FEBRUARY 2018 Volume 38 Number 2 ISSN 0160-3345

The NORTH SLOPE OIL PATCH Jobs in some of the na on’s largest oil fi elds PAGE 4 By NEAL FRIED

BEST ESTIMATES SHOW ONGOING JOB LOSS Clearing up some poten ally confusing data PAGE 9 By DAN ROBINSON

GAUGING ALASKA’s ECONOMY PAGE 14

To sign up for a free electronic subscrip on, read past issues online, or purchase a print subscrip on, visit labor.alaska.gov/trends. Contact the editor at (907) 465-6561 or [email protected].

ALASKA DEPARTMENT of LABOR Dan Robinson Sara Whitney Sam Dapcevich and WORKFORCE Chief, Research and Analysis Editor Cover Ar st DEVELOPMENT Bill Walker Governor ON THE COVER: BP’s Lisburne Produc on Center at Prudhoe Bay, photo courtesy of BP Alaska Heidi Drygas ON PAGE 4: Deadhorse, photo by Flickr user Dan Love Commissioner h ps://crea vecommons.org/licenses/by-nc-sa/4.0/legalcode

Alaska Economic Trends is a monthly publica on meant to objec vely inform the public about a variety of economic issues in the state. Trends is funded by the Employment and Training Services Division of the Alaska Department of Labor and Workforce Development and is published by the department’s Research and Analysis Sec on. Material in this publica on is public informa on, and with appropriate credit may be reproduced without permission.

2 FEBRUARY 2018 ALASKA ECONOMIC TRENDS Gas project will create thousands of jobs for Alaskans

The oil and gas sector is the largest pri- training centers, career and technical vate economic driver in the state, and education providers, registered appren- this month’s edition of Trends examines ticeship programs, and the University of employment in North Slope oil fi elds. Alaska system. We are fortunate to have The Alaska Department of Labor and the only comprehensive pipeline industry Workforce Development has been helping training center in the United States: the generations of Alaskans gain the skills Fairbanks Pipeline Training Center. required to work in this important indus- try, and we are building on that legacy While a fi nal agreement on the Alaska as we prepare Alaskans for thousands of LNG project is not expected until the jobs that will be created by the Alaska end of 2018, construction may begin as LNG project. soon as 2019, and expanding the capacity of Alaska’s existing training programs Heidi Drygas The Alaska LNG project is designed and institutions is critical to ensuring Commissioner to move Alaska’s North Slope gas to maximum Alaskan employment on the tidewater, with offtake points along project. the 807-mile pipeline that will provide natural gas for in-state customers. At the Maximizing Alaska resident hire also pipeline’s terminus in Nikiski, the gas requires we increase the number of will be liquefi ed and shipped by sea to programs helping Alaskan high school Asia. Its construction will create an es- students transition to postsecondary timated 12,000 direct jobs with another education or training, registered ap- 1,000 long-term jobs for the operation of prenticeship, and university programs. the project. The economic impact of this A key component to achieving this will project is also expected to create thou- be increasing the number of qualifi ed ca- sands of indirect jobs. reer and technical education instructors for secondary, postsecondary, and ap- In his state of the state address, Governor prenticeship training. The department’s Bill Walker announced his support for workforce development plan will call for Follow the Alaska a strong project labor agreement, which deeper investment in career and technical Department of will put skilled Alaskans fi rst in line to education to ensure the next generation Labor and Workforce Development on work on the project. One of our missions of Alaskans enter the workforce prepared Facebook (facebook. at the department is to ensure the Alas- for employment on the Alaska LNG proj- com/alaskalabor) kan workforce has the skills and experi- ect and beyond. and Twi er (twi er. ence necessary to build and operate the com/alaskalabor) Alaska LNG project. To meet this goal, The Alaska Department of Labor and for the latest the department is developing a strategic Workforce Development is committed news about jobs, to ensuring the Alaskan workforce has workplace safety, workforce development plan to align ex- and workforce isting resources and amplify the ability the skills and experience required for development. to train Alaskans for high-demand jobs this project. With a strong project labor associated with the project. agreement that puts Alaskans fi rst, and a coordinated effort to align and increase Alaskans are already training for the the capacity of training and education myriad occupations required to construct partners throughout the state, we will and operate the Alaska LNG project. succeed in preparing Alaskans for the Alaska has a robust network of regional thousands of job opportunities that will training centers, joint apprenticeship be created by the Alaska LNG project.

ALASKA ECONOMIC TRENDS FEBRUARY 2018 3 THE NNorthorth SlopeSlope OOilil PPatchatch

Jobs in some of the na on’s largest oil fi elds

By NEAL FRIED Industry makeup of the Slope

rudhoe Bay, home to the na on’s largest oil Alaska began producing employment numbers for the fi eld, is what single-handedly transformed Slope in 1986, seven years a er oil began to fl ow down PAlaska into an oil-producing powerhouse and the Trans-Alaska Pipeline and 18 years a er the fi eld’s became the source of the state’s greatest wealth. discovery. (See Exhibit 1.) These numbers include both According to historian Terrance Cole, “The balance direct oil and gas industry jobs and all other employ- sheet of Alaskan history is simple: One Prudhoe Bay ment in the oil patch because in this unique loca on, is worth more in real dollars than everything that has nearly everything is ed to oil and gas ac vity. been dug out, cut down, caught, or killed in Alaska since the beginning Many of those suppor ng jobs are cat- of me.” And yet, it’s a place few The oil industry egorized in professional and business Alaskans ever visit. services and include everything from represents 3 percent engineering and geological fi rms to fa- of Alaska’s employ- cility support services and waste man- Oil-related ac vity has since spread agement and remedia on. The Slope well beyond Prudhoe Bay, and this ment, and two- also has a substan al number of jobs ar cle uses the terms “oil patch” thirds of those jobs in the leisure and hospitality sector, and “the Slope” to refer to the en- are on the Slope. employers that operate the camps and re oil industrial complex in the other facili es that feed and house the area, including Prudhoe Bay and Ku- large workforce. (See Exhibit 2.) paruk but also Moose’s Tooth to the west, Point Thomson to the east, and any other area Two other large categories of oil patch employers are in the North Slope Borough that is touched by oil. construc on and transporta on companies, as there’s plenty to build and maintain as well as thousands of Because this ar cle’s focus is oil and gas-related ac- workers and materials to transport. vity, it excludes employment in the North Slope Bor- ough’s eight Inupiat communi es. For more on those Some industries are notably absent on the Slope. For communi es, see “When The North Slope is Home” example, there’s almost no employment in retail or in the September 2016 edi on of Alaska Economic government. In contrast, these represent over a third Trends. of all jobs statewide.

4 FEBRUARY 2018 ALASKA ECONOMIC TRENDS Prudhoe Bay oil fi eld, photo by Flickr user North Slope Pamela A. Miller Oil Patch

The oil patch’s Historical Oil Patch Employment historical ebb and fl ow 1 A , 1986 2017* Over the last 30 years, employment levels

have fl uctuated from year to year, some- 14,000 $120 mes considerably. During the fi rst two 12,000 decades, Slope employment reached a $100 high of 6,524 in 1990, two years a er 10,000 oil produc on peaked, then dwindled to $80

4,816 by 1999. (See Exhibit 1.) 8,000 Employment $60 At the me, this waxing and waning 6,000 seemed drama c and vola le, but in $40 hindsight, the bandwidth of oil patch 4,000 employment stayed mostly within a ght Oil prices $20 range of 5,000 to 6,000. 2,000

0 $0 The overarching declining trend that 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015 2017* began in the 1990s was punctuated by periods of recovery, but jobs remained *Based on the fi rst nine months of 2017 Note: Employment numbers include all industries. below the 6,000 mark un l 2003. In the Source: Alaska Department of Labor and Workforce Development, Research and Analysis early 2000s, with produc on declining, it SecƟ on seemed unlikely that Slope employment would ever top 6,000 again. Consequently, the oil in- In 2006, the oil patch resumed adding jobs at a strong dustry’s long-term job outlook was bleak. The 10-year pace. In addi on to the tonic provided by four years industry forecast we published in 2006, for example, of above-average oil prices, maintenance and work predicted no growth from 2004 to 2014. on a number of new fi elds breathed life into the in- dustry, and early that year a sec on of BP’s pipeline Oil produc on was down to less than half its peak in sprung a leak that turned into the largest oil spill in 2006, and the downward trajectory was broadly ac- North Slope history and resulted in millions spent on cepted as permanent with employment levels expect- repairs. ed to follow. But that didn’t happen. Oil prices began to rise in 2003 and by 2005 had more than doubled. This “mature” oil province was now the fastest-grow-

ALASKA ECONOMIC TRENDS FEBRUARY 2018 5 ing employment area in the state and oil was the fast- est growing industry. In 2007, oil patch employment Jobs Mostly in Oil Produc on hit a new record of 7,781 — more than 1,000 jobs O ’ , 2016 above the 1990 peak. 2 Government All Other 0.3% 1.9% The price of oil soared to a record $133 per barrel in Leisure/ 2008, and employment reached 10,000 that year for Hospitality the fi rst me. Prices so ened some during the U.S. 5.3% recession but remained above $100 per barrel from Professional 2011 to 2014. and Business Services 15.3% A er 2008, employment hit a new record each year Financial 0.4% TransportaƟon1.8% un l topping out at 12,540 average monthly jobs for Retail 0.2% Wholesale0.3% ConstrucƟon Oil and Gas 2015. Monthly employment hit a high of 13,485 that 3.7% March. (See Exhibit 3.) 70.8% Less oil produced per worker

Employment was at an all- me high even as produc- on con nued a long-term declining trend in what Source: Alaska Department of Labor and Workforce Development, Re- had been North America’s most produc ve oil fi eld. search and Analysis SecƟ on (See Exhibit 4.) Most new fi elds were smaller, requir- ing more investment and more workers to produce a barrel of oil. The high job levels didn’t last

In terms of the produc on-to-worker ra o, the peak Oil prices began to fall in late 2014 and dropped to was 372 barrels a day per worker in 1988, which fell to $40 a barrel by 2015. Employment began to follow less than half that by 2000 (163 barrels). The ra o con- suit later that year. By the second quarter of 2017, nued to drop, dipping below 100 barrels per worker in employment had fallen to 2007 levels, erasing most of 2005 and reached a low of 45 in 2016. the past decade’s gains.

Two-Year Job Decline from 2015 Peak About the data N , 2015 2017 3 This article’s data include Employment peak employment in all industries in the North Slope’s oil patch, 13,485 12,843 as running these virtually self- 12,517 12,325 12,330 12,106 11,741 11,898 contained camps on the North Slope requires a range of sup- 10,369 Lowest level since 2007 9,946 port workers in addition to oil 9,652 9,443 9,563 9,645 8,923 industry workers.

In this article, “the oil patch” and “the Slope” refer only to the oil fi elds and related oil ac- tivity and exclude North Slope Borough employment in its eight communities: Anaktuvuk Pass, Atqasuk, Utqiagvik (was Barrow), Kaktovik, Nuiqsut, Point Hope, Point Lay, and Jul Jun Jun Jun Oct Oct Apr Apr Apr Sep Feb Feb Sep Feb July Dec Dec Aug Aug Nov Nov Mar Mar Mar May May May Wainwright. To learn more Jan 2015 Jan Jan 2016 Jan 2017 Jan about these communities, see the September 2016 issue of Note: Employment numbers include all industries. Source: Alaska Department of Labor and Workforce Development, Research and Alaska Economic Trends. Analysis SecƟ on

6 FEBRUARY 2018 ALASKA ECONOMIC TRENDS Na onal oil industry employment began to recover in late 2016, and Alaska’s overall oil and gas job losses Long Decline in Produc on began to moderate in 2017. (See Exhibit 5.) 4 N , 1978 2017 Prices have inched up to around $70, possibly high enough to stabilize the industry, and planned explora- Millions of barrels on and maintenance on the North Slope in 2018 are 2,000 also likely to stem further losses. Another posi ve for 1,800 1,600 the industry is that oil produc on has increased over 1,400 the past two years. 1,200 1,000 800 A mostly imported workforce 600 One of the most striking aspects of the oil patch 400 workforce is that it’s almost en rely imported from 200 0 other parts of the state and na on. Slope workers 1978 1981 1984 1987 1990 1993 1996 1999 2002 2005 2008 2011 2014 2017 stay in camps that house thousands of people and work shi s such as one or two weeks on before fl ying Source: Alaska Department of Revenue home to other parts of Alaska or the Lower 48. Wages are a major a rac on Although data for residency and work loca on aren’t available for just the oil patch, looking at numbers for Although Alaska has the most seasonal workforce in the en re North Slope Borough s ll show how unusu- the na on and no other place in the U.S. has winters al the oil patch workforce is. Statewide, 67 percent of as harsh as the North Slope, oil-related work there workers live and work in the same area. In the North carries on year-round and is much less seasonal than Slope Borough, less than one in fi ve workers also live in other parts of the state. There’s less construc on there. Nearly half make the long commute from other and maintenance in the winter, but ice road build- parts of Alaska and 35 percent commute from out of ing and explora on con nue through the winter and state. (See Exhibit 6.) early spring.

Alaska’s Oil Job Losses Taper A er Hard Fall 5 C , 2015-17

1,105 929912 928

500 349 92 Jul Jul Jun Jun Jun Oct Oct Apr Apr Apr Sep Sep Feb Sep Feb Feb Dec Dec Nov Nov Mar Mar Mar July May May May Aug Aug Aug -82 -241 -260 Jan 2016-17 Jan 2015-16 Jan 2014-15 -623 -771 -737 -778 -883 -942 -1,201 -1,219 -1,446 -1,587 -1,540

-2,071 -2,148 -2,253 -2,318 -2,298 -2,283 -2,379 -2,412 -2,543 -2,618 -2,578 -2,678

Note: Employment in this exhibit is for the oil and gas industry only, and covers the industry statewide. Source: Alaska Department of Labor and Workforce Development, Research and Analysis SecƟ on

ALASKA ECONOMIC TRENDS FEBRUARY 2018 7 Mostly Imported Workforce Slope Wages Top The List 6 N , 2016 7 A , 2016 Percent of 2016 state avg Locals 18% Alaska $53,160 100% Nonresidents North Slope oil patch $107,361 202% 35% North Slope Borough $96,324 181% Northwest Arctic Borough $64,464 121% Live Southeast Fairbanks Census Area $64,332 121% elsewhere Aleutians West Census Area $55,896 105% in Alaska Anchorage, Municipality $55,668 105% 47% Valdez-Cordova Census Area $53,160 100% Bristol Bay Borough $51,624 97% Juneau, City and Borough $51,012 96% Aleutians East Borough $50,772 96% Source: Alaska Department of Labor and Workforce Development, Fairbanks North Star Borough $50,508 95% Research and Analysis SecƟ on Nome Census Area $47,376 89% Kenai Peninsula Borough $46,908 88% Ketchikan Gateway Borough $45,264 85% Tough and remote condi ons and the need for a Denali Borough $44,268 83% highly skilled workforce mean high wages. In 2016, Sitka, City and Borough $43,392 82% Kodiak Island Borough $42,480 80% average earnings in the oil patch were more than Dillingham Census Area $42,456 80% double the statewide average of $53,000. (See Exhib- Matanuska-Susitna Borough $41,808 79% it 7.) The Slope’s average is driven up even further by Petersburg Census Area $41,040 77% the long hours — usually 60 to 70 or more per week. Skagway, Municipality $40,680 77% Bethel Census Area $40,452 76% Earnings are highest for those employed by the oil Yakutat, City and Borough $39,540 74% Prince of Wales-Hyder Census Area $39,168 74% producers, at $192,283 on average in 2016. For those Yukon-Koyukuk Census Area $38,748 73% in construc on and transporta on, wages averaged Wrangell, City and Borough $37,776 71% $118,773 and $106,858, respec vely — over 50 per- Lake and Peninsula Borough $36,696 69% cent higher than their statewide industry averages. Hoonah-Angoon Census Area $35,832 67% Haines Borough $35,748 67% Wages for support work were considerably lower but Kusilvak Census Area $26,100 49% s ll well above their industry averages statewide. For Note: The Slope’s average wage includes all industries. example, leisure and hospitality jobs in the oil patch Source: Alaska Department of Labor, Research and Analysis averaged $56,251, but just $23,316 Alaska-wide. SecƟ on

Nonresident share grows people to commute such long distances.

As noted, these high wages a ract workers from all The numbers of residents and nonresidents in the oil over the country. Over the past decade, the share of industry both declined in 2016, however. Resident oil industry workers who aren’t Alaska residents has employment fell by 18 percent and nonresident em- grown, ranging from 28 percent nonresident in 2009 ployment by 14 percent. to 37 percent in 2016. While no breakout exists for the Slope, which represents two-thirds of Alaska’s oil industry, its nonresident percentage is likely even Neal Fried is an economist in Anchorage. Reach him at (907) 269- higher because of the work schedules that allow 4861 or [email protected].

8 FEBRUARY 2018 ALASKA ECONOMIC TRENDS BBestest EstimatesEstimates ShowShow OOngoingngoing JobJob LossLoss

Clearing up some poten ally confusing data

By DAN ROBINSON Two Data Sets Tell Confl ic ng Jobs Story C laska has been shedding jobs 1 for a li le more than two 4,000 years, and there’s a lot of in- ^ƐƟŵĂƚĞƐ Preliminary CES A 2,000 ĞƐƟŵĂƚĞƐƐƵŐŐĞƐƚ terest in when the numbers will turn job recovery posi ve again. We’ll come back to 0 that, but fi rst it’s necessary to ex- -2,000 plain how a familiar set of numbers -4,000 ... But more reliable on our Web site may have tripped ůĂƐŬĂĞƐƟŵĂƚĞƐ data show job loss up people hungry for signs of a re- -6,000 ŚĂƐĐŽŶƟŶƵĞĚ covery. (See Exhibit 1.) -8,000 -10,000 Jul Jun Oct Apr Sep Feb Dec Dec ƵŐ Nov Nov Mar Es mates eventually May Jan 17 turn into counts Oct 16 Sources: Current Employment StaƟ sƟ cs; and Alaska Department of Labor and Workforce The most recent job numbers re- Development, Research and Analysis SecƟ on ported by us or anyone else are es mates, usually based on a survey of a small but sta s cally signifi cant percentage of em- ployers. the numbers are reliable and much more accurate than the job es mates. The quarterly numbers have Eventually these es mates become closer to actual a roughly six-month lag, but once they become avail- counts, thanks to the quarterly reports that nearly able, the original es mates’ usefulness expires. all employers are required to fi le under state unem- ployment insurance laws. These reports include the Specifi c to our current meframe, the more com- number of people who worked each month and the plete data are available and published through the amount of money they were paid over the quarter. second quarter of 2017, and third quarter data are nearly complete and provide solid informa on about That repor ng isn’t perfect — some employers make jobs through September. From that data, we know mistakes or fail to report — but because the report- with a high degree of certainty that Alaska con nued ing is mandatory and there are legal consequences to lose jobs through at least September 2017. (See for failing to report or for deliberately misrepor ng, exhibits 1 and 4.)

ALASKA ECONOMIC TRENDS FEBRUARY 2018 9 Revisions to CES Es mates Have Been Large 2 C , 2012 2017 Total jobs Preliminary CES esƟmate Final revised esƟmate 375,000

365,000

355,000

345,000

335,000

325,000

315,000

305,000 July July July July July Aug Nov Nov Nov Nov Mar Mar Mar Mar Mar May May May May May Sept Sept Sept Sept Jan 16 Jan Jan 15 Jan Jan 14 Jan Jan 13 Jan Jan 12 Jan

Source: Alaska Department of Labor and Workforce Development, Research and Analysis SecƟ on

Two diff erent federal-state Why the U.S. Bureau of Labor Statistics made the change sta s cal programs States work with the U.S. Bureau of Labor Sta s cs on Although the loss of state control over the CES esti- a handful of programs that produce key labor market mates and the more mechanized estimation process informa on: jobs, wages, wage rates, and unemploy- produces less reliable data for Alaska’s uses, the ment rates. These programs have names and acro- change benefi tted the program at the national level. nyms that only the highest-end users need to know or One concern that precipitated the change was that dur- care about. If government sta s cal agencies do their ing big shifts in economic trends — the beginning of a job well, users shouldn’t need specialized knowledge recession, for example — the national CES estimates of processes or acronyms to answer important eco- captured the turning point but states as a group weren’t nomic ques ons such as whether the state is adding able to identify the shifts as quickly. or losing jobs.

State-level use of the estimates doesn’t always match Explaining the accuracy of recent job es mates is an national-level use, either. In Alaska, being able to pro- excep on to the rule about not burdening users with vide reliable over-the-year job growth information is im- behind-the-scenes details, and that requires looking portant, but seasonally adjusted monthly job numbers fi rst at two of these federal-state programs: the Cur- get little use. rent Employment Sta s cs program and the Quarterly Census of Employment and Wages. For national-level analysis, it’s useful for all 50 states’ estimates to be comparable in the way they’re pro- The easier of the two to explain is the Quarterly Cen- duced and to be of similar reliability. BLS determined sus of Employment and Wages, which accesses the that the increased month-to-month volatility at the state quarterly informa on employers fi le under unemploy- level was an acceptable price to pay for that. ment insurance laws discussed above and converts it to employment and wage data broken out by indus- try and geography down to the county level, which equates to boroughs and census areas in Alaska. It’s because of the QCEW program, for example, that we know there were 100 construc on jobs in Bethel in June of 2017 and that gas sta ons in the Kenai Penin-

10 FEBRUARY 2018 ALASKA ECONOMIC TRENDS sula Borough paid about $772,000 in wages in the second quarter of CES Es mates Paint Muddy Picture 2017. 3 C , 2012 2017 The other program, Current Em- Year-ago change Final year-ago change ployment Sta s cs, is designed to at Ɵme of prelim esƟmate aŌer revisions do what the fi rst word in its name 10,000 suggests: es mate the current 8,000 number of jobs in an economy. The Bureau of Labor Sta s cs has 6,000 produced na onal employment es- 4,000 mates since 1915 and has worked 2,000 with state agencies to produce es- mates for all 50 states since 1949. 0 -2,000 How accurate are -4,000 the CES es mates? -6,000 -8,000

For the last several years, the CES -10,000 es mates have become misleading enough that we’ve stopped talking -12,000 July July July July July Nov Nov Nov Nov Mar Mar Mar Mar Mar May May May May May Sept Sept Sept Sept about them in our monthly eco- Sept Jan 16 Jan Jan 15 Jan Jan 14 Jan Jan 13 Jan nomic press release or in Alaska 12 Jan Economic Trends. Source: Alaska Department of Labor and Workforce Development, Research and Analysis SecƟ on Alaska is one of the smallest states in the country and the most sea- sonal, both of which make producing reliable sample- Another strength when state economists had control based es mates more diffi cult. over the es mates was that they showed li le direc- onal bias. In other words, they weren’t consistently Another complica ng factor is that since 2011, states too high or too low. Summing the diff erence between have had less control over their es mates. Un l then, the seven years of monthly preliminary es mates and states had wide la tude to adjust them when state the fi nal revised data shows the es mates were on av- economists felt it was warranted. erage 400 jobs low per month, meaning state analysts showed a small bias on the low side over the extended Using that approach, the average diff erence between period. Alaska’s preliminary es mate and the fi nal revised number was 1,900 jobs over the 2004-2010 period. Knowing that the methods BLS implemented in 2011 That meant the es mates were revised by well under 1 were more mechanical and done primarily by na onal percent on average. technicians with substan ally less local knowledge,

CES numbers go through revisions, are reliable as a historical series

Although the Current Employment Statistics preliminary es- QCEW data through the third quarter of 2017 will guide timates are problematic, they become reliable as a historical those revisions, although subsequent months — from Octo- series once the estimates go through their fi rst major revision, ber 2017 forward — will also be revised in a process called which relies heavily on Quarterly Census of Employment and “re-estimation.” Those numbers are likely to be more accu- Wages data. (See the article for an explanation of QCEW.) rate than the original estimates, but could still be volatile.

For states, that revision occurs in the fi rst few months of the To make all this clear, our Web site will switch from the CES year and covers the period through the third quarter of the numbers to our alternate employment estimates for October previous year. So, for example, the next round of CES an- 2017 onward and make it clear that the numbers from Octo- nual revisions is in progress now and revised numbers will ber on are produced by Alaska analysts rather than the CES be published in March. program. http://live.laborstats.alaska.gov/ces/index.cfm

ALASKA ECONOMIC TRENDS FEBRUARY 2018 11 Job Losses That Began in 2015 Con nue 4 A, , 2015 2017 6,000

4,000

2,000 0

-2,000

-4,000 -6,000

-8,000 -10,000 Jul Jul Jul Jun Jun Jun Oct Oct Oct Apr Apr Apr Sep Feb Feb Sep Feb Sep Dec Dec Dec Aug Aug Aug Nov Nov Nov Mar Mar Mar May May May Jan 16 Jan 17 Jan 15

Source: Alaska Department of Labor and Workforce Development, Research and Analysis SecƟ on

we alerted users at the me that the es mates would become more vola le and advised cau on about read- Alternate es mates based ing too much into the monthly swings or apparent new on QCEW projec ons trends. A er fi rst con nuing to publish the CES job es mates The average revision in Alaska’s job numbers since BLS in our monthly press release with a warning about took control of the es mates has been 3,400 jobs, and their reduced reliability, we decided they were doing the bias has been dis nctly on the low side, with the more harm than good and instead included only the summed diff erence between the es mates and the unemployment rate as the key monthly labor market fi nal revised data being low by an average of 1,500 measure in the press release. jobs a month. But giving the public some idea of what’s happening Even more problema c were the longer stretches with the state’s job count — one of the best measures when CES es mates were especially high or low — if of broad economic health — is important enough that taken at face value, they erroneously show turning since July 2016, we’ve generated alternate employ- points in Alaska’s economy. ment es mates based on projec ons of the reliable though less current QCEW data and included them in From May through December of 2013, for example, our monthly press release. the es mates showed Alaska down an average of 2,300 monthly jobs from their year-ago levels, enough We revise our QCEW-based es mates as soon as a new of a decline that if accurate would have signaled quarter of QCEW data becomes available, so we’re Alaska was entering a recession. The revised numbers always discussing employment es mates and revised showed, as state economists expected they would, data in which we have confi dence. that Alaska consistently added a modest number of jobs over that period. To date, we’ve con nued to publish the CES es mates on our Web site with a warning that the es mates “are Overall, the es mates have tracked with Alaska’s likely to see especially large revisions” and a link to our seasonal pa ern, but they’ve substan ally underes - monthly economic press release for a more accurate mated summer job counts in 2012-2014 and substan- es mate of overall state employment. ally overes mated summer job counts in 2016. (See Exhibit 2.) What the es mates said about over-the year That brings us back to how someone could get the losses or gains painted a muddled picture of the 2012- wrong impression about the direc on of the state’s 2016 economy, a period during which the revised data economy. CES numbers taken from our Web site show showed a clear growth trend that shi ed to a clear re- the state’s job count going from well below year-ago cessionary trend of job loss. (See Exhibit 3.) levels in May 2017 to suddenly more than 2,000 above

12 FEBRUARY 2018 ALASKA ECONOMIC TRENDS year-ago levels in June, as shown by Exhibit 1 at the be- ginning of this ar cle. The numbers then remain above year-ago levels through October before dipping again during the last two months of the year. As noted, based on published QCEW data through the second quarter of 2017 and nearly complete data through third quarter, it’s almost certain that job loss- es con nued through at least September. Beyond that, there are more ques on marks — but historical pat- terns strongly suggest Alaska con nued to lose a mod- erate number of jobs through the end of 2017. Exhibit 4 shows what we believe are the most accurate Alaska employment numbers through December. Making online jobs data more clear Beginning this month, we will replace the preliminary CES es mates on our Web site with our alternate QCEW-forecast based es mates and a note making that clear. Alaska’s CES es mates will s ll be available on the BLS site, and we’ll publish a link to that data on our site.

Dan Robinson is the chief of Research and Analysis in Juneau. Reach him at (907) 465-6040 or [email protected].

Regional estimates also available

This article focuses on statewide job numbers, but we also produce regional employment estimates, and the Bureau of Labor Statistics works with states to produce CES estimates for “metropolitan statistical areas.” In Alaska, those are the Anchorage/Mat-Su Region and Fairbanks. (Note BLS publishes the prior as “Anchor- age MSA.”)

State analysts produce job estimates for other parts of the state, including Anchorage and the Southeast, Gulf Coast, Northern, Southwest, and Interior regions. We publish those estimates on our Web site (the Anchor- age/Mat-Su and Fairbanks data are also on the BLS site) and revise them annually using Quarterly Census of Employment and Wages data.

http://live.laborstats.alaska.gov/ces/

ALASKA ECONOMIC TRENDS FEBRUARY 2018 13 Gauging Alaska’s Economy

14 FEBRUARY 2018 ALASKA ECONOMIC TRENDS Gauging Alaska’s Economy

Four-week moving average ending with the specifi ed week

ALASKA ECONOMIC TRENDS FEBRUARY 2018 15 Employment by Region Percent change in jobs - 5.6% December 2016 North Slope to December 2017 Northern Region Northwest Arctic

Interior - 0.9% Nome Region -1.0% Fairbanks Yukon-Koyukuk Southeast Statewide Denali Fairbanks

Matanuska- Kusilvak Susitna Anchorage/Mat-Su Skagway Region Valdez-Cordova Haines Bethel Yakutat Juneau Kenai Dillingham Peninsula Petersburg Southwest Anchorage Hoonah- Wrangell Region Gulf Coast Region Sitka Bristol Bay -1.2% Southeast Kodiak Island Region Lake & Anchorage/ -1.9% Peninsula Mat-Su Prince of Wales- +1.1% Hyder Ketchikan Aleutians West Aleutians -1.2% East

Unemployment Rates Seasonally adjusted Not seasonally adjusted

Prelim. Revised Prelim. Revised 12/17 11/17 12/16 12/17 11/17 12/16 United States 4.1 4.1 4.7 United States 3.9 3.9 4.5 Alaska 7.3 7.2 6.6 Alaska 7.3 7.1 6.6

Regional, not seasonally adjusted

Prelim. Revised Prelim. Revised Prelim. Revised 12/17 11/17 12/16 12/17 11/17 12/16 12/17 11/17 12/16 Interior Region 7.5 7.3 6.6 Southwest Region 12.3 11.4 11.2 Southeast Region 7.2 6.9 6.4 Denali Borough 20.6 20.2 18.2 Aleu ans East Borough 5.9 4.3 4.7 Haines Borough 13.4 12.7 11.1 Fairbanks N Star Borough 6.6 6.3 5.7 Aleu ans West 5.4 4.4 4.6 Hoonah-Angoon 18.7 16.6 14.0 Southeast Fairbanks 10.8 10.9 9.8 Census Area Census Area Census Area Bethel Census Area 13.2 13.0 12.2 Juneau, City and Borough 5.1 4.9 4.5 Yukon-Koyukuk 18.0 17.6 16.8 Bristol Bay Borough 14.3 11.4 13.3 Ketchikan Gateway 7.0 7.0 6.4 Census Area Dillingham Census Area 11.1 10.4 11.4 Borough Kusilvak Census Area 19.8 19.4 18.0 Petersburg Borough 10.1 9.2 9.8 Northern Region 10.8 11.2 10.4 Lake and Peninsula 15.4 15.1 12.7 Prince of Wales-Hyder 12.3 11.7 11.6 Nome Census Area 12.2 12.0 11.4 Borough Census Area North Slope Borough 6.3 7.0 5.9 Sitka, City and Borough 5.0 4.7 4.2 Gulf Coast Region 8.8 8.2 8.0 Northwest Arc c Borough 14.5 15.6 15.5 Skagway, Municipality 20.0 21.2 20.6 Kenai Peninsula Borough 8.9 8.7 8.0 Anchorage/Mat-Su Region 6.4 6.3 5.7 Wrangell, City and Borough 8.8 8.6 7.7 Kodiak Island Borough 7.3 4.8 6.7 Anchorage, Municipality 5.7 5.7 5.1 Yakutat, City and Borough 11.0 11.4 9.0 Valdez-Cordova 9.9 9.5 9.3 Mat-Su Borough 8.7 8.3 8.0 Census Area

16 FEBRUARY 2018 ALASKA ECONOMIC TRENDS How Alaska Ranks

Private Sector Unemployment Rate1 Job Growth2 Job Growth2 1st 1st 50th 1st Hawaii 50th Nevada Oregon 50th 2.0% 7.3% 3.3% -1.0% 3.3% -1.3%

State Government Retail Trade Job Growth2 Job Growth2

1st 50th 1st 39th 50th Tennessee 40th Wisconsin Nevada Louisiana 4.9% -1.3% -3.6% 3.0% -1.6% -3.8%

Alaska tied for 39th with Alabama

1December seasonally adjusted unemployment rates 2December employment, over-the-year percent change. Alaska numbers are sourced only from Alaska Department of Labor and Workforce Development, Research and Analysis Sec on.

Sources are U.S. Bureau of Labor StaƟ sƟ cs and Alaska Department of Labor and Workforce Development, Research and Analysis SecƟ on, unless otherwise noted. Other Economic Indicators Current Year ago Change

Anchorage Consumer Price Index (CPI-U, base yr 1982=100) 218.616 1st half 2017 216.999 +0.75%

Commodity prices Crude oil, Alaska North Slope,* per barrel 63.79 Dec 2017 53.26 +19.77% Natural gas, residential, per thousand cubic ft 13.50 Oct 2017 13.74 -1.75% Gold, per oz. COMEX 1,333.40 1/19/2018 1,204.90 +10.66% Silver, per oz. COMEX 17.03 1/19/2018 17.00 +0.16% Copper, per lb. COMEX 318.60 1/19/2018 261.05 +22.05% Zinc, per MT 3,376.00 1/18/2018 2,767.00 +22.01% Lead, per lb. 1.17 1/18/2018 0.91 +28.57%

Bankruptcies 97 Q3 2017 106 -8.5% Business 7 Q3 2017 5 +40.0% Personal 90 Q3 2017 101 -10.9%

Unemployment insurance claims Initial fi lings 6,008 Dec 2017 6,955 -13.62% Continued fi lings 56,686 Dec 2017 59,106 -4.09% Claimant count 15,519 Dec 2017 16,416 -5.46% *Department of Revenue es mate Sources for pages 18 through 21 include Alaska Department of Labor and Workforce Development, Research and Analysis SecƟ on; U.S. Bureau of Labor StaƟ sƟ cs; U.S. Bureau of Economic Analysis; U.S. Census Bureau; COMEX; Bloomberg; Infomine; Alaska Department of Revenue; and U.S. Courts, 9th Circuit

ALASKA ECONOMIC TRENDS FEBRUARY 2018 17

This page intentionally left blank. Safety Minute

What the law says about toxic substances at work

It is your right under the law to know about toxic and other • Employers must keep Safety Data Sheets on fi le for hazardous substances or physical agents in your work- each toxin or hazard and make them available during place. AS 18.60.068 requires employers to display this in- the work shift. If employers can’t provide workers with formation in a prominent place. For a free printable copy of an SDS within 15 calendar days of a request, they the Safety and Health Protection on the Job poster, please must remove employees from possible exposure. visit: http://labor.alaska.gov/lss/forms/right-to-know.pdf.

The law also requires the following: The Alaska Department of Labor and Workforce Develop- ment will help employers develop SDS programs, review • Employers must inform employees about the locations programs on-site, and conduct safety seminars. For more and nature of operations that could result in exposure information, contact the department’s Labor Standards and to these toxins. Safety Division, Occupational Safety and Health Section at: http://labor.alaska.gov/lss/oshhome.htm. • Employers must educate employees about the health effects of exposure and the purpose, proper use, and Safety Minute is wri en by the Labor Standards and Safety Division of the limitations of personal protective equipment. Alaska Department of Labor and Workforce Development.

ALASKA ECONOMIC TRENDS FEBRUARY 2018 19