The diffusion and adoption of Mass Appraisal methods by municipalities in the Netherlands: A qualitative study on CAMA’s and AVM’s within the Special Act for Real Estate Assessment

(1)

1

The diffusion and adoption of Mass Appraisal methods by municipalities in

the Netherlands

A qualitative study on CAMA’s and AVM’s within the Special Act for Real Estate Assessment

Student: Peter van Erk, s1349937 Mentor: Johan Christensen Leiden University

(2)

2

Abstract

Since approximately 2005, algorithm based mass appraisal methods (CAMA and AVM) were completely diffused within Dutch municipalities. These mass appraisal methods are functioning as a helping hand to manual valuations in determining the WOZ-value of a property for tax purposes. Since that time, municipal organizational size, culture, infrastructure, learning and interorganizational relations had influence on the decision of municipalities to make full use of CAMA’s and AVM’s as the best course of action available. In this research the specific influences of these concepts on the adoption of mass appraisal methods by municipalities are explained and the process of diffusion of these methods over time is described.

Keywords Information Technology, traditional manual valuation, CAMA, AVM, Hybrid

valuation and Special Act for Real Estate Assessment

Acronyms

CAMA Computer Assisted Mass Appraisal

AVM Automated Valuation Model

IAAO International Association of Assessing Officers

EAVMA European AVM Alliance

RICS Royal Institute of Chartered Surveyors

TEGoVA The European Group of Valuers Associations

Used translations

Waarderingskamer Netherlands Council for Real Estate Assessment

Wet WOZ Special Act for Real Estate Assessment

VNG Cooperation Agency of the Association of Netherlands Municipalities

Waardepeildatum Valuation date

(3)

3

Definitions

Adoption: “The decision of an organization to make full use of an innovation as the best

course of action available” (Rogers, 1995:171).

Algorithm: A kind of recipe (a list of instructions) in which you tell a computer how to reach

a certain goal (Vaandrager, n.d.).

Analyst Assisted AVM: “A Hybrid Valuation that relies on the experience and judgment of a

professional, but not necessarily a qualified surveyor, to validate and supplement the output of an AVM” (EAVMA, 2017:6).

AVM Assisted Appraisal: “A Hybrid Valuation that relies on the experience and judgment of

a qualified surveyor, to translate the output of an AVM into a legally compliant valuation. This is obtained without conducting a physical inspection of the subject property, although it is supported by Comparable Evidence, which may or may not incorporate data from the AVM” (EAVMA, 2017:6).

Batch Valuation: “The process where a large number of AVM results are obtained without

individual manual submission. Batch valuations are used to value a large number of properties, e.g. for capital modelling, provisioning, whole loan trading, surveyor management etc” (EAVMA, 2017:3&7).

Communication channel: “A communication channel is the means by which messages get

from one individual to another” (Rogers, 1995:36).

Computer Assisted Mass Appraisal: “The process of using a computer to help assessors

update property values and keep them up-to-date for ad valorem tax purposes. Basically, the computer can be seen as a device to store a massive file of current, accurate information about property characteristics, recent real estate transactions, cost, depreciation, and other economic factors for the real estate market. This information can be analyzed in a number of ways to generate value estimates for all properties and to produce a number of reports that aid the assessor” (Ryan & Le Vasseur, 2002:5).

Diffusion: The spread of a new technology across the social system (Rosenberg, 1972:23; &

Kee, 2017:1)

Drive By: “The valuation produced by a qualified surveyor following only the external

inspection of a property. This is typically conducted literally through a drive-by, merely checking the property's existence, apparent external condition and neighbourhood characteristics, without stopping to assess any of its individual characteristics” (EAVMA, 2017:5).

House Price Index: “A time series capturing the price development of residential properties

over time” (EAVMA, 2017:6).

Hybrid Model: “An analysis that incorporates elements from different models, e.g.

Comparable Based, Hedonic, Indexation etc” (EAVMA, 2017:6).

Hybrid Valuation: “Generic term used to include all of AAAVM, SAAVM and AVMAA”

(EAVMA, 2017:6).

(4)

4

enterprise software, storage systems and audio-visual systems which enables users to acess, transmit, store and manipulate information (FOLDOC, 9 sept 2008).

Innovation: “An idea, practice or object that is perceived as new by an organization”. (Rogers,

1995:10)

Innovativeness: “The degree to which an organization is relatively earlier in adopting new

ideas (regarding the innovation) than other members of a social system” (Rogers, 1995:37).

Market value: “The estimated amount for which a property should exchange on the date of

valuation between a willing buyer and a willing seller in an arm’s-length transaction after proper marketing wherein the parties had each acted knowledgeably, prudently, and without compulsion.” (International Valuation Standards Council, 2017:18)

Portfolio Valuation: “The circumstances and purpose often defining a distinct AVM

product”(EAVMA, 2017:7).

Professional appraiser: A professional who prioritises the quality of his services and acts in

the public interest and has the necessary knowledge and expertise and practices the profession independently, objectively and with integrity (NRVT, n.d.)

Social system: “A social system is a set of interrelated units that are engaged in joint problem-

solving to accomplish a common goal” (Rogers, 1995:37).

Surveyor Assisted AVM: “A Hybrid Valuation that relies on the experience and judgment of a

qualified surveyor, to validate and supplement the output of an AVM. The modifications or manipulations introduced by the surveyor onto the AVM output and/or the Comparable Evidence removes the objectivity and integrity of the fully automated process and it may compromise its unbiased nature” (EAVMA, 2017:6).

Traditional manual valuation: The manual valuation of a property by a recognized

professional appraiser.

(5)

5

Chapter 1./ Introduction

Most of the time algorithms are used in order to process available data and to do something useful with the information a computer contains. An algorithm is a kind of recipe (a list of instructions) in which you tell a computer how to reach a certain goal (Vaandrager, n.d.). In the private sector this is a proven method. In the public sector there are also used algorithms, for examples ones that automate or assist with the process of decision making. Some of these decisions and assessments don’t have that big of an impact. The algorithms of municipal cars that drive around and automatically determine whether you need to receive a parking fine or not clearly have a smaller impact than an algorithm that determines whether to offer someone council housing or to give someone probation. The logic of the algorithm behind this kind of decisions are therefore more serious in terms of consequence (Copeland, 2018).

In a Dutch Parliament debate in October 2017, the minister of Economic Affairs stated that he actually doesn’t know what kind of algorithms are used within the whole of the Dutch government and what could be the possible downsides of this use.1 The minister promised a research on the use of algorithms within the Dutch government, which is currently carried out by his people. As this is the master thesis of student on Public Administration, this is not a research on the up- or downsides of a particular algorithm used within the Dutch government. Those more quantitative questions are better to be answered by econometrists and data scientists.

Where this research contributes is by providing important insights in the diffusion and adoption of different types of algorithms used by Dutch municipalities. To be more specific, the algorithms of Computer Assisted Mass Appraisal (CAMA) systems and Automated Valuation Models (AVM’s), which assist as a helping hand during the yearly property appraisal for real estate taxation by Dutch municipalities.2 The usage of algorithm based methods in valuation of real estate properties is one of the fastest rising trends in the art of information technology (IT) and real estate valuation and the usage of these method has increased considerably all over the world in the last years (Droj and Droj, 2017:7).

1_{The House of Representatives 2017 -2018, 26 643, nr. 8 Stenogram Digitalisering Infrastructuur.}

2_{Two examples of different types where algorithms are used to determine the market value for the Special Act}

for Real Estate Assessment are Geotax (https://www.geotax.nl/taxeren-en-waarderen/waardebepaling/) and Ortax (http://insights.ortec-finance.com/nl-nl/2017/07/14/machine-learning-het-ultieme-woz-waarderingsmodel)

(7)

7

1.1 Research question

On October 27 in 2004 there was a debate in the Dutch Parliament on a bill called ‘Alteration of the Special Act for Real Estate Assessment and of any other laws’. The main goal of this bill was to make changes in the Special Act for Real Estate Assessment in order to stimulate a more efficient execution of the act at that time. During this debate the growing use of mass appraisal methods was one of the main topics under discussion. Some members of Parliament demanded certification or regulation regarding the companies that were producing mass appraisal methods. Also the certificating of the methods itself was one of the suggested options. Other members of Parliament, and also the Secretary of State, were not willing to agree upon this kind of a plan. Doing this in the beginning development phase of these mass appraisal methods would discourage creativity of the developers and would not lead to the desired innovation of the methods.3

Till this time, 2018, there is not implemented any certification or further regulation on mass appraisal methods or their suppliers within the Special Act for Real Estate Assessment. This master thesis is oriented towards the collection and analysis of empirical data related to the diffusion and adoption of these algorithm based mass appraisal methods within the Act. Where CAMA’s and AVM’s can be used for both the appraisal of homes and non-residential properties. This study focuses particularly on the CAMA’s and AVM’s that are used for the appraisal of homes. The research questions are thereby as follows, divided in a more descriptive (1) and an explanatory (2) part:

1) How did the use of mass appraisal methods by Dutch municipal organizations diffuse over time?

2) How did internal organizational characteristics of Dutch municipal organizations influence the innovative adoption of CAMA’s and AVM’s, since the complete diffusion of mass appraisal methods as a helping hand in determining the market value of properties for the Special Act for Real Estate Assessment?

The reason for also incorporating a descriptive question in this research is because in order to understand adoption it is necessary to trace back historical events as these led to the current organizational dilemma’s on adoption (Lyytinen & Damsgaard, 2001:185).

3

(8)

8

1.2 Social and academic relevance

Detailed features on the diffusion and adoption of CAMA’s and AVM’s in Dutch municipal real estate appraisal are at this moment mostly unknown or only in hands of a small group of professionals. In this research, practical insights around Dutch CAMA’s and AVM’s are gained which helps in order to get a clearer picture on the diffusion and adoption of algorithms in municipal organizations. In the light of the current political discussion and research on the use of algorithms by government this is socially relevant. Even more socially relevant regarding this specific research topic is that the board of the Netherlands Council for Real Estate Assessment is at this moment discussing on the developments regarding the use of mass appraisal methods within the Special Act for Real Estate Assessment. A central question asked by the board is for example if the desired modernization of the valuation systems /

models is sufficient? Hereby the board focuses on the different use of CAMA’s and AVM’s

within the Act. This research provides in answering this and other important questions on the development of valuation systems/models within the Act.

In a major textbook from the 80’s, called ‘Management Information Systems’, is stated that it is ultimately not the technology itself, that leads to higher innovative adoption of new technologies by organizations but this rather helps in enabling new approaches and forms. Once in position, these approaches and forms are getting confirmed through management action. These often slow and unexpected painful adoption of IT innovations within organizations made researchers seek to understand and predict these processes (Lyytinen & Damsgaard, 2001:174). But, research concerning IT innovation is in its infancy. What is missing at the moment are reliable evidence-based studies on current innovatively adoption of IT in organizations (Bretschneider & Mergel, 2011:199).

1.3 Further structure

Before the focus in chapter 3 will went to the relevant literature for this research, you first find a chapter on the background and important definitions in this research. It essential to have this knowledge in order to continue reading. The fourth chapter contains information on the research design and data collection. The theoretical framework of chapter 3 was tested by interviewing thirteen professionals in the Netherlands who presented four different perspectives on the development of AVM’s: user, developer/researcher, interest group and supervising authority. You can find the results of these and other empirical data in the analysis of chapter 5 were both research questions are answered. In the conclusive chapter a

(9)

9

summary of the results, the limitations of the research, possible avenues for future research and policy recommendations are stated down. In the appendix you can find the interview guide that was used during the research.

(10)

10

Chapter 2./ Background and important definitions

In this chapter I define and discuss the key definitions employed in this study: market value, traditional manual valuation, CAMA, AVM and hybrid valuations. Thereby I already provide some background information on the recent rise of mass appraisal methods in the Netherlands. More detailed information on this can be found in the descriptive part of this thesis in chapter 5. The explanatory part of this thesis will start where this chapter ends. Namely from the fact that every municipality in the Netherlands is at this moment in some way or the other using mass appraisal methods as a helping hand in determining the market value of a property for tax purposes. This mixed valuation approach is also called a hybrid valuation. The explanatory part will focus on how internal organizational characteristics of Dutch municipal organizations influence the decision of municipalities to make full use of CAMA’s and AVM’s as the best course of action available.

2.1 Background

The Special Act for Real Estate Assessment was initiated in 1995 in the Netherlands. The WOZ-value that is at the core of this Act is the value of a property determined by Dutch municipalities for tax purposes and is a reflection of the market value of a property:

‘’Market Value is the estimated amount for which a property should exchange on the date of valuation between a willing buyer and a willing seller in an arm’s-length transaction after proper marketing wherein the parties had each acted knowledgeably, prudently, and without compulsion.’’ (International Valuation Standards Council, 2017:18)

Despite traditional manual valuation being extensively employed by professional appraisers during the 90’s, over the last several years there has been a move towards mass appraisal methods as a helping hand in the valuation process (Francke, 2010:5). A big reason for this change is the situation of a large number of homes to be yearly valuated, doing this manually can be extremely time consuming (Matysiak, 2017: 2). These mass appraisal methods are based on appraisal models. The International Association of Assessing Officers (IAAO) defines an appraisal model as a representation, in words or as an equation, that explains the relationship between the final sale price and variables representing demand and supply (Moore, 2018:34).

(11)

11

2.2 Definitions

But, right now in 2018, there is recognized a problem of various terms regarding methods in mass appraisal. Kindt and Metzner (2017:6) found the terms AVM, CAMA, Portfolio Valuation and Batch Valuation in different researches conceptualized identically whereas there are clear differences between these terms. So, before this research continues it is important to determine which of these forms are used by Dutch municipalities, which not and how they can be conceptualized. Within the Special Act for Real Estate Assessment there are used, as just mentioned, manual and mass appraisal methods that come together in hybrid valuations. But on the side of mass appraisal methods some municipalities use a CAMA system whereas other municipalities use an AVM. This is a remarkable point of this research because normally, seen from international perspective, a CAMA is linked to governmental use and an AVM to mortgage lending. Portfolio Valuation and Batch valuation are not methods that are used by Dutch municipalities. These methods are only used by lenders. From now on the focus of this research will continue on the mass appraisal methods CAMA and AVM, within the Act for Real Estate Assessment in the Netherlands.

There is not one universal type of CAMA or AVM that is used all over the Netherlands and there are important differences between the two approaches (table 1). Every CAMA and AVM is adapted to the local municipal situation. Examples of different used methods by municipalities that fall under CAMA’s or/and AVM’s are direct comparison, clustering, (hedonic) regression analysis and trend calculations (Bervoets, 2018:2). In the abstract you can find the conceptualization of CAMA that is used in this research. In short, a CAMA is an informational system for storing, collecting and valuating properties (Droj & Droj, 2015:2) As just mentioned, a critical point for this research is that a CAMA is a mass valuation that is done by a governmental agency using a software solution for the purpose of computing property tax. The main difference between a CAMA and an AVM in the Dutch context is that a CAMA is less or not based on statistical methods in comparison with an AVM. To point out the differences between CAMA and AVM some of their differences, determined by the European AVM Alliance (EAVMA) members from The Netherlands, Spain, Italia, Norway, United Kingdom, Germany, Austria and Sweden, are listed:

(12)

12

CAMA AVM

Valuation purpose 4 Tax on property Mortgage lending

Recipient of the valuation 5 Municipality or national

government

Bank or other party in mortgage industry

Adjustments of the model Frequent adjustments by

surveyor/appraiser

No adjustments in mathematical algorithm

Frequency of valuation Yearly valuations or few

years basis

Daily or monthly basis

Table 1: Internationally defined differences between CAMA and AVM (EAVMA, 2017:4).

When it comes to AVM’s there can be distinguished four commonly used definitions. AVM’s are seen as the main instrument in the mass appraisal of property by the IAAO (Droj & Droj, 2015:2). According to the IAAO standard an AVM is:

“A mathematically based computer software program that produces an estimate of

market value based on market analysis of location, market conditions, and real estate characteristics from information that was previously and separately collected. The distinguishing feature of an AVM is that it is an estimate of market value produced through mathematical modeling. Credibility of an AVM is dependent on the data used and the skills of the modeler producing the AVM. ”(IAAO, 2003:5)

A working group on AVM standards of the Royal Institute of Chartered Surveyors (RICS) also defined an AVM whereby they explicitly exclude human intervention during the process:

“Automated Valuation Models use one or more mathematical techniques to provide an estimate of value of a specified property at a specified date, accompanied by a measure of confidence in the accuracy of the result, without human intervention post-initiation,” (RICS, 2013:5)

The European Group of Valuers Associations (TEGoVA) came to the following definition of an AVM in their European Valuation Standard 6:

“AVMs are statistically-based computer programmes which use property information to generate property-related values or suggested values. This statistical method of

4

Not completely true for The Netherlands, AVM also used for property tax. 5

(13)

13

valuation seeks to arrive at the value of a property directly through the application of a mathematical algorithm to a data base of transaction prices and property characteristics whilst omitting an explicit traditional valuation, such as a comparative or income based approach valuation and a professional valuer's qualitative assessment of the value. (TEGoVA, 2016:2)”

In some contrary to the more ‘valuer based’ definition of TEGoVA. As TEGoVA is one of the major interest groups defending the interests of the professional appraiser. The EAVMA, which its interests can be seen as the opposite of TEGoVA, defines an AVM as:

“A system that provides an estimate of value of a specified property at a specified date, using mathematical modelling techniques in an automated manner.”(EAVMA, 2017:1).

Because of this many different definitions of an AVM with different related interests there is not chosen one main definition in this research. The main point which is taken into account in this research is that all AVM definitions exclude any professional appraiser involvement in the process of arriving at an estimate of value. So this is the determining factor in order to be an AVM within the Special Act for Real Estate Assessment. Only the end value of an AVM can be changed when there is a good substantiated motivation to do so. Within a CAMA the professional appraiser or analyst has the possibility of model adoption during and after the process, this is a difference between the two approaches as also came forward in table 1.

In the process of determining the market value there are used manual and mass appraisal methods at the same time in order to come to the best value. Manual approaches are using traditional valuation, models in words, drive by, asking value etc. Mass Appraisal methods are using purchase price, CAMA, AVM, House Price Index, hybrid models etc. Beside these two extremes of valuation types there can also be observed hybrid valuation which is formed out of these two sides. These hybrid valuations contain AVM Assisted Appraisal (AVMAA), Surveyor Assisted AVM (SAAVM), Analyst Assisted AVM (AAAVM). Comparable hybrid valuations can be thought of related to CAMA. These hybrid valuations are now lying at the core of the work that is carried out by municipalities within the Act for Real Estate Assessment in order to determine property values for taxation purposes.

(14)

14

Chapter 3./ Theory

This chapter reviews the literature on the diffusion of innovation and the thereafter following adoption of it by organizations, together with the organizational factors that have influence on this. As the innovation of algorithm based mass appraisal methods can be placed in the field of IT, there will be started with a brief general paragraph on IT in order to get a clear picture of what this precisely is. After this paragraph the two general theoretical questions of this research are stated down with as goal to find the most relevant literature and to help forming a theoretical framework out of this. Diffusion theory is used to answer the descriptive theoretical question on diffusion of innovations over time. Out of the literature review there is formed a theoretical framework on the descriptive question. Also for the explanatory part about adoption there is formed a theoretical framework out of a literature review. In this framework the existing organizational size, culture, infrastructure, learning and interorganizational relations have been determined as factors that affect the adoption of new technology by organizations. Each of these factors is given its own section and rationale and is in the end transformed into a hypothesis. The descriptive part is also given one hypotheses in the end of its section.

3.1 What is information technology?

IT is commonly defined as communication systems, computers and all the necessary enterprise software, storage systems and audio-visual systems which enables users to access, transmit, store and manipulate information (FOLDOC, 9 sept 2008).

In the last 50 years there occurred huge changes in IT. Through that time predictions were made by management scholars, social scientists and public administration researchers on possible future changes within the field of IT and its influence on social, political and economic institutions (Emery, 1964:1). These predictions focused on the structure and processes of organizations. Examples of such predictions are that the decision-making process would speed up and costs would be reduced as a consequence of IT. But despite these sort of claims, many of them remain unrealized today, particularly and generally so in government (Bretschneider & Mergel, 2011:187).

The availability and use of IT and related systems is now recognized almost to the point of being a commodity at the same level as labor (Dewett & Jones, 2000:313). The extent to which a technology succeeds in making an organization innovative is not dependent on the

(15)

15

technology itself but on the adaption of it by the bureaucratic or political organization and the barriers that appear between institutions (Bretschneider & Mergel, 2011:199). This observation is an important starting point for the sequel of this literature review and this further research as this will focus on the internal organizational characteristics that influence the adoption of algorithm, and so in a sense IT based, mass appraisal methods.

3.2 Theoretical questions

But before the literature review will start it is important to focus on the theoretical questions of this research. These are at the base of the literature review. By answering the theoretical questions it is possible to form answers on the research questions related to mass appraisal methods later on in this research. The two theoretical questions to be answered in this chapter are:

- How does a new technology diffuse over time? (paragraph 3.3 & 3.4)

- How do internal organizational characteristics influence the adoption of a new information technology in an organization? (paragraph 3.5 & 3.6)

As already mentioned on the previous page there have been developed many theories by management scholars, social scientists and public administration researchers which attempt to give an explanation on the relationship between technological change and social, political and economic institutions. Out of all these separate theories on the spreading of technology there can be considered the two broad groups of diffusion and adoption theory (Ruttan, 2001). Adoption is hereby called the decision of an organization to make full use of an innovation as the best course of action available and diffusion is the spread of a new technology across the social system (Rosenberg, 1972:23; & Kee, 2017:1). These two theories form the basis for our understanding of the descriptive question on technology diffusion and the explanatory question on the adoption of technology. Where the descriptive part is more aimed at diffusion theories related to new technique in general as there is not produced specific literature on information technology regarding this broad phenomenon. The explanatory part is more focused on information technology as there are more specific adoption theories available and applicable.

(16)

16

3.3 Literature review: Diffusion of innovation

An innovation is an idea, practice or object that is perceived as new by an organization (Rogers,1995:11). The main elements in the diffusion of new technologies can be described as an innovation which is communicated through certain channels, over time and among members of a social system (Rogers, 1995:35-37). A social system is hereby a set of interrelated organizations that are engaged in joint problem-solving to accomplish a common goal (Rogers, 1995:37). The theory of diffusion is, during the years, broadly applied to all kind of technologies ranging from new ideas to new machines (Bayer & Melone, 1989:162) . This makes diffusion theory forming the basis for our understanding of the spreading of information technology. The theory is more easily described as individuals that adopt something new and when they come in contact with others they communicate their experience about the new product or service (Coleman, Katz & Menzel, 1957:258). The communication channel they use for this is the means by which messages get from one individual to another (Rogers, 1995:36). The diffusion theory is derived from the theory of contagion, this theory was applied to the disease process in the eighteenth century (Mahajan & Peterson, 1985). The landmark study on diffusion appeared in 1943 when the rural sociologists Gross and Ryan published a work on the diffusion of hybrid corn seed in two farming communities in Iowa. This study caused a growth in diffusion studies in the years thereafter. With rural sociology as the starting point, the research paradigm of diffusion was in the end of the 1960s also spread to economics, public health, education, communication and other fields (Kee, 2017:2).

When it comes to diffusion studies specifically focused on social sciences, it is Everett Rogers, a professor of rural sociology, who published the first edition of the book Diffusion of

innovations in 1962. In that book the till that time over 508 scattered studies on diffusion

where synthesized (Hanel & Niosi, 2007:5; Kee, 2017:2). By doing this he developed theories on the diffusion and adoption of innovations among individuals and organizations. The book

Diffusion of innovations and later publications of Rogers are one of the most cited on the topic

of diffusion. Diffusion itself is the second most cited social science theory in general (Rice, 2009:489).

3.4 Theoretical framework

The four main elements of diffusion, as just described, are the nature of the innovation, communication channels, time, and social system. As Rogers (1995:10) states, these elements are to a certain extent returning in every diffusion and adoption research. Beside these

(17)

17

elements there is, by Grajek and Kretschmer (2009:239) argued that the economic variable of price also has an important influence on the diffusion of technology. It are the elements on the independent variable ‘time’, as described by Rogers in 1995, that are needed in order to answer the descriptive theoretical question of this research. The inclusion of the independent variable of time is one of the strengths of diffusion research according to Rogers, but at the same time he realized that this variable can be criticized as that is a non-isolated variable (Rogers, 1995:20). At the base of these ‘time’ theories lies the paper by Ryan and Gross (1943:23), wherein the S-shaped curve of diffusion and the first classification of adopters was developed (Hanel & Niosi, 2007:4). Innovativeness, the rate of adoption of the innovation and the innovation-decision process of new technology have been recognized as elements of time (Rogers, 1995:36). These elements are now turned into a three sided theoretical framework.

3.4.1 Innovativeness

In his book Rogers (1995:150) describes five categories of adopters: innovators, early adopters, early majority, late majority and laggards. The plotting over time of categories of adopters of an innovation usually follows a normal, bell shaped curve (fig.1).

Fig. 1 Categorization of adopters on the basis of innovativeness (Frank et al., n.d.)

These five categories are ideal types to categorize the innovativeness of the members of the social system. Innovativeness is the degree to which an organization is relatively earlier in adopting new ideas than other members of a social system (Rogers, 1995:37). To provide with a short description of these categories (Rogers, 1995:263-266):

- Innovators: The first organizations to adopt an innovation, hereby they launch the new idea by importing the innovation into the social system.

- Early Adopters: The second fastest category of organizations to adopt an innovation. The early adopter decreases uncertainty about a new idea by adopting it.

(18)

18

- Early Majority: This category adopt new ideas just before the average member of a social system. They take a longer time to adopt the innovation in comparison with innovators and early adopters.

- Late Majority: This category adopts a new idea just after the average member of the social system. They do this with a certain extent of skepticism.

- Laggards: This is the last category in the social system to adopt the innovation. These organizations have little or no opinion leadership.

Notwithstanding the wide use of this categorization there are also clear points of critique on it. In some situations the adoption by a social network happens at once. It is not clear what would mean early and what would mean late in this case. There can also be thought of processes where an innovation gets dropped by organizations, this is a sort of reversed process. Some researchers also found contradictory behaviors where laggards were more visionary in their uses of new technologies than those who are called innovators by Rogers (Lyytinen & Damsgaard, 2001:183).

3.4.2 Rate of adoption

The five categories of adopters can be placed in a so called S-curve where the cumulative number of adopters is plotted over time (fig. 2) (Fisher, 1971:76). This curve shows the rate of adoption, which is the relative speed with which an innovation is adopted by the members of a social system (Rogers, 1995:35-37). This curve rises slowly in the beginning when there are only a few adopters. It then accelerates to a maximum until half of the organizations have adopted the innovation. After this the rise gets slower because fewer remaining organizations adopt the innovation (Rogers, 1995:257).

(19)

19

So, the diffusion curve is divided into sequential stages. There are observations that suggest that the curve doesn’t follow the predicted shape of figure 2. This comes as a consequence of the points of critique from the first section which make the curve automatically getting another shape. There are also observed feedback loops in other studies. Hereby the local history, available information and the dynamics of the diffusion process affected the shape of the diffusion curve. An example of this kind of a study is Damsgaard (1996) who carried out a study on Trade link, the negative stigma around this firm resulted in continued failures for the product and in the end the inability to move beyond the initial stage of the curve.

3.4.3 The innovation-decision

The third part of this theoretical framework on technology diffusion over time provides with knowledge on the decision-making process of organizations or individuals with regard to new technology. There are determined three types of innovation decisions: optional innovation-decisions, collective-innovation decisions (choice made by consensus among the members of a system) and authority innovation-decisions (choices made by relatively few individuals in a system who possess power, status or technical expertise). A fourth category consists of a sequential combination of two or more of these types of innovation decisions (Rogers, 1995:35).

A point of critique on this concept of time is that this element should always be analyzed before the in paragraph 1 and 2 described elements. As mentioned in section 1, in some situations the adoption by a social community happens at once which has consequences for the whole process of diffusion.

This all results in the following hypothesis for the descriptive part. Later on this hypothesis will be linked to mass appraisal methods, in chapter 4 is explained how this is done:

Hypothesis 1: The categorization of adopters, the S-curve and a clear innovation-decision can, to a certain extent, also be recognized in the diffusion of mass appraisal methods within the whole of Dutch municipal organizations related to the Special Act for Real Estate Assessment.

(20)

20

3.5 Literature review: Adoption of innovation

Adoption is the decision of an organization to make full use of an innovation as the best course of action available. So this adoption can be done for a whole new innovation or for an innovation that is already used by organizations but wherefore new ideas have been developed in order to further improve the innovation (Rosenberg, 1972:23). Rejection is a clear decision not to adapt an innovation. If there are some uncertainties or if there is limited information about the consequences of the innovation on forehand, than one way to cope with this uncertainty is to try the new idea on partial basis (Hanel & Niosi, 2007:3). Rogers (1995:171) states that most organizations will not adopt an innovation without trying it first on a probationary basis in order to determine its usefulness in their own situation.

When it comes to the adoption of new information technology, there can be focused on the decision-making process within organizations that is related to this new information technology (Bretschneider & Mergel, 2011:188). Also for this topic of research, the work of Rogers can be seen as the main international recognized source of information. But in more recent years there derived several academic publications which also try to find the internal characteristics of organizations that have influence on the adoption of new technologies by organizations. This explosion of literature started since the publication of the in the first paragraph used study of Nathan Rosenberg from 1972. In this study Rosenberg provided several arguments on the adoption of new technologies that added to a perspective and program for further research (Hanel & Niosi, 2007:5). In the 1970s the economics and management took the lead in the area of diffusion which was previously held by sociology. In continuation of the descriptive part of this chapter, also for the explanatory part it is still Rogers, who has summarized the main contributions on adoption of innovations specifically for the social sciences in the 4th edition of his major book Diffusion of innovations.

Out of all the studies that were carried out after 1972 there can be distinguished a clear role for the size of the organization as the main organizational characteristic that affects the process of adoption. Beside Rogers (1995:379) who mentioned this, other publications on the side of IT are for example these of Dewett & Jones (2001:330), Bretschneider & Mergel (2011:190) and Jordan (2015:40). In these studies arguments were made on the consequences of the size of a governmental organization regarding the adoption of a new information technology. Also other organizational characteristerics have influence on the level of adoption

(21)

21

via the size of the organization. The organizational culture is one of this key features (Dewett & Jones, 2001:331; Jordan, 2015:69). Organizational cultural traditions and investments to change the existing (technological) infrastructure have influence on the adoption of a new information technology. There is no doubt that these factors cause heated discussions within organizations (Lyytinen & Damsgaard, 2001:185). The ability of the people in an organization to learn and assimilate new ways of behaving and thinking can play a decisive role in the adoption of a new information technology or the further improvement of the ‘old’ one (Reger, Gustafson, Dematie & Mullane, 1994:565 ). Finally, the process by which the adopted technology is adapted to governmental organizations involves human interaction between organizations on the specific form wherein the technology is best used (Rogers, 1995:333).

3.6 Theoretical framework

In the theoretical framework, that is following now, there is considered what the effects of the existing constructs of organizational size, culture, infrastructure, learning, and interorganizational relations are on the adoption of IT by organizations. This in order to answer the theoretical question. The clear focus of this part lies on theories on the adoption of IT within governmental organizations as this is the main theme of this master thesis. Hereby is searched for recent literature regarding the adoption of enterprise software and storage systems, as these parts of IT are the closest related to the mass appraisal methods CAMA and AVM. However, there are no consistent definitions of the measures of the used constructs, that is what Wisdom, Chor, Hoagwood and Horwitz (2013:480) argue in their review of theories and constructs regarding innovation adoption. They suggest that this lack of precise definitions and ways to measure the constructs point out the need for further work on the adoption of innovations in order to increase our understanding of it.

3.6.1 Size

When you look at the size of an organization and its ability to adopt a new information technology there can be made distinctions between small, medium and large organizations. Internal expertise, IT competencies and the availability of social and physical networks were found factors that motivate small and medium size organizations to adopt new technologies (Jordan, 2015:40). In contrast to this, for large organizations there was concluded that organizational leadership and senior managers are factors that had more influence on the

(22)

22

adoption of technology than any other factor (Jordan, 2015:143). Another research on the influence of size on the adoption of IT in governmental organizations resulted in the conclusion that larger organizations adopt new technologies earlier because they are better able to invest in a new information technology than smaller organizations (Hall & Kahn, 2002:22). Also Rogers (1995:379) came to this observation in his chapter on innovation in organizations. This leads to the following hypothesis on size:

Hypothese 2: In large organizations, organizational leadership and senior managers have a larger influence on the decision of a municipality to make full use of mass appraisal methods as the best course of action available, than any other organizational characteristics.

3.6.2 Culture

A study of Green (2007), on how to lead a workforce, led to an idea that is picked up by studies on the adoption of IT within governmental organizations. Green argued that internal organizational differences within a workforce could be a barrier to adopt new technologies in governmental organizations. A recently conducted research on the adoption of innovations showed that there is an important role for the community structure in the adoption of innovations. Proposals are more likely to be accepted if organizations are homogeneous. It is harder to adopt an innovation if the organizational culture is heterogeneously distributed (Sáenz-Royo, Gracia-Lázaro & Moreno, 2015:1 & 11). So if the opinion of the complete organization is in a sense not one sided this will affect the adoption of a new information technology. Bayer and Melone (1989:162) also recognized that innovations need to be compatible with existing values, needs of organizations or individuals and past experiences. This all leads to the following hypothesis on the factor culture:

Hypothese 3: The presence of a homogenous organizational culture is positively related to the decision of a municipality to make full use of mass appraisal methods as the best course of action available.

3.6.3 Infrastructure

Learning, that comes after this factor, is a central feature of the adoption of new information technology. It represents costs that cannot be recovered, if you adopt the new information technology the costs on learning to work with it are irreversible. But another important cost is the disposal of the previous used technology or methods. Many potential technology users

(23)

23

prefer to delay the adoption in order to reduce uncertainty around the new information technology (Hanel & Niosi, 2007:3). Also Bretschneider & Mergel (2011:190) agree on this point, preexisting structures determine the form and use of the new information technology. According to Lyytinen & Damsgaard (2001:183) the past decision history is not regarded important in the widely used theories of Rogers. However, they see clear influences of path dependencies because new information technology forms an add-on to the technology base that was already there.

Jordan (2015:144) built on this insight in his research on the adoption of enterprise software in government institutions. Responses in his research, from government employees, support the idea that the adoption of enterprise technology depends on how well the technology fits within organizational traditions. Hereby the respondents also expressed the belief that the complexity of a technology could put up a barrier in the adoption of enterprise technology. This brings us to the following expectation:

Hypothesis 4: The previous used technology or method has a negative influence on the decision of a municipality to make full use of mass appraisal methods as the best course of action available.

3.6.4 Learning

The in the literature review mentioned Rosenberg argued in his 1972 article that the skill level of employees is crucial for successful implementation and operation of a new information technology. If this requires complex new skills which are time consuming or costly to acquire this might cause the adoption of a new information technology to be slow (Hall & Khan, 2002:9). In a later research there was concluded that the lack of skilled manpower is often something that naturally comes along with the adoption of a new information technology but this could indeed form a barrier for adopting a new information technology (Olatokun & Busola, 2011:4). In a case like this it all comes down to a so called Rosenborg-type ‘learning by using’ process in which the employees learn how to deal with the system by working with it. Employees can also actively search for knowledge. This can be done from literature but also from other users. This creates feedback effects that promote organizational innovation (Hanel & Niosi, 2007:3). Learning from others is an important point in the next final sector on

(24)

24

interorganizational relations. The hypothese that accompanies this sector is as follows:

Hypothese 5: The skill level of workers has a positive influence on the decision of a municipality to make full use of mass appraisal methods as the best course of action available.

3.6.5 Interorganizational relations

Rogers (1995:381) describes this factor as the degree to which organizations are linked to each other by interpersonal networks. If an organization has a higher network connectedness this will result in new ideas flowing more easily into the organization. On this statement other researchers found that variations in knowledge between organizations might cause misunderstandings and deter the adoption of a new information technology (Jordan, 2015:136). Another study, not directly related to the adoption of IT within governmental organizations but certainly a noted peace in the field of research, is the study by Coleman, Katz and Menzel (1966) which studied the medical innovation adoption. They found that communication between doctors influenced the individual adoption of an innovative antibiotic by doctors. In this way the study highlighted the important role of connectedness during the process of the adoption of a new information technology (Kee, 2017:2). As already mentioned, this factors has a strong relationship to the factor of learning because it might stimulate this factor (Dewett & Jones, 2001:334).

Hypothesis 6: The presence of interorganizational relations between municipalities has a positive influence on the decision of a municipality to make full use of mass appraisal methods as the best course of action available.

3.7 Brief overview of the theory

The bottom line of this chapter is that the diffusion process spreads technology over time, but the already existing internal characteristics of an organization affect the adoption of these changes. The final impact and potential form of new technologies in governmental organizations is not set on forehand. This bottom line was caught in a descriptive and explanatory part. The descriptive part focused specifically on the diffusion of new technology and the theories that are developed on this. Whereas the explanatory part examined theories on the adoption of new information technology within governmental organizations.

(25)

25

The topic of the next chapter is the research design and data collection of this research. The hypotheses will be conceptualized and operationalized here. In the fifth chapter the hypotheses are empirically tested and the evidence out of this process will be discussed. The conclusions follow in the final chapter.

(26)

26

Chapter 4./ Research design and data collection

Out of the theoretical discussion there was developed an interview guide (Appendix) in order to collect the desired data on the diffusion and adoption of mass appraisal methods within Dutch municipalities. Thirteen professionals, representing the sides of user, developer/researcher, supervising authority and interest group where interviewed. As the greatest added value of this research lies on the explanatory part, two third of the interview questions were focused on this part. The division of the research hypotheses assumed a comparable reasoning. There was gathered additional quantitative data from a public database on the characteristics of the diffusion of mass appraisal methods between 2005 and 2017. So, in this chapter the conception and methods for measuring the variables are explained. After this, there is explained how the interviewing partners where selected and specific points on the method of data gathering are brought forward, together with experiences from this process. The chapter ends with a discussion on the limitations of the research design and data collection, finished with a paragraph on the possible threats to inference with regard to the research results.

4.1 Concepts and operationalization

Both qualitative and quantitative studies can measure the innovative adoption of new information technologies. Using a qualitative manner provides the possibility to focus on one or some cases and conduct semi-structured interviews. In these kind of studies researchers may measure innovative adoption via analyzing questions on “how” and “why”, for example why certain agencies are more innovative than others (Demircioglu, 2016:4). Also by Roulston (2010:60-63) is argued that the selection of multiple viewpoints and different backgrounds, which is the case in this research with the selection of the stakeholders into four divisions, does not allow the researcher to conduct a strictly structured interview. A semi-structured interview extended with asking expansive questions on the direction the interview leads to helps to direct the interviewee to talk about topics in which he or she has expertise. This research is therefore based on a qualitative approach, conducting semi-structured interviews.

The conception of the dependent variables diffusion and adoption was already formulated in the previous chapter on theory. Translated into this specific research this means that diffusion is the spread of mass appraisal methods across municipalities, whereby specifically is focused

(27)

27

on the element of time in this diffusion. The concept of adoption is the decision of a municipality to make full use of mass appraisal methods as the best course of action available. However, as already came forward in the theoretical part of this research, there are no consistent definitions of the measures of the used concepts. An overview of them can be found in figure 3 and 4. Wisdom, Chor, Hoagwood and Horwitz (2013:480) suggested that this lack of precise definitions and ways to measure the concepts point out the need for further work on the adoption of innovations in order to increase our understanding of it. For the concepts on the influence of time on diffusion there didn’t occur this kind of possible validity problem. These concepts are pretty straightforwardly and universally applicable through the consistent way to measure them via the three concepts of time.

Looking to the whole picture, there was till this time not determined any standard set of variables that were supposed to influence diffusion and adoption of mass appraisal methods. Therefore, the literature review on the diffusion and adoption of IT and the therefrom following theoretical frameworks are of great importance for this study. These theoretical frameworks culminate in conceptual frameworks which are presented as a schema (figure 3 and 4). These conceptual frameworks include all the relevant concepts and are schematic renderings of the hypotheses that followed out of answering the theoretical questions in the previous chapter. In the analysis there will be considered if these hypotheses lead to a satisfactory answers the two main research questions.

(28)

28

Of course it is possible to argue that there are also other variables and/or concepts influencing diffusion and adoption rather than the ones in this research. The explanations of academics are not the irrefutable truth. There is no one in the world with a monopoly on the truth, also not me. Theoretical explanations are not set in stone but are always attempts to gain a better understanding of the reality. The current available literature on the research topic is extendedly analyzed in order to come up with the most complete possible frameworks that lead to the most valid results in the end. To include all possible factors, related to adoption, in

this research would be simply unfeasible. For example elements of the technology itself or

other external factors could also have a significant influence on the diffusion and adoption of mass appraisal methods. However, including more factors would almost certainly lead to a more quantitative approach. But this would not lead to the desired results for this specific

research. Questions on “how” and “why” could not have been asked and only positive or

negative relations could be confirmed or disconfirmed on an abstract level. Also the

interviewing time of an hour per person would be too short to ask all the necessary questions

and I am thereby convinced that by deducting multiple semi-structured interviews, possible very important factors, outside the main research scope of time and the internal organizational characteristics, automatically come to the surface. If this was the case, these factors and their explanation can be red in the analysis of this research.

Operationalization means that the concepts from the theoretical framework are made suitable for empirical research. The empirical research is carried out through conducting semi-structured interviews. As may be expected, the topics for the interviews are driven by the theoretical discussion and the framework that flew out of this discussion. Let’s start with the operationalization of the concepts on diffusion. In the theoretical framework of diffusion was mentioned that the concept of innovation-decision should be analyzed before the other two concepts (innovativeness and rate of adoption), in order to be better able to tell something about the diffusion of mass appraisal methods. The three types (optional innovation-decisions, collective-innovation decisions or authority innovation-decisions) form the conception of the innovation-decision. This was empirically researched in the interviews by asking the question what the most crucial point in the development of mass appraisal methods in the Netherlands was to the opinion of the interviewed person. In this way there is gathered insight in what the starting point of the use of mass appraisal methods was. One of the three types of innovation-decision could come to the surface through this manner.

(29)

29

The second concept is innovativeness. This is the degree to which an organization is relatively earlier in adopting new ideas than other members of a social system (Rogers, 1995:37). In the empirical research there is made an analyses of publicly available quantitative data which contains information on the mass appraisal methods every municipality in the Netherlands used between 2005 and 2017. Beside to this operationalization of the concept, there was in every interview asked if the interviewed person was able to define different periods in the spreading of mass appraisal methods (AVM’s/CAMA’s) within the Netherlands. This question and accompanied data were also used in answering the concept called rate of adoption. For this concept the cumulative number of adopters needs to be plotted over time in a curve. This curve then shows the rate of adoption, which is the relative speed with which the mass appraisal methods were adopted by Dutch municipalities. The interview question thereby tries to give insight in the speed of the adoption by using the time spans of the different periods that were determined by the interviewed persons.

Before the switch was made to questions on the concepts regarding the internal characteristics that possibly influence the adoption of mass appraisal methods, there were asked several questions on the current situation regarding the use of mass appraisal methods by municipalities. This helps in order to be able to put the general research results in diffusion and adoption in a more in depth analysis in the next chapter. The first question was aimed at getting information about the current parameters which are used in mass appraisal methods and whether these methods are comparable to a certain extent. As a follow up question there was asked whether there should be developed more regulation on the construction or use of mass appraisal methods by municipalities. There was also asked whether the use of mass appraisal methods fits within the current regulation of the Special Act for Real Estate Assessment. Another type of question was if mass appraisal methods, or the producers of them, should be certified and what is the best possible way to do this.

Also the theoretical framework of the adoption of mass appraisal methods is operationalized. As already mentioned, this operationalization is less set on precise definitions and ways to measure the concepts. Size, culture, infrastructure, learning and interorganizational relations are used to tell something about the influence of the independent variable ‘internal organizational characteristics’ on the adoption of mass appraisal methods by municipalities. These concepts are conceptualized here, specifically for the case of mass appraisal methods. In the theoretical framework the concept of size was for example divided in small, medium

(30)

30

and large organizations. In this research there is only focused on small/medium and on the other side large organizations. Municipalities or administrative organizations wherein municipalities work together, with more than 50.000 houses within their municipal borders are seen as large. Municipalities or administrative organizations with less than 50.000 houses are seen as small/medium. The interview guide contained three questions on this concept. There was asked to the opinion of the interviewed person on more and more municipalities working together in so called administrative organizations. The second question focused on the possibility of having one overreaching type of mass appraisal method that valuates all the real estate within the Netherlands or certain region’s and as third question there was asked which down- and upsides would this might bring. The goal of these questions was to discover what the consequences are of working in larger organizations in relation to the used mass appraisal method. This to be able to confirm or disconfirm the hypothesis on this concept.

A second concept related to the internal characteristics of an organization is culture. This is an abstract concept that cannot easily be measured. To get empirical insight in this concept, the first question that was asked was if there can be spoken of a certain change in organizational culture by adopting mass appraisal methods within municipalities. The follow up question was focused on getting the opinion of the interviewed people about the use of these mass appraisal methods by municipalities in the Netherlands. Infrastructure is to be measured more straightforward than the somewhat abstract one of culture. The concept of infrastructure seeks to find out whether preexisting structures determine the form and use of mass appraisal methods. In the interview guide this was operationalized by incorporating the question what the motivations for a municipality are to choose a specific kind of mass appraisal method. Thereby there were incorporated additional questions on a possible changes that might have occurred in organizational structure due to the adoption of mass appraisal methods in the organization.

The concept of learning and interorganizational relations are to a certain extent related to each other. Both concepts try to get insight in to which extent people search for knowledge in order to be better able to work with mass appraisal methods. The questions for the hypothesis on learning started with asking whether there is enough knowledge within the Netherland and/or the municipalities to innovatively work with mass appraisal methods. If this was not the case, a follow up question need to be asked on how this situation could be further improved. The set of questions for the concept of interorganizational relations contains three questions. The

(31)

31

first obvious question to ask is whether the municipalities share knowledge with each other in applying mass appraisal methods. To get more insight in possible other ways of knowledge sharing there was asked how the suppliers of mass appraisal methods working together with their counterparts. As a last question of this concept there was asked how the Netherlands, regarding the application of mass appraisal methods, performs in international perspective.

The final question of the interview guide was aimed at possible future developments. Beside the development of CAMA’s and AVM’s, there was asked if there are other techniques within the Special Act for Real Estate Assessment that contribute to higher innovation and will play an important role in the further development of real estate valuation. This could provide in some input regarding possible future recommendations related to this research.

4.2 Data collection

For this specific research it is extremely important to interview people that have a wide experience with the Special Act for Real Estate Assessment and the use of mass appraisal methods within it. As this research tries to get insight in the whole development of mass appraisal methods over more than 20 years from the start of the Act till now. Therefore there was decided to approach people in senior and leading positions in organizations because it was expected that they would have the best ‘helicopter view’ of the field and already worked for the longest period of time in the sector. So, this was an important condition to be met. To be ensured of a good representivity of the interviewed people over the different parts of the research field there was decided to divide the group in four sides: user, developer/researcher, interest group and supervising authority. For the specific hypothesis on the size of the municipalities it was important to speak with employees from organizations that handle more than 50.000 objects. It was thereby also important to speak at least with one provider of CAMA as well as at least one provider of an AVM. By doing this, all possible different opinions on the research topic where encapsulated. The different groups are now shortly introduced.

User

This group contains employees from municipalities or municipal organizations with more than 50.000 properties within their municipal borders and it need to be persons that use or did use a CAMA or AVM in practice. This group can provide information on the experiences around the practical use of mass appraisal methods and are an important source when it comes

The diffusion and adoption of Mass Appraisal methods by municipalities in the Netherlands: A qualitative study on CAMA’s and AVM’s within the Special Act for Real Estate Assessment