999.0

Dec 2008

94.094.795.996.898.598.9

Jan 2009

94.095.595.797.597.899.5

Feb 2009

95.296.796.498.298.199.7

Mar 2009

95.297.496.398.597.899.7

Apr 2009

96.697.897.498.798.699.8

May 2009

96.898.197.698.998.699.9

Jun 2009

97.098.697.499.397.9100.1

Jul 2009

96.698.996.699.696.9100.3

Aug 2009

96.997.797.098.197.498.7

Sep 2009

99.098.099.498.5100.199.0

Oct 2009

98.899.899.3100.4100.0101.1

Nov 2009

99.2100.799.5101.6100.0102.5

Dec 2009

99.3100.999.2101.699.3102.5

Jan 2010

99.3101.199.2101.599.3102.1

Feb 2010

99.8101.499.5101.699.4102.1

Mar 2010

100.4101.4100.2101.4100.2101.7

Apr 2010

100.9101.2100.7101.198.3101.3

May 2010

101.0100.8100.8100.8100.7101.1

Jun 2010

101.3100.6100.9100.6100.7101.0

Jul 2010

101.5100.5101.1100.598.2100.8

Aug 2010

101.7100.5101.2100.3100.6100.6

Sep 2010

101.7100.7100.9100.5100.0100.8

Oct 2010

102.3101.0101.2100.999.7101.1

Nov 2010

102.5101.5101.2101.199.4101.2

Dec 2010

102.3101.9100.8101.498.9101.3

Jan 2011

102.4102.4100.8101.798.7101.3

Feb 2011

102.7103.3101.1102.699.2102.4

Mar 2011

103.7103.8102.0103.199.9102.9

Apr 2011

104.3104.0103.0103.5101.4103.5

May 2011

104.7104.3103.8104.3102.7104.7

Jun 2011

104.6104.3103.8104.7103.1105.5

Jul 2011

104.4104.0103.7104.5103.1105.4

Aug 2011

104.3103.7103.6104.1103.2105.1

Sep 2011

104.1103.6103.5104.1103.4105.2

Oct 2011

104.2103.8103.5104.3103.1105.2

Nov 2011

104.3104.1103.4104.4102.6105.2

Dec 2011

104.4104.3103.5104.6102.5105.3

Jan 2012

105.0105.0103.9105.0102.7105.4

Feb 2012

105.1105.9104.0105.6102.8105.8

Mar 2012

105.4106.0104.5105.6103.5105.7

Apr 2012

105.7106.1104.8105.7103.8105.9

May 2012

105.2105.8104.4105.7103.5105.9

Jun 2012

105.4105.8104.5105.6103.5105.9

Jul 2012

105.1105.5104.1105.3103.1105.5

Aug 2012

105.0105.5104.1105.2103.1105.4

Sep 2012

105.2105.6104.3105.2103.3105.3

Oct 2012

105.3105.8104.5105.4103.7105.4

Nov 2012

105.6106.2104.6105.9103.4106.1

Dec 2012

105.7106.5104.5106.2103.0106.4

Jan 2013

105.7107.1104.6106.7103.1106.8

Feb 2013

106.3107.2105.1106.8103.5106.8

Mar 2013

106.4107.4105.2106.8103.6106.8

Apr 2013

106.7107.7105.5107.0103.8106.8

May 2013

106.8107.6105.5106.8103.8106.6

Jun 2013

106.4107.8105.1106.9103.3106.4

Jul 2013

106.4107.6105.0106.6103.2106.1

Aug 2013

106.4107.3105.0106.3103.2105.8

Sep 2013

106.3107.6104.9106.4102.9105.8

Oct 2013

106.5107.6105.1106.5103.2105.7

Nov 2013

106.7107.8105.1106.6103.0105.8

Dec 2013

106.4108.0104.6106.7102.0105.9

Jan 2014

106.5108.1104.6106.7101.8106.0

Feb 2014

107.1108.6105.2107.1102.3106.3

Mar 2014

107.3108.6105.3107.1102.4106.2

Apr 2014

107.8109.0105.7107.4102.6106.4

May 2014

108.1108.9105.8107.3102.4106.4

Jun 2014

107.9108.4105.5106.9101.8106.0

Jul 2014

108.2108.6105.7106.9101.9105.9

Aug 2014

108.6108.7105.9106.7101.7105.4

Sep 2014

108.4108.7105.6106.7101.3105.4

Oct 2014

108.7109.0105.9107.1101.5105.7

Nov 2014

108.5109.2105.5107.2100.8105.9

Dec 2014

108.3109.4105.1107.2100.0105.8

Jan 2015

109.0109.3105.8107.2100.9105.8

Feb 2015

109.2109.9106.0107.8101.0106.4

Mar 2015

109.4110.2106.2108.0101.1106.5

Apr 2015

109.8110.5106.6108.2101.6106.5

May 2015

109.7110.6106.4108.2101.3106.5

Jun 2015

109.9110.5106.5108.2101.3106.5

Jul 2015

109.7110.2106.2107.7100.9105.9

Aug 2015

110.0109.8106.3107.3100.5105.5

Sep 2015

110.5109.8106.7107.2100.6105.3

Oct 2015

110.5109.8106.6107.2100.4105.2

Nov 2015

110.6110.2106.5107.499.9105.2

Dec 2015

111.0110.1106.9107.4100.4105.3

Jan 2016

111.5110.6107.3107.9100.7105.8

Feb 2016

111.8111.2107.7108.5101.2106.4

Mar 2016

112.0111.4107.8108.5101.1106.2

Apr 2016

112.2111.2108.0108.2101.3105.8

May 2016

111.9111.0107.6108.0100.7105.6

Jun 2016

111.9110.8107.4107.7100.1105.2

Jul 2016

111.1110.7106.8107.7100.0105.0

Aug 2016

111.8110.3107.3107.499.8104.7

Sep 2016

111.5110.3106.9107.299.5104.6

Oct 2016

111.3110.6106.7107.599.1104.9

Nov 2016

110.9110.6106.4107.699.0105.0

Dec 2016

111.1110.9106.5107.898.8105.1

Jan 2017

112.0111.9107.4108.999.8106.3

Feb 2017

111.8111.9107.3109.0100.0106.5

Mar 2017

112.1111.7107.3108.799.5106.0

Apr 2017

112.1111.7107.3108.699.3105.9

May 2017

111.9111.6107.1108.399.2105.5

Jun 2017

112.0111.1107.1107.899.1104.9

Jul 2017

111.9110.4106.9107.098.4103.9

Aug 2017

111.8110.2106.6106.697.9103.4

Sep 2017

111.4109.8106.3106.197.6102.8

Oct 2017

111.5109.7106.5106.097.9102.7

Nov 2017

112.0109.9106.8106.497.4103.2

Dec 2017

111.4110.7106.3107.297.9104.0

Jan 2018

111.9111.0106.9107.698.7104.4

Feb 2018

111.8110.8106.9107.498.9104.2

Mar 2018

111.2110.8106.3107.498.3104.2

Apr 2018

111.4110.3106.7106.999.3103.7

May 2018

111.1110.5106.4107.198.8104.1

Jun 2018

110.9110.6106.3107.299.1104.2

Jul 2018

111.3110.5106.7107.299.4104.3

Aug 2018

111.4110.2106.8106.999.5104.0

Sep 2018

111.3109.8106.8106.699.8103.9

Oct 2018

111.2109.6106.8106.5100.0103.9

Nov 2018

111.5109.8107.0106.799.9104.1

Dec 2018

110.7110.0106.3106.999.6104.2

Jan 2019

111.3110.8106.8107.6100.0104.8

Feb 2019

111.7111.0107.2107.7100.2104.9

Mar 2019

111.6111.5107.1108.199.9105.2

Apr 2019

112.0111.5107.4108.2100.1105.2

May 2019

112.2111.3107.6108.0100.3105.2

Jun 2019

111.7111.0107.2107.9100.6105.2

Jul 2019

111.9110.7107.4107.6100.6104.9

Aug 2019

111.5110.3106.9107.2100.2104.6

Sep 2019

111.6109.9107.1106.7100.1104.1

Oct 2019

111.9109.8107.3106.6100.3104.1

Nov 2019

111.3109.9106.8106.6100.0104.1

Dec 2019

111.2110.4106.8107.099.8104.3

Jan 2020

111.8111.0107.4107.7100.4105.1

Feb 2020

112.2111.4107.7108.2101.0105.7

Mar 2020

112.7110.9108.2107.7101.5105.2

When Worlds Converge: Statistics Agencies Learning from Each Other during the Pandemic

We never know when our worlds are going to converge. I have used this blog to tell you about how BLS operations are continuing—and changing—due to the COVID-19 pandemic. I also plan to tell you about our international activities and will continue writing about the BLS Consumer Price Index (CPI) and other programs. Today, all three of these topics converge into one.

The COVID-19 pandemic has compelled BLS and statistical agencies worldwide to examine our processes and concepts to ensure the information we collect and publish reflects current conditions. For BLS, this means suspending all in-person data collection and relying on other methods, including telephone, internet, and email. Adding to our toolbox, BLS is now piloting video data collection. To be flexible, we have changed some collection procedures to accommodate current conditions. For example, we are now doing all of our work at home instead of in our offices. We are learning more every day about teleworking more effectively, and we are training our staff as we learn.

Once we collect the data, we are examining how we need to adapt our processing and publication. Will our typical procedures to account for missing data still apply? Will seasonal patterns in the data change due to COVID-19? Will we be able to publish the level of detail our data users have come to expect? These and more are open questions. We will make informed decisions as we learn more about the pandemic’s impact on our data and operations. What I do know is that BLS has a long practice of sharing its procedures and methods, including any changes. We already have extensive information about COVID-19 on the BLS website, and we continue to update that information. We also provide program-specific information with each data release to alert users to any unique circumstances in the data.

Since BLS has long been known for producing gold-standard data, information about our procedures and methods is also of great interest to our international colleagues. In fact, BLS has helped statistical organizations throughout the world with the collection, processing, analysis, publishing, and use of economic and labor statistics for more than 70 years. We provide this assistance primarily by our Division of International Technical Cooperation. They strengthen statistical development by organizing seminars, consultations, and meetings for international visitors with BLS staff. This division also serves as the main point of contact for the many international statistical organizations that compile information, publish comparable statistics worldwide, share concepts and definitions, and work to incorporate improvements and innovations.

A hallmark of our international activities has been onsite seminars at BLS, often attended by a multinational group of statistical experts and those working to become experts. At these seminars, BLS technical staff present details on every aspect of statistical programs, including concept development, sampling, data collection, estimation procedures, publishing, and more. In recent years, funding, travel restrictions, and other limitations have reduced the number of in-person events, replaced to some extent by virtual events. And of course, the current COVID-19 pandemic and related travel restrictions mean all such events are now being held virtually. But they still go on.

Recently, our international operations converged with our COVID-19 response when the International Technical Cooperation staff set up a virtual meeting between BLS staff primarily from our Consumer Price Index program and their counterparts at India’s Ministry of Statistics and Programme Implementation (MOSPI). They met to discuss challenges in producing consumer price data during the ongoing pandemic. The discussion was largely about methodology: what to do with missing prices and how to adjust weights to reflect real-time shifts in spending that consumers are making in response to the pandemic. It is helpful to hear from worldwide colleagues who are facing similar challenges. These issues are unprecedented, and we know the potential solutions for one country may not be ideal for the nuanced conditions in another country.

In India, for instance, commerce has been limited to essential commodities—food, fuel, and medicine. This will likely leave them unable to publish some indexes. While this is unfortunate in the present time, it is fairly straightforward; they can’t publish what they don’t have. It gets more complicated a year from now. What does it mean to have an annual price change when the denominator is missing? The CPI deals with this by having a fairly robust imputation system—basically “borrowing” price change from similar areas and items—but we will be monitoring the situation closely to make sure our assumptions about what is similar remain valid.

One advantage BLS has over MOSPI is that we are able to collect data by telephone, email, or on the web. MOSPI has traditionally only done in-person collection. Both agencies are transitioning to different modes of collection, but we have significantly greater experience.

Sharing information with our international colleagues, about the CPI and other programs, and about our COVID-19 experience, is a key part of the BLS mission. These worlds continue to converge, not just during organized meetings but also on websites and wikis maintained by statistical organizations and through participation in expert groups and conferences. For example, the United Nations Economic Commission for Europe hosts a ”statswiki” that currently has pages dedicated to COVID-19 and Official Statistics. It is a small world after all, and the worldwide social distancing we are all experiencing makes it clear that we are all in this together. And together, BLS and our international colleagues, reacting to COVID-19 and making adjustments to consumer price indexes and other statistics, will continue to provide vital information that tracks changes in the world economy.

Paid Leave Benefits When You Are Unable to Work

Many American workers have lost jobs or had their work hours reduced as a result of the COVID-19 pandemic and response efforts. Many other workers still have jobs, but their work environment probably has changed since March. It’s reasonable to assume more people are working from home now than the 29 percent we reported who could work at home in 2017–18. At BLS we are still working to provide you with the latest economic data and analysis, but nearly all of us are now working from home, instead of in our offices.

Still, there are many jobs that just can’t be done from home. In these challenging times, I know we all are grateful for the healthcare workers who are treating patients who have COVID-19 and other medical conditions. We’re grateful for our emergency responders and for the truck drivers, warehouse workers, delivery workers, and staff in grocery stores, pharmacies, and other retail establishments that provide us with the necessities of daily life. As much as I think of these men and women as superheroes, I know they are humans. Even extraordinary humans can get sick, or they may need to take care of family members who get sick. Let’s look at the leave benefits available to them if they need it.

According to our National Compensation Survey, 73 percent of private industry workers were covered by paid sick leave in 2019. Among state and local government workers, 91 percent were covered by paid sick leave. The availability of sick leave benefits varied by occupation, ranging from 94 percent of managers in private industry to 56 percent of workers in construction and extraction occupations.

The share with paid sick leave also varies by industry, pay level, size of establishment, and other characteristics of jobs and employers. The following chart shows sick leave availability for employers of different sizes.

Percent of workers in private industry with access to paid sick leave by establishment size, March 2019

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

Paid sick leave plans commonly provide a fixed number of days per year. The number of days may vary by the worker’s length of service with the employer. The average in private industry in 2019 was 7 paid sick leave days.

Average number of paid sick leave days per year for workers in private industry, by length of service and establishment size, March 2019

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

About half of workers with such a plan could carry over unused days from year to year.

We recently posted a new fact sheet on paid sick leave that provides even more detail.

In the past few years, some states and cities have mandated that certain employers provide their workers with paid sick leave. We include these mandated plans in our data on paid leave. A Federal law passed in March 2020 requires paid sick leave for certain workers affected by COVID-19.

In addition to paid sick leave, some employers offer a short-term disability insurance plan when employees can’t work because of illness. These plans are sometimes called sickness and accident insurance plans. This was traditionally a blue-collar or union benefit, and it often replaces only a portion of an employee’s pay. In 2019, 42 percent of private industry workers had access to such a benefit. Like sick leave, the availability of short-term disability benefits varies widely across worker groups. Some states provide Temporary Disability Insurance plans that provide similar benefits.

While the National Compensation Survey asks employers what benefits they offer to workers, the American Time Use Survey recently asked workers whether paid leave is available from their employer and whether they used it. In 2017–18, two-thirds of workers had access to paid leave at their jobs. These data include information on age, sex, and other characteristics. For example, younger workers (ages 15–24) and older workers (age 65 and older) were less likely to have access to paid leave than were other workers.

Percent of workers with access to paid leave by age, 2017–18 averages

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

While the survey did not ask workers to classify the type of leave, they were asked the reasons they could take leave. Of those with paid leave available, 94 percent could use it for their own illness or medical care, and 78 percent could use it for the illness or medical care of another family member.

I hope you and your loved ones remain healthy and are able to take care of each other in these challenging times. High-quality data will be vital in the public health response to the COVID-19 pandemic. High-quality data also will be vital for measuring the economic impact of the pandemic and recovery from it. My colleagues at BLS and our fellow U.S. statistical agencies remain on the job to provide you with gold standard data.

Percent of workers in private industry with access to paid sick leave by establishment size, March 2019
Establishment sizePercent

1–49 workers

64%

50–99 workers

68

100–499 workers

80

500 workers or more

89
Average number of paid sick leave days per year for workers in private industry, by length of service and establishment size, March 2019
Length of serviceAll establishments 1 to 49 workers50 to 99 workers100 to 499 workers500 workers or more

After 1 year

76678

After 5 years

77679

After 10 years

77779

After 20 years

77779
Percent of workers with access to paid leave by age, 2017–18 averages
AgePercent

Ages 15–24

35.4%

Ages 25–34

70.3

Ages 35–44

71.7

Ages 45–54

74.4

Ages 55–64

74.2

Age 65 and older

51.7

How Have We Improved the Consumer Price Index? Let Me Count the Ways

Soon after I became Commissioner, the top-notch BLS staff was briefing me on the many programs and details that make up BLS. I asked the staff working on the Consumer Price Index (CPI) if they could list for me some of the improvements that have occurred over the past few years. It has been nearly a quarter century since the Boskin Commission studied the CPI and recommended enhancements. I knew many of these enhancements had been implemented, along with other improvements. But I was shocked to see my staff come back with an 8-page, detailed listing of 77 substantive improvements that have been implemented since 1996.

You may have thought a price index that has been around since 1913 is happy to rest on its laurels. Well, you’d be wrong. There are improvements to the CPI going on all the time. As I reviewed the list, I noticed a number of improvements involve the introduction of new or changed goods and services, such as cell phones or streaming services. I also noticed improvements in how we handle product changes, as I wrote recently in a blog about quality adjustment. But these topics only scratch the surface.

I’m not going to describe every CPI enhancement that has taken place over the past 24 years; you can find much more detail on the CPI webpage. But I will whet your appetite by highlighting a few categories of improvements.

Keeping the CPI market basket up to date

The goal of the CPI is to track the change in the prices consumers pay for a representative market basket of goods and services. Let’s look at the what and where of that market basket.

  • What we collect — goods and services. We collect prices for a market basket of goods and services, designed to represent what consumers are buying. In January 2002, we switched from updating that market basket every 10 years to every 2 years, providing a better representation of current spending patterns.
  • What we collect — housing. We track the cost of housing by a separate sample of housing units. In 2010, we increased that sample to improve accuracy. In 2016 we began rotating that sample every 6 years. Previously, the housing sample was only rotated when new geographic areas were introduced after the U.S. census.
  • Where we collect — outlets. We collect prices from stores and businesses that are chosen based on where consumers shop and buy goods and services. In January 1998, we switched from updating this sample of “outlets” every 5 years to every 4 years. And in January 2020, we switched the source used to determine those outlets to the Consumer Expenditure Survey, which is also the source of our spending information for the market basket of goods and services people buy. Previously we used a separate survey of households to identify outlets.
  • Where we collect — geography. We collect prices for goods and services in selected geographic areas, designed to represent all urban areas of the United States. In January 2018, we updated the geographic areas, designed to represent current population trends. We last updated these areas in 1998.

Collecting CPI information

According to folklore, CPI data collection was accomplished by staff who dressed up in high fashion, the ladies in fancy hats and white gloves and the gentlemen in the finest haberdashery, who then went shopping to determine the latest prices. That’s not how it happens. CPI staff are not paid to “shop” to collect prices. We use trained experts who are skilled at gaining cooperation from many different types of businesses; ensuring they are obtaining price information for goods and services that are consistent from one month to the next or making appropriate adjustments; and gathering information from thousands of outlets about hundreds of thousands of goods and services over a short data-collection period. Let’s look at how the data-collection process has improved:

  • When we collect — In June 2005, the CPI switched from a collection period that spanned the first 15–18 days of the month to collection across the entire month. This provides more representative data, especially for items that frequently vary in price within the same month.
  • How we collect — In January 1998, the CPI began using computerized data-collection tools, which automate certain math functions and screen for errors or inconsistencies. We continue to upgrade our processes; CPI data-collection staff recently began using a new generation of tablet computers.
  • Alternative collection — Not all price information comes from traditional collection with stores. Some information comes from websites, corporate data files, third parties that combine data from different sources, and more. In fact, the CPI and other BLS programs are focused on identifying even more alternative collection methods in the coming years.

Calculating the CPI

Once we collect the prices on all these goods and services, we need to calculate an index. In simple terms, we find the difference between the price in month 1 and the price in month 2, and express that difference as a rate of change from month 1. We publish rates of change and also express current prices as an index, which is equal to 100 in a base period.

Many factors and decisions go into combining data for an item and then combining data for all items into the published CPI. We’ve improved those calculations in several ways over the past few years.

  • Geometric mean — In January 1999, the CPI switched the formula for calculating price changes at the component item level from an arithmetic mean to a geometric mean. This allows the overall index to capture substitutions consumers make across specific products within a component item category when the prices of those products change relative to one another. With the geometric mean formula, BLS does not assume consumers substitute hamburgers for steak, which are in different component categories. The formula only captures substitution within a component category, such as among types of steak.
  • More decimal places — In January 2007, the CPI began publishing index numbers to 3 decimal places, which improved consistency between published index numbers and rates of change.

New information available to the public

While the CPI has been around for over a century, we have added a number of new indexes over time, to provide a variety of inflation figures. Here are some of the newest products in the CPI family:

  • Research (Retroactive) series — The CPI Research Series incorporates many of the improvements that came out of the Boskin Commission. The series provides a pretty consistent way to measure price changes from 1978 up to the most recent full year.
  • Chained CPI series – The Chained CPI uses an alternative formula that applies spending data in consecutive months to reflect any substitution that consumers make across component item categories in response to changes in relative prices. For example, this index would capture consumer substitution of hamburger for steak. This measure is designed to come closer to a “cost-of-living” index than other BLS measures. The series was first produced in 2002.
  • Elderly research series – The CPI for the Elderly reweights the component CPI data based on the spending patterns of elderly households. This series, mandated by Congress, began in 1988. In 2008, we extended the series retrospectively back to 1982.

As I have mentioned in the past, we are always working to improve the CPI. We recently contracted with the Committee on National Statistics, part of the National Academy of Sciences, to provide guidance on a variety of issues. I’ll use this space to report on the Committee’s work, as well as other improvements underway in the CPI.

State Productivity: A BLS Production

We have a guest blogger for this edition of Commissioner’s Corner. Jennifer Price is an economist in the Office of Productivity and Technology at the U.S. Bureau of Labor Statistics. She enjoys watching theatrical performances when she’s not working.

I recently had the pleasure of attending a high school play. The cast was composed of a male and female lead and at least a dozen supporting actors. The program listed the performers and acknowledged many other students, parents, teachers, and administrators. They all played some important role to bring the play to life—lighting, sound, painting props, sewing costumes, creating promotional materials, selling tickets, working concessions. All of these pieces came together harmoniously to make the performance a success.

Setting the Stage: New Measures of State Productivity

We can view the health of the nation’s economy through the same lens. Our diversified economy is made up of lead performers and supporting roles in the form of industries. Some industries contribute more heavily to growth in output or productivity, playing the star role. Other industries are supporting characters, contributing to a smaller, but necessary, share of growth. Our productivity program recently published a webpage that examines how industries contribute to the nation’s private business output and productivity growth.

We also can examine these roles geographically. Until recently, BLS productivity measures were only produced at the national level. Last June, BLS published experimental measures of state labor productivity for the private nonfarm business sector. These measures, which cover the period from 2007 to 2017, will help us learn more about productivity growth in each state and how each state contributes to national productivity trends.

Measuring productivity for all states allows us to credit the role played by each state, not just the total performance of the national economy or region. Just as each person, no matter how small their role, was necessary for the success of the school play, each state contributes to how we evaluate national or regional productivity. When we examine the contribution of each state to total productivity trends, we find that, like actors, no two states perform identically. Similar individual growth rates may have different impacts on the productivity of the nation or region. By analyzing state productivity trends over the long term, we learn more about regional business cycles, regional income inequality, and the role of local regulations and taxes on growth.

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.

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

We estimate each state’s annual contribution to national or regional productivity growth by multiplying the state’s productivity growth rate by its average share of total current dollar national or regional output. The economic size of each state influences its contribution to national and regional estimates. From 2007 to 2017, California was our lead performer, with the largest contribution to national productivity growth. The state’s productivity grew 1.7 percent per year on average, and its large economy means it contributed more than one-fifth of the 1.0-percent growth in national labor productivity.

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

Supporting actors included Texas and New York. Making a cameo appearance was North Dakota; despite having the largest productivity growth rate, it ranked 28th in terms of its contribution to national productivity growth. Stars in each region included Illinois (Midwest), New York (Northeast), Texas (South), and California (West). Understudies—those states with the largest growth rates—were North Dakota (Midwest), Pennsylvania (Northeast), and Oklahoma (South). Oregon and Washington shared this role out West.

Second Act

For now, our new measures cover the private nonfarm sector for all 50 states and the District of Columbia from 2007 to 2017. These measures include output per hour, output, hours, unit labor costs, hourly compensation, and real hourly compensation. Our measures of labor productivity for states are experimental, meaning we’re still assessing them and considering ways to improve them. In the second act, we will be looking into producing state-level measures for more detailed sectors and industries.

For an encore performance, check out our state labor productivity page. We’d love to hear your feedback! Email comments to productivity@bls.gov.

Annual percent change in labor productivity in the private nonfarm sector, 2007–17
StateAnnual 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
States with the largest contributions to national labor productivity, average annual percent change, 2007–17
StateState contribution to U.S. labor productivity

California

0.22

Texas

0.10

New York

0.08

Pennsylvania

0.06

Washington

0.04

Massachusetts

0.04

Illinois

0.03