A Preliminary Framework to Fight Tax Evasion in the Home Renovation Market

A Preliminary Framework to Fight Tax Evasion in the Home Renovation Market

Cataldo Zuccaro, Michel Plaisent, Prosper Bernard
Copyright: © 2021 |Pages: 22
DOI: 10.4018/978-1-7998-4963-6.ch015
(Individual Chapters)
No Current Special Offers


This chapter presents a preliminary framework to tackle tax evasion in the field of residential renovation. This industry plays a major role in economic development and employment growth. Tax evasion and fraud are extremely difficult to combat in the industry since it is characterized by a large number of stakeholders (manufacturers, retailers, tradesmen, and households) generating complex transactional dynamics that often defy attempts to deploy transactional analytics to detect anomalies, fraud, and tax evasion. This chapter proposes a framework to apply transactional analytics and data mining to develop standard measures and predictive models to detect fraud and tax evasion. Combining big data sets, cross-referencing, and predictive modeling (i.e., anomaly detection, artificial neural network support vector machines, Bayesian network, and association rules) can assist government agencies to combat highly stealth tax evasion and fraud in the residential renovation.
Chapter Preview


This article introduces the reader to the role of analytics and how governments can fight tax evasion in the construction and home improvement sector. According to the Institute for Operation Research and the Management Sciences (INFORMS, 2017), analytics aims at assisting organizations to fulfill their objectives by using scientific and mathematical methods to analyze their data, describing and enlighten the meaning of past facts, forecasting and giving advice for future decision-making. The following sections will guide the reader on the economic importance of the home improvement market and its underground economies while presenting how analytics is already used in several contexts to combat tax evasion and to present a structured approach, inspired by CRISP–DM methodology (IBM, 2016), that can be applied to tackle the thorny problem of tax evasion in the home renovation industry. After identification of the main sources of data, a list of relevant studies helpful to fight tax evasion is presented.

The ‘do-it-yourself’ (DIY) home improvement and renovation markets overlap to constitute an important market which is expected to grow in the next years (Reportlinker, 2020). The major players in the market are the stores selling tools and goods (lumber, paint, structural material, electrical, plumbing and heating supplies), trades’ people and individual households. To service the needs of its customers, hardware stores in the USA create roughly 668,459 jobs (IBISWord, 2019). On the consumer side, the Harvard’s Joint Center for Housing Studies, estimated that roughly 400 billion was spent by homeowners for renovation supplies and equipment in 2017 (Hunter, 2020). A compilation in 2018 by Statista (2020) stated that sales in the home renovation market were already 394 billion USA dollars and in 2019 rose to almost 400 billion dollars. In addition, according to a study by Riquier (2019), the market was expected to grow to 450 billion dollars by 2023 and Bustanete (2018) predicted the market to reach 680 billion dollars by 2025. The retail hardware store market is largely dominated by Home Depot, Lowe’s, Menard and McCoy (IBISWord, 2019). The 2018 sales for the main retailers of the USA were estimated (Statista, 2000) at 110 billion dollars for Home Depot, 72 billion dollars for Lowe’s and a combined 38 billion dollars for Sherwin-William, Kingfisher and Fastenal (Statista, 2020).

In this very dynamic market, 80% of homeowners will hire someone to do the work according to Bustamante (2018) but only 51% of these would be licensed tradesmen and this is a major problem in detecting fraud and tax evasion in the home renovation market. An experiment conducted by Doeer & Necker (2018) with 2900 business illustrates the extend of this complicity in the home improvement in Germany, in which they found that 56% of businesses responding to an advertising asking for painting services, offered to evade tax payment. Bitzenis et al. (2016) summarize the situation by the fact that the shadow economy overlaps legal, illegal economy and household economy.

Fraud and tax evasion problems with construction are present in many countries that have seen the home renovation market grow dramatically during the last decade. For example, in Quebec, second most populated province of Canada, the market for home renovation is estimated to be nearly 50 billion Canadian dollars (Bedford, 2020) and according to tax and income department Revenu Quebec (2020), the construction industry is responsible for 40% of total tax evasion. In France, hidden work (undeclared but taxable) is responsible for between 6.8 and 8.4 billion euros (Chemin, 2019), and half of which was detected in construction (ACOSS, 2020).

Tax evasion and tax fraud can wreak havoc on the treasuries of most countries. The mean percentage for Europe was estimated to be 18.6% (Scheinder et al., 2015). Two clear examples of this are Italy and Greece where the shadow economy has been estimated respectively to be 22% of GDP and 20% while only 6% in the USA (Vousinas, 2017). In Australia, the shadow economy in the construction industry accounted for 10% for a similar period (Chancellor & Abbot, 2015). A previous study from Cebula & Feige (2012) based on time series from the International Monetary Fund estimated the shadow economy to 18–23% of a country’s total revenues.

Obviously, it is impossible to eradicate fraud and tax evasion. There is a need to vigorously pursue and prosecute tax evaders with both the legal and technical tools available to governments. Legal remedies are specific to each country; however, today technical solutions are available to governments to reduce significantly fraud and tax evasion. This is how data mining can come to the rescue.

Key Terms in this Chapter

Methodology: Set of appropriate research methods and techniques applied to study a particular field or determined problems in order to find a solution or reach a goal. The CRISP-DM methodology is a sequenced list of tasks to be performed in order to reach conclusions on data.

Fiscal Fraud: Subterfuge, illegal acts used by a taxpayer (individual or business) in order to evade the tax burden to which he would normally be submitted by the law as a result of generating personal income.

Anomaly detection: Highlight, discovery, identification of an irregularity, a deviation from expectations arising from significant deviation (more or less) with a standard or a majority of cases a priori similar and potentially indicative of fraud, error or fault. Synonym for outlier detection.

Accuracy: The degree of proximity of a measurement in relation to the target true value. Sometimes called “trueness”, it is often confounded with “precision” which refers to the variability of the results within one set of measurement.

Data Mining: Data analysis technique allowing extraction of new information, hidden correlations difficult to see under a mass of data, trends, anomalies, associations, mainly by the use of statistical processing, often in the context of big data.

Algorithm: A finite sequence of operating rules executed on data and which allow a result to be obtained. The sequence may comprise some cycles and conditional branching. In daily life, it compares to a recipe.

Big Data: A huge and moving set of data of various varieties, structure, and format, produced by various sources, at various frequencies that must be processed quickly using specialized tools, other than common database management software.

Data Modeling: The development of a statistical or algorithmic model to describe, explain and predict a well-defined phenomenon or problem by employing a large variety of variables and observations (cases).

Attribute: Characteristic attached to an entity (person, phenomenon, and situation) and used to describe it in a specific way, on a particular aspect. Information, normally considered as not decomposable, which determines and characterizes an entity.

Complete Chapter List

Search this Book: