NOTICE: SLIP OPINION
(not the court’s final written decision)
The opinion that begins on the next page is a slip opinion. Slip opinions are the written opinions that are originally filed by the court. A slip opinion is not necessarily the court’s final written decision. Slip opinions can be changed by subsequent court orders. For example, a court may issue an order making substantive changes to a slip opinion or publishing for precedential purposes a previously “unpublished” opinion. Additionally, nonsubstantive edits (for style, grammar, citation, format, punctuation, etc.) are made before the opinions that have precedential value are published in the official reports of court decisions: the Washington Reports 2d and the Washington Appellate Reports. An opinion in the official reports replaces the slip opinion as the official opinion of the court. The slip opinion that begins on the next page is for a published opinion, and it has since been revised for publication in the printed official reports. The official text of the court’s opinion is found in the advance sheets and the bound volumes of the official reports. Also, an electronic version (intended to mirror the language found in the official reports) of the revised opinion can be found, free of charge, at this website: https://www.lexisnexis.com/clients/wareports. For more information about precedential (published) opinions, nonprecedential (unpublished) opinions, slip opinions, and the official reports, see https://www.courts.wa.gov/opinions and the information that is linked there.
Filed
Washington State
Court of Appeals
Division Two
March 28, 2023
IN THE COURT OF APPEALS OF THE STATE OF WASHINGTON
DIVISION II
LANDIS+GYR MIDWEST, INC., No. 56877-2-II
Appellant,
v.
STATE OF WASHINGTON, DEPARTMENT PUBLISHED OPINION OF REVENUE,
Respondent.
GLASGOW, C.J.— Puget Sound Energy (PSE) hired Landis+GYR Midwest, Inc. (Landis)
to collect data from PSE’s electric and natural gas meters located at each customer’s property, convert the data into a usable form, and transmit the usable data to PSE. In 2014, the Department of Revenue began auditing Landis. In 2019, the Department told Landis that the services Landis provided PSE were digital automated services subject to retail sales tax and no exemption applied. Landis paid the tax and then sued for a refund.
Landis moved for summary judgment, arguing the services it provided to PSE were data processing services, which are exempt from the retail sales tax imposed on digital automated services. The Department filed a cross motion for summary judgment, arguing Landis’s service was not a data processing service because its primary purpose was collecting and transmitting PSE’s data. The superior court denied Landis’s motion and granted the Department’s, dismissing Landis’s lawsuit.
No. 56877-2-II
Landis appeals. It argues that the trial court erred by denying its summary judgment motion and granting the Department’s. Landis contends that because the primary purpose of its service to PSE was the extraction and conversion of certain meter data into a readable and usable format, not merely the transmission of data, it provided a data processing service that is exempt from retail sales tax.
We agree with Landis and reverse. Landis’s service meets the statutory definition of a data processing service and is thus exempt from the retail sales tax. We remand for entry of summary judgment in Landis’s favor.
FACTS
PSE hired CellNet Data Services in 1997 to perform automated readings of PSE’s electric and natural gas meters. In 2004, the Department informed CellNet that the state of Washington did not collect retail sales tax on meter reading services. A. Adoption of Tax Code for Digital Services In 2009, the legislature expanded the list of transactions subject to retail sales tax to include the sale of digital goods, codes, and digital automated services. LAWS OF 2009, ch. 535, § 301(8)(a); see also RCW 82.04.050(8)(a), .250(1). A digital automated service is “any service transferred electronically that uses one or more software applications.” RCW 82.04.192(3)(a). The definition of “digital automated service” has 16 exceptions that are digital automated services but are not subject to retail sales tax. RCW 82.04.192(3)(b)(i)-(xvi). The legislature added data processing services to the list of exceptions in the statute in 2010. LAWS OF 2010, ch. 111, § 203(3)(b)(xv). The exception was retroactive to July 2009. Id.§ 902(1).
3
No. 56877-2-II
A data processing service is a “primarily automated service . . . where the primary object of the service is the systematic performance of operations by the service provider on data supplied in whole or in part by the customer to extract the required information in an appropriate form or to convert the data to usable information.” RCW 82.04.192(3)(b)(xv) (emphasis added). The definitional statute lists examples of data processing services, such as “check processing, image processing, form processing, survey processing, payroll processing, claim processing, and similar activities.” RCW 82.04.192(3)(b)(xv). The original draft of the bill that became the statute more broadly defined a “data processing service” as a primarily automated service “that involves the systematic performance of operations on data” to extract or convert usable information. H.B. 2620, at 15, 61st Leg., Reg. Sess. (Wash. 2010) (emphasis added). A bill report explained the change, stating that narrowing the definition “clarifies that data processing services are services where the primary object . . . relates to data processing, but some incidental services may be involved.” H.B. REP. ON SUBSTITUTE H.B. 2620, at 3, 61st Leg. Reg. Sess. (Wash. 2010). B. Landis’s Service Shortly after the change in the law, Landis acquired CellNet. In 2011, Landis entered a contract with PSE to continue the automated meter reading services.
The agreement covered approximately 1.2 million electric meters and 800,000 natural gas meters. The contract stated that Landis “shall enter and store” data from the meters in a database and “shall provide Good Billing Reads” to PSE. Clerk’s Papers (CP) at 414. A “good billing read” was defined as “a complete meter read that contain[ed] all requested energy and demand information . . . delivered to PSE from a Supported Meter” within a time window that allowed PSE to efficiently bill its customers. CP at 406. The data Landis had to provide for electric meters
4
No. 56877-2-II
included the meter number, information about the cumulative kilowatts per hour used, the date and time of that information, and information on whether there had been a power outage or the meter showed signs of tampering. For natural gas meters, Landis had to provide the meter number, the cumulative cubic feet used, the data and time of that information, and information about possible tampering. PSE could also pay for more detailed information, such as when power was used throughout the day.
The individual meters automatically generated data as customers consumed electricity and natural gas. Landis owned radio modules that were attached to the meters, which broadcasted that raw data to collection devices. The radio modules on the electric meters broadcasted once every five minutes and the ones on the natural gas meters broadcasted once every 15 minutes.
At midnight each day, the collection devices sent all of the individual meters’ data for that day to Landis’s data center in Kansas. There, “in a process that [took] several hours, the data [was] automatically converted into a format that [could] be read by PSE’s computer systems and compiled into a database.” CP at 327. The next morning, Landis sent PSE the converted data in a form that PSE could use to calculate the amount of power consumed by a given customer, track outages, and identify malfunctioning meters or tampering. If Landis’s system could not provide the information automatically, Landis had to send someone to manually read the meters. The network could not transmit any other kind of data.
PSE paid Landis based on the number of “good billing reads” Landis provided. Landis did not charge or pay retail sales tax for its service.
5
No. 56877-2-II
C. Audit
In 2014, the Department began auditing Landis for the time period beginning in January 2010. In 2019, the auditor informed Landis that they had forgotten to include the tax due on Landis’s meter reading services. The auditor explained that because of the 2009 change in the law, Landis’s services were “no longer exempt” because they “qualif[ied] as a digital automated service . . . subject to retail sales tax.” CP at 166. The audit result stated that Landis owed approximately $17.7 million in unpaid retail sales tax and about $880,000 in unpaid retailing business and occupation tax for the audit period.
Landis objected to several items in the audit, including the application of the retail sales tax to its meter reading services.1 Landis argued its meter reading service was a data processing service exempt from the tax because the primary object of the service was “to convert the gas and electricity usage data (which [was] extracted via automation from data provided by PSE at the meter) to information in an appropriate form” that PSE used to bill its customers. CP at 217. Landis compared its service to payroll processing, where a service provider receives hours, pay rates, and other information “from its customers and converts this data to useable information each pay cycle” so its customers pay their employees correctly. Id.
In March 2020, the Department produced a report explaining its belief that Landis’s service was not a data processing service exempt from retail sales tax. The Department reasoned that Landis’s service included data collection, transmission, processing, storage, real time meter information services, and service interruption monitoring, so “the full service contracted for” went too far beyond data processing to qualify for the retail sales exemption. CP at 261.
1 Landis also challenged additional audit findings, but they are not at issue in this appeal.
6
No. 56877-2-II
In May 2020, Landis sought review of the audit through the Department’s internal process.
In September 2020, the Department issued a postaudit report explaining that it believed Landis’s service to PSE primarily involved the collection and transmission of data from the meters to PSE, not the conversion of that data into a useable format. Therefore, the Department maintained that Landis’s service was not a data processing service and was subject to the retail sales tax.
Before the Department’s internal process finished, Landis paid the assessed back tax.
Landis then timely sued for a refund of the tax payment on the grounds that Landis’s service was primarily a data processing service and the Department had improperly found the service was subject to retail sales tax.2 D. Superior Court Proceedings Both Landis and the Department moved for summary judgment on the issue of whether Landis’s meter reading service qualified as a data processing service exempt from retail sales tax. Both parties argued that there were no genuine issues of material fact. Landis conceded that the service met the broad requirements of an automated digital service that would be subject to retail sales tax if it did not meet the claimed exception.
Landis submitted a declaration from a technician who helped provide the meter reading service to PSE. The technician explained how the meter reading system worked and explained the purpose of the contract:
The primary object of the services provided by [Landis] to PSE [was] to extract meter data from PSE-owned meters and convert that data to a form usable by PSE.
There would be no reason for PSE to contract with [Landis] unless [Landis] could provide meter data extracted from [PSE’s] meters and returned in a format that
2 Landis also asserted causes of action for the improper taxation of temporary worker services and a violation of the Washington Taxpayers’ Rights and Responsibilities, chapter 82.32A RCW. Landis does not raise either of these arguments on appeal.
7
No. 56877-2-II
[could] be used in PSE’s customer information systems. The timely and accurate delivery of readable meter data to PSE allow[ed] PSE to download the information into its billing system so that its customers [could] be billed timely and accurately.
CP at 327. The technician also described some ancillary services Landis provided, such as outage detection services and readings on request.
Landis submitted two declarations from the person who managed PSE’s meter data management system and the receipt of files from Landis. The manager stated, “PSE’s primary purpose, in fact only purpose, for entering into the contracts with [Landis], was to obtain meter data extracted from PSE gas and electric meters that [was] processed into a format that [could] be used by PSE.” CP at 675. The manager also explained that “[w]ithout the conversion of the meter data performed at the [Landis] data center, the data would have been unreadable and unintelligible by PSE’s computer systems.” CP at 676. Landis also submitted examples of the raw data pulled from the meters and contrasted it with the usable format the data was converted into before it was loaded into the database.
The manager explained that for PSE to pay Landis for providing a “good billing read,”
Landis “was required to successfully extract PSE meter data from a PSE meter, transmit it to [Landis]’s data center, convert the meter data into a form usable by PSE, and deliver a complete set of meter data for each supported meter back to PSE.” CP at 1217. “In short, a ‘Good Billing Read’ as defined in the 2011 [contract] was synonymous with the successful processing of meter data from a PSE meter in accordance [with] the terms of that contract.” CP at 1218.
In a deposition, the manager agreed that, because of the arrangement with Landis, PSE did not have to pay employees to manually read its meters. The automated service resulted in “more
8
No. 56877-2-II
accurate information that . . . was processed properly and sent to PSE, and limited issues that you run into in the field, like access issues.” CP at 930.
Neither party disputed any of the facts as described above. Their disagreement centered on how the law should be applied to those facts. The parties applied the “primary purpose” test from Qualcomm, Inc. v. Dep’t of Revenue, which evaluates the purpose of the service from the perspective of the purchaser or customer. 171 Wn.2d 125, 127-28, 249 P.3d 167 (2011). Landis argued that the object “of the service from PSE’s perspective” was to obtain extracted meter data processed into a form that could be used in PSE’s customer billing system. CP at 694. The Department argued that the primary purpose of Landis’s service was to collect and transmit data from individual meters, not process it.
After a summary judgment hearing, the trial court ruled that the primary purpose of the transaction from PSE’s perspective was “to obtain good billing reads,” and while there was data processing involved, “that [was] not the primary purpose or object of the service.” Verbatim Rep. of Proc. (VRP) at 29. The trial court reasoned that Landis “provides automated meter reading services . . . to allow [PSE] to access consumption data from [its] meters. This information is then electronically transmitted to a database managed by and owned by [Landis] and then accessed remotely by [PSE].” Id. at 27. The trial court also stated that the contract between Landis and PSE said that its “primary purpose was to allow PSE to ‘obtain good billing reads’” and that Landis did not invoice PSE for data processing. Id. at 28. The trial court granted the Department’s motion for summary judgment, denied Landis’s motion for summary judgment, and dismissed Landis’s lawsuit. Landis appeals.
9
No. 56877-2-II
ANALYSIS
On appeal, the parties identify no disputed issue of fact and they each argue they should prevail as a matter of law. The Department reasons that the primary purpose of Landis’s service is collecting and transmitting data, “which is an information service, not data processing.” Br. of Resp’t at 31. We disagree.
We review grants of summary judgment de novo. FPR II, LLC v. Dep’t of Revenue, 16 Wn. App. 2d 706, 713, 482 P.3d 320 (2021). The interpretation of a statute is a question of law that we also review de novo. Qualcomm, 171 Wn.2d at 131. Summary judgment is proper when the moving party shows “that there is no genuine issue as to any material fact and that the moving party is entitled to a judgment as a matter of law.” CR 56(c). If there are no genuine issues of material fact and we determine that a party who did not prevail below is entitled to summary judgment, we can remand to order the entry of judgment in that party’s favor. See Impecoven v. Dep't of Revenue, 120 Wn.2d 357, 365, 841 P.2d 752 (1992).
The Department argues that we must narrowly construe exemptions to tax statutes. It is true that there must be unambiguous legislative intent to make a tax exemption, and we will not create tax exemptions by implication. TracFone Wireless, Inc. v. Dep’t of Revenue, 170 Wn.2d 273, 297, 242 P.3d 810 (2010). And in general, “statutory exceptions are construed narrowly in order to give effect to the legislative intent underlying the general provisions.” Foster v. Dep’t of Ecology, 184 Wn.2d 465, 473, 362 P.3d 959 (2015). However, “[w]e construe ambiguous tax statutes in favor of the taxpayer.” Qualcomm, 171 Wn.2d at 131. A taxpayer claiming to be exempt from a tax has “the burden of establishing the exemption” applies. Gartner, Inc. v. Dep’t of Revenue, 11 Wn. App. 2d 765, 773, 455 P.3d 1179 (2020). See also RCW 82.32.180.
10
No. 56877-2-II
A. Statutes and Regulations A digital automated service is “any service transferred electronically that uses one or more software applications.” RCW 82.04.192(3)(a). The sale of a digital automated service is a retail sale subject to retail sales tax. RCW 82.04.050(8)(a), .250(1).
The legislature has provided several exceptions to the retail sales tax for digital automated services, including one for data processing services. RCW 82.04.192(3)(b)(xv). A data processing service must be “primarily automated,” must be “provided to a business or other organization,” and “the primary object of the service [must be] the systematic performance of operations by the service provider on data supplied in whole or in part by the customer to extract the required information in an appropriate form or to convert the data to usable information.” RCW 82.04.192(3)(b)(xv) (emphasis added). “Data processing services include check processing, image processing, form processing, survey processing, payroll processing, claim processing, and similar activities.” Id.
The Department’s excise tax rules restate the statutory definition of a data processing service and then provide three examples. WAC 458-20-15503(303)(o). The first example describes a product that is a digital automated service but not a data processing service: automated technology that searches a computer and gathers documents relevant to a lawsuit. WAC 458-20- 15503(303)(o) ex. 20. The technology also allows the customer to organize, store, and notate the gathered documents. Id.
The second example describes a product that is a data processing service exempt from the retail sales tax: a check processing service that receives check scans from a bank and returns “extracted and reformatted data,” including the dollar amount, account number, and verification
11
No. 56877-2-II
of signature, for the bank “to reconcile its customer’s accounts.” WAC 458-20-15503(303)(o) ex. 21. A third example explains that the check processing service is still a data processing service even if it receives check images from both the bank and the bank’s customers. WAC 458-20- 15503(303)(o) ex. 22. B. Other Cases Analyzing Taxation of Information Services Both parties in this case apply the “primary purpose” or “true object” test. The Washington Supreme Court has used this test to determine whether “the primary purpose [of a service] was information processing or information transmission” in a case where information processing systems were exempt from a tax applied to network telephone services. Qualcomm, 171 Wn.2d at 140.
Qualcomm sold a system that provided trucking companies with information about vehicle location and operation. Id. at 128. Mobile terminals in the trucks collected location data, for example, and relayed it to Qualcomm’s network management facility, a component that Qualcomm referred to as the “service.” Id. The facility processed the data, making it readable for customers, and stored the data. Id. Trucking companies could access the data via software on their computers. Id. “The service [was] useless without the mobile terminals and software, which in turn [were] useless without the service.” Id. at 128-29.
The Supreme Court made clear that it was analyzing whether “the primary purpose [of Qualcomm’s system] was information processing or information transmission.” Id. at 140 (emphasis added). The statute at issue in Qualcomm, excluded from the definition of telecommunications services subject to retail sales tax, “[d]ata processing and information services that allow data to be generated, acquired, stored, processed, or retrieved and delivered by an
12
No. 56877-2-II
electronic transmission to a purchaser where such purchaser’s primary purpose for the underlying transaction is the processed data or information.” RCW 82.04.065(27)(a). The Supreme Court held that Qualcomm provided an exempt service in part because “[t]he raw data gathered by the . . . system’s hardware would be unintelligible without the calculations performed by Qualcomm.” Qualcomm, 171 Wn.2d at 143-44. This court recently summarized Qualcomm, stating that “the data processing service component of [Qualcomm’s] service was the true object because [Qualcomm’s] customers sought to acquire the specific information, not raw data.” Gartner, 11 Wn. App. 2d at 783.
The Supreme Court has considered the manipulation of transmitted information in other tax cases. For example, when the City of Seattle taxed Comcast’s high-speed cable Internet service as a network telephone service subject to retail sales tax, the Supreme Court disagreed on several grounds. Cmty. Telecable of Seattle, Inc. v. City of Seattle, 164 Wn.2d 35, 44, 186 P.3d 1032 (2008). Importantly, the definition of “network telephone service” expressly excluded the provision of Internet service. Id. at 43. And Comcast was “not engaging in the mere ‘provision of transmission’” of data described in the network telephone service statute. Id. at 44 (quoting RCW 82.04.065(2)). Instead, it was “transform[ing] and manipulat[ing] data” that passed through the network, which was “an integral and necessary part of the provision of Internet services.” Id. (internal quotation marks omitted).
In sum, the Supreme Court has, in other tax cases, drawn a clear distinction between services involving the mere transmission of data and services primarily involving the conversion or manipulation of data.
13
No. 56877-2-II
C. Landis’s Automated Meter Reading Service Landis’s service collected data from PSE’s meters, transmitted the data to a computing center where it was converted into a useable format, and returned the converted data to PSE. Landis was paid based on the amount of converted data, or “good billing reads,” it provided PSE. The parties do not dispute any facts about how the relevant system works, and it is undisputed that PSE’s primary purpose in hiring Landis was to obtain “good billing reads,” but the parties attach different connotations to this term.
The Department maintains that the primary purpose of Landis’s service providing “good billing reads” was electronic meter reading, or in other words, collecting and transmitting information from PSE’s meters to PSE’s computer system. The Department contends that data processing is a necessary but minor component of Landis’s service, not the primary purpose. We disagree and conclude that the primary purpose of Landis’s service is data processing.
The Department is correct that the legislature edited the definition of data processing to clarify that the exemption covered “services where the primary object of the service relates to data processing, but some incidental services may be involved.” H.B. REP. ON SUBSTITUTE H.B. 2620, at 3, 61st Leg. Reg. Sess. (Wash. 2010) (emphasis added). The Department contends that the collection of data from the meters is not incidental. But Landis rightly asserts that the full statutory definition of data processing is disjunctive, not conjunctive: the service must “extract the required information in an appropriate form or convert the data to useable information.” RCW 82.04.192(3)(b)(xv) (emphasis added).
Here, Landis owned all of the components of the network except for the utility meters themselves. Landis collected data from the meters, extracted the relevant data points PSE wanted,
14
No. 56877-2-II
and converted them to an appropriate form (a “good billing read”) in a process that took several hours. Landis and PSE employees repeatedly explained that the raw data in PSE’s meters was functionally useless because it was not compatible with PSE’s computer system. Landis also submitted examples of the raw data from the PSE meters compared with the final product it returned to PSE. At the trial court, the person managing this process stated that PSE’s alternative to hiring a company like Landis was to manually read the meters. But PSE further explains that in the digital era, the company’s only modern alternative would be building its own data processing center that could convert the meter data, as well as a transmission network.
Landis did not merely transmit information from point A to point B, manipulating the data only for transportation purposes. Instead, every day, Landis’s radio modules collected data points from 1.2 million of PSE’s electrical meters and from 800,000 of PSE’s natural gas meters. Landis’s radio modules transmitted those data points to a facility that converted the data points into usable information, like the quantity of electricity or gas used, and identified information about outages or tampering. Landis then sent the resulting information to PSE in a useable form. A system where PSE’s meters generated “good billing reads” each month and Landis’s service merely transmitted those files from the meters to PSE’s headquarters would fall outside the data processing exception. But it is undisputed that is not the system at issue here. All of these facts establish that the primary purpose of Landis’s service was data processing.
Although the Department characterizes Landis’s service as a system “to collect, transmit and electronically transfer” billing information, the service bears more similarity to the check processing examples in the Department’s regulation examples, 21 and 22, which depict data processing. Br. of Resp’t at 40. Like the check processor in the examples, Landis collects data
15
No. 56877-2-II
from PSE and individual customers, extracts the needed portions of that data, processes the data by converting it into a format that PSE can use, and sends it to PSE. This similarity also shows that data processing is the primary object of Landis’s service.
Finally, the Department argues that like the service at issue in Qualcomm, Landis’s systems were “information services” focused on transmitting information for the customer. “Like Qualcomm’s trucking customers, [PSE] is interested in timely and reliable information, not the mode or method by which that information is created.” Br. of Resp’t at 40-41. But the Department’s reliance on Qualcomm is misplaced because this case involves different statutory tax classifications. And although Landis’s services would almost certainly qualify as “information services” as that term is used in Qualcomm, the processing Qualcomm performed on the data it received was a crucial part of what exempted Qualcomm’s service from the definition of “telecommunication services.” Qualcomm, 171 Wn.2d at 144. This court has noted that “the data processing service component of [Qualcomm’s] service was the true object because [Qualcomm’s] customers sought to acquire the specific information, not raw data.” Gartner, 11 Wn. App. 2d at 783. Thus, Qualcomm tends to support Landis’s position.
In sum, undisputed evidence in the record shows that the primary purpose of Landis’s service was extracting certain data from the raw data points produced by PSE’s meters and converting that data into a usable format, meeting the requirements of the data processing exception. PSE’s primary purpose for hiring Landis was to have Landis’s technology and network extract raw data from PSE’s meters and convert it into usable “good billing reads.” This satisfies the definition of a “data processing service” in RCW 82.04.192(3)(b)(xv). The trial court erred by granting the Department’s motion for summary judgment and denying Landis’s.
16
No. 56877-2-II
CONCLUSION
Landis’s service meets the definition of a data processing service under RCW 82.04.192(3)(b)(xv) and is thus exempt from retail sales tax. We reverse. And because there are no issues of material fact, we remand for entry of summary judgment in Landis’s favor.
Glasgow, C.J.
We concur:
Lee, J.
Cruser, A.C.J.