Classroom Companion: Business

 
Chapter 20 · Big Data Economics

---

321
20
and sometimes disastrous—decisions. Handled with care and expertise, big data is 
a formidable competitive tool in the digital economy improving customer satisfac-
tion and perfecting sales and marketing operations by precisely targeted informa-
tion bulletins and ads.
The challenge is that the firm must possess deep knowledge in advanced data 
management tools such as artificial intelligence, machine learning, expert systems, 
and data mining. The result may be that the company is not able to discover and 
utilize the huge amount of data it may possess about its business operations and 
customers.
Big data offers big opportunities in several sectors, for example, in health care, 
marketing, digital service provision, statistics, and management of public services. 
Large amount of data about people is collected by government bodies, both public 
and clandestine, by intercepting Internet traffic; receiving data captured by social 
media providers, telecommunications operators, and application providers; storing 
information received from surveillance cameras; and storing biometric informa-
tion about inhabitants and visitors. This information is used for crime prevention, 
criminal investigation, and antiterrorism. The same data may be misused for social 
control of the population and for identifying, tracking, and harassing dissidents 
and political opponents.
 
?
Questions
1. Is causation and correlation the same thing? Explain.
2. From which sources do data brokers collect information? Hint: see, for exam-
ple, ProPublica (Everything We Know About What Data Brokers Know About 
You) and Clearcode (What Is a Data Broker and How Does It Work?).
 
v
Answers
1. No. That two events are causally related means that one event is caused by the 
other event. That two events are correlated means that a linear statistical rela-
tionship exists between them (e.g., both increase at statistically proportionate 
rates—or one increases, while the other decreases at a proportionate rate). The 
events may be correlated because the events are causally related, both are 
caused by a third event and not causally related themselves, or they are entirely 
unrelated but varies in the same way (spurious correlation). If the relationship 
between two variables is nonlinear, then the correlation between them is zero so 
that causation does not imply correlation.
2. The data brokers may buy or retrieve information from several resources, for 
example:
5
Publicly available information (e.g., birth certificates, criminal registers, taxa-
tion registers)
5
Webpages mentioning the person
5
Loyalty cards for shops, hotels, etc.
5
Membership lists of organizations
5
App owners
5
Dealers of products that can be connected to data about the customer (e.g., 
car dealers and realtors)