The USPTO Trademark Case Files Dataset: Descriptions, Lessons, and Insights

• Published date: 01 December 2013
• Date: 01 December 2013
• DOI: http://doi.org/10.1111/jems.12035

The USPTO Trademark Case Files Dataset:

Descriptions, Lessons, and Insights

STUART J.H. GRAHAM

U.S. Patent & Trademark Office and Scheller College of Business GeorgiaInstitute of Technology

800 West Peachtree Street Atlanta GA 30308

Stuart.Graham@scheller.gatech.edu

GALEN HANCOCK

Department of Economics University of California

530 Evans Hall Berkeley CA 94720

galen@econ.berkeley.edu

ALAN C. MARCO

U.S. Patent & TrademarkOffice

Office of the Chief Economist

600 Dulany Street Alexandria VA 22314

alan.marco@uspto.gov

AMANDA FILA MYERS

U.S. Patent & TrademarkOffice

Office of the Chief Economist

600 Dulany Street Alexandria VA 22314

amanda.fila@uspto.gov

This article describes the “USPTO Trademark Case Files Dataset” of trademark applications

and registrations derived from the U.S. Patent and Trademark Office (USPTO) main database

for administering trademark case files. The dataset provides detailed information on 6.7 million

trademark applications filed with, or registrations issued by, the USPTO between January 1870

and January 2012, including ownership, mark characteristics, classification, prosecution events,

and renewal and maintenance history. This article provides a comprehensive description of the

dataset, including discussions of the legal framework affecting and the administrative processes

generating these data. We provide a “first look”at the information the dataset captures and present

key trends in trademark applications, registrations, and renewals. We highlight data elements

valuable to researchers and the general public, and discuss issues that may arise in using these

data. In describing these data, we aim to encourage new streams of research on trademarks and

what they indicate about their users, the strategies for employing them, and the wider economic

impacts that these data will help uncover.

1. Introduction

Scholars in economics, management, and related fields have rarely used trademark

data to conduct empirical analyses. This paucity of use may be due to its inaccessibil-

ity relative to more commonly used patent information, securities filing records, and

other conveniently available data sources. Trademarkdata represent a new opportunity

for conducting research in areas such as marketing, advertising, brand use, strategy,

innovation, and new product and service introduction. They may also contain valu-

able information on innovative activity not well captured by traditional sources. Unlike

Published 2013. This article is a U.S. Government work and is in the public domain in the USA.

Journal of Economics & Management Strategy, Volume22, Number 4, Winter 2013, 669–705

670 Journal of Economics & Management Strategy

patents, which are limited to technological innovations, trademarks cover a broader set

of participants in the economy because almost every firm, regardless of size, market,

or business strategy, has goodwill to protect. Accordingly, trademark data may capture

innovations that are not patented, either because they are not patentable or because their

inventors have chosen not to seek patent protection.

This article describes a new dataset—the “USPTO TrademarkCase Files Dataset”—

of trademark applications and registrations derived from the USPTO’s main database

of trademark case files, available at: http://www.uspto.gov/ip/officechiefecon/

economicsdata.jsp. This dataset contains detailed information on 6.7 million trademark

applications filed with, or registrations issued by, the USPTO between January 1870

and January 2012, including prosecution history, ownership, mark text, classification,

related marks, and renewal history. Although the USPTO provides trademark case files

to the public in electronically downloadable formats through the Data.gov web site,

these records consist of document images and XML files that are not well suited to

large-scale, comprehensive analysis. The USPTO Office of Chief Economist (OCE) be-

gan a process in 2011 to convert these structured files into regularized data tables to be

more compatible with standard statistical software packages. Our intent in doing so is

to eliminate the necessity of numerous individual researchers engaging in duplicative

programming, cleaning, and converting efforts and, thereby, free up research capacity

for more substantive inquiry into innovation, law, organizations, and management.

In this article, we describe the dataset and provide a “first look” at the large

amount of information it contains. We highlight some key elements in the data likely to

be valuable to researchers and other users. However, thereare numerous potential uses

for the dataset beyond those we discuss in this article. We offer this view in order to

encourage a new stream of research on trademarks and what they may indicate about

their users, the strategies under which they are employed, and the wider economic

impacts that these data are able to help uncover.

The article is structured as follows. Section 2surveys some past literature involving

trademarks and trademark data, noting that such previous research has been limited.

In Section 3, we provide a brief background on trademarks and the Federal trademark

registration process, and describe the lifecycle of a registered trademark, highlighting

the USPTO processes that generate the administrative data from which the dataset is

derived. Section 4describes the organizational structure of the dataset and provides a

first look at the data therein. In Section 5, we offer some concluding observations.

2. Prior Literature

Although scholars have written on trademarks and the trademark system, empirical

research is limited. The extant literature mostly examines the system from a historic or

legal perspective (e.g., Schechter, 1927), or through economic theory (e.g., Landes and

Posner, 1987). The marketing literature has consideredtrademarks within the context of

strategic brand management (e.g., Elliot and Percy, 2006) and intangible asset valuation

(e.g., Wilkins, 1992). Where valuation is done, it tends to be more art than science and

subject to significant subjective considerations (Smith, 1997).

Empirical literature on trademarks, although limited, has recently been expanding,

particularly in Europe. Sandner (2009) investigates how patent and trademark portfo-

lios impact the market value of European firms. He observes complementarities between

patents and trademarks and finds that holding each contributes incrementally to firm

USPTO Trademark Case Files Dataset 671

value. Sandner and Block (2011) estimate the effect of trademarks on firm market value,

employing a methodology frequently used with patent citations, and find associations

between firm value and indicators associated with trademark value, such as seniority,

oppositions, and the number of goods and service classes. Employing data from the Ox-

ford Firm Level Intellectual Property Database, Greenhalgh et al. (2011) show a positive

link between trademark registration activity and firm performance in productivity, em-

ployment, wages, and growth rates in the United Kingdom. Graevenitz (2007)studies

trademark oppositions, examining the reputational effects of European firms defending

established brands by opposing potential imitator mark applications. He finds that a

reputation for tough opposition has a strong effect on the probability of settlement and

suggests that firms may benefit from such a reputation through reduced opposition

costs.

There has been some empirical research on the relationship between trademarks

and innovative activity, particularly related to service and high-tech sectors (Schmoch,

2003; Mendonca et al., 2004). Millot (2009) studies the link between trademark filing and

innovating, observing that trademark data can identify innovation outside the fields of

engineering and science, such as in marketing and service innovation. Likewise, Gotsch

and Hipp (2011) find trademarks to be useful indicators of innovation in knowledge

intensive services (KIS) industries based on German survey data. Schmoch and Gauch

(2009), more generally, consider the advantages and limitations of using trademarks

to measure innovation in service industries. Heimonen (2012) examines the factors

that affect innovation, as measured by intellectual property rights including patents,

trademarks, utility models, and registered designs, in small and medium-sized firms

in Finland. Using data on Portuguese firms, Ferreira and Godinho (2011) estimate an

innovation function on a regional scale to distinguish the importance of local factors,

including research and development, the structureof human resources, and the existence

of KIS and technology companies, in the growth of innovation. They model a composite

measure of innovation, using both patents and trademarks, to emphasize the diversity

of intellectual capital outcomes of the innovation process.

Studies employing U.S. data are less common. Somaya and Graham (2006)find

complementarities between trademark and copyright use by U.S. packaged software

firms during the 1990s using both trademark litigation and registration data. Ceccagnoli

et al. (2010) use trademarks registered to firms as a proxy for marketing capabilities

and assets complementary to firm innovation. Krasnikov et al. (2009) use data compiled

from the USPTO’s Trademark Electronic Search System (TESS) to measure the financial

returns of firms’ branding efforts. They find that the total number of brand-association

trademarks available to firms increases their financial performance. Port (2008) examines

U.S. trademark litigation strategies, presenting evidence indicating that litigation can be

used to deter market entry.In another U.S. study, Beebe (2006) examines the consistency

of how the likelihood of confusion test has been applied across different federal courts.

Despite these few examples, studies using U.S. administrative data remain scarce.

To help remedy this paucity of research, the USPTO Trademark Case Files Dataset

provides a large volume of relevant data to the research community for analysis. To

facilitate greater use, these data have been provided in a relational database in formats

compatible with standard statistical software packages. Nevertheless, because these

data are not commonly used in the research community, a comprehensive description

of these data is desirable. In the next sections, we introduce the user to these data

and the institutional environment under which they are generated by companies and

individuals registering trademarks at the USPTO.