rational, calculated decisions regarding costs and benefits but are under
the infl uenc
fl
e of social context criteria that together form the concept of
institutions. Hence, institutionalism emphasizes the persuasive control
over practices of individuals or organizations under the institution’s sway
(King et al., 1994). Persuasion can be achieved not only through directives,
but also through more gentle means like deployment of specifi c
fi knowledge,
subsidies of activities deemed ‘appropriate’, standard-setting and raising
awareness (King et al., 1994).
Various authors have explicitly or implicitly analyzed vertical channels
of communication (emphasizing activities undertaken by actors outside the
set of potential adopters) through which persuasive control over adoption
of innovations is exerted. Adoption at any time is supposed to be depen-
dent on the number of potential adopters that has yet to adopt the innova-
tion and prior adopters do not infl uenc
fl
e potential adopters. Thus, adoption
begins rapidly and slows down as the number of adopters increases. The
formal description of this model is presented in Table 14.2.
As opposed to vertical channels of communication and persuasion,
Rogers (1995) has identified horizontal channels of communication and
persuasion between potential adopters through which innovations are
promoted through processes of mimicking. Innovation by means of mim-
icking is likely to occur under the conditions that the innovations are
socially visible (Mahajan & Peterson, 1985); causes, conditions and con-
sequences are known (absence of causal ambiguity); and the success of
the innovation is unlikely to be determined by path dependencies (Loh
& Venkatraman, 1992). Adoption at any time in this line of reasoning
is related to the number of adopters, as well as the number of potential
adopters (see Table 14.2).
Bass (1969) has also identified a mixed-infl uenc
fl
e model as a rival model
to both the internal as well as the external model and in which adoption is
both determined by vertical as well as horizontal channels of communica-
tion and persuasion. The formal description (Table 14.2) yields an asym-metrical S-shaped adoption function in which external influenc
fl
e results in
more rapid early adoption than with imitation alone.
Diff
ffusion of Personalized Services 187
Table 14.2 Summary and Formal Descriptions of Three Rival Diffusion Mo
ff
dels
Labels
Formal Description of Model
External infl uence (Mahajan
fl
dN /d = p[m – N ]
t
t
t
& Peterson, 1985; Mahajan,
which (after integration) equals to the adoption
Muller & Bass, 1990)
function:
N = m[1 – e-pt]
t
N : cumulative number of adoption at time
t period t
p: coefficient of external infl
ffi
uence (p>0)
fl
m: number of potential adopters (m>0)
Mixed infl uence (Bass, 1969)
fl
dN /d = [p + q.N ][m – N ]
t
t
t
t
which (after integration) equals to the adoption
function:
N = m[p(m – m )/(p + q. m )].e-[(p + qm).t].
t
0
0
[1 + [q(m – m0)]/(p + qm )]. e-[(p + qm).t]]–1
0
N : cumulative number of adoption at time
t period t
p: coefficient of external infl
ffi
uence (p>0)
fl
q: coefficient of internal infl
ffi
uence (q>0)
fl
m: number of potential adoptersm
0: number of adopters at t=0
Internal infl uence (Mahajan
fl
dN /d = q.N [m – N ]
t
t
t
t
& Peterson, 1985), ‘word of
which (after integration) equals to the adoption
mouth’ diff
ffusion (Wang &
function:
Doong), imitation (Loh &
Venkatraman, 1992),
N = m / (1 + ([m – m ]/m ). e-qmt)
t
0
0
institutional isomorphism
N(t): cumulative number of adoption at time
(DiMaggio & Powell, 1983)
period t
q: coefficient of internal infl
ffi
uence (q>0)
fl
m: number of potential adoptersm
0: number of adopters at t=0
Additionally, according to the so-called Scandinavian Institutionalism, inno-
vations can be viewed as “ideas” as much as they can be viewed as artifacts. In
order for ideas (such as “personalization”) to spread (either through internal or
external infl uenc
fl
e), they must be translated into a success story or tale. During
the travel, the idea itself is likely to change (Czarniawska & Sevon, 2005). As
such, the idea of translation is a much more complex change than the notions
of “mimicking” and “direction” suggest, and it adds to the diffusion literature
the notion that diff
ffusion is an intricate social process that involves translation
activities of experts, boundary spanning agents and knowledge brokers. In
188 Vincent
Homburg and Andres Dijkshoorn
the analysis of his paper, we attempt to provide a diffusion model that takes
the above notions of change into account.
4 METHODS AND DATA
In order to explain the diff
ffusion of personalized e-government services among
relatively autonomous Dutch municipalities, we employ two methods.
First, we fi t t
fi
hree quantitative diffusion-of-innovation models (see Table
14.2) for the purpose of comparing and specifying relevant communication and persuasion channels in the adoption of personalized e-government services (phase 1 of the study). The data that are used in the analysis have been
extracted from a larger data set that was commissioned by the Dutch Ministry
of the Interior and composed by the “Government has an answer” program
committee. The data set covers e-government characteristics in the time frame
2006–2010.2 The fi t
fi ting procedure requires time series of a minimum of five
fi
consecutive observations (Mahajan & Peterson, 1985), a condition to which
our data satisfy and were performed using basic statistics software.3 The analytical procedure is as follows: (1) parameters of alternative models are esti-
mated; (2) all models are tested against the null hypothesis that diffusion is
a random event (White Noise), and (3) remaining models are contrasted to
determine the best diff
ffusion model (Wang & Doong, 2010).
Second, in line with our objective to further extend the e-government
body of knowledge, we adde
d a phase 2 of the study and analyzed adoption
processes in more detail in ten selected municipalities, fi
five early adopters
and fi
five laggards, selected from the data set described above. As the e-gov-
ernment literature consistently reports city size as being a major determi-
nant of e-government adoption in general (see Table 14.1), we selected both the adopters as well as the laggard from substrata of the population, based
on city size. In each of the selected municipalities, qualitative interviews
were held with key stakeholders using a topic list. Responses were recorded,
transcribed, and analyzed4 using back-and-forth coding techniques (Miles
& Huberman, 1994).5 The categories resulting from the coding techniques in the selected municipalities allowed us to compare characteristics of both
adopters with non-adopters in various, and through induction to explain
diff
ffusion of personalized e-government services.
5 ANALYSIS: EXPLAINING THE DIFFUSION
OF PERSONALIZED E-GOVERNMENT
5.1 Description of Personalized E-Government
Services in Dutch Municipalities
Table 14.3 lists the prevalence of attributes of personalized electronic service delivery by Dutch municipalities in the years 2006, 2007, 2008, 2009
and 2010.
Diff
ffusion of Personalized Services 189
Table 14.3 Prevalence of Personalization Attributes in Dutch Municipal
E-Government Services
2006
2007
2008
2009
2010
(n=458) (n=443) (n=443) (n=441) (n=418)
DigiD authentication
20.7%
56.7%
76.3%
88.2%
94.6%
Personalized newsletter
16.4%
21.2%
21.2%
N/A
27.9%
Tracking & tracing
10.0%
16.0%
28.2%
26.5%
41.3%
Payment
15.9%
42.4%
61.4%
80.0%
91.6%
Pre-completed forms
N/A
N/A
17.8%
19.1%
33.9%
Personalized counters (MyGov.nl)
5.2%
14.2%
23.7%
28.8%
40.9%
Personalized policy consequences
N/A
N/A
19.4%
18.7%
22.2%
5.1.1 Phase 1: Models of Diffu
ff sion
To determine which infl uenc
fl
e model best explains adoption, we fi
fit each of
the three models described in Table 14.2 using an iterative non-linear sum of squared residuals regression analysis6 and apply it to the time series of prevalence of personalized counters (see Table 14.4).
As all R2 indicate a reasonable fi
fit, and p and q estimates are all positive,
additional procedures must be taken into account as to compare alterna-
tive diff
ffusion models. In a pairwise comparative test, if one or more of the
alternative models fail to reject the White Noise model, there is no need to
proceed further (Mahajan & Peterson, 1985). From Table 14.5 it can be concluded that all three rival models can reject the null hypothesis (which
states that diff
ffusion is a random event).
Table 14.4 Parameters for Best Fit for E-Government Personalization Adoption in
Municipalities in the Netherlands 2006–2010
Influ
fl ence Model
External
Internal
Mixed
p 0.11
-
0.079
q -
0.04
0,000
Adjusted R2
0.96
0.94
0.50
Table 14.5 Model Comparisons against White Noise Model
Alternative Models
Internal Infl uence
fl
Mixed Infl uence
fl
External Influ
fl ence
H : White noise
t = 3.116 (p<0,05)
t = 2.830 (p<0,05) t = 2.879 (p<0,05)
0
190 Vincent Homburg and Andres Dijkshoorn
Table 14.6 Model Comparisons among Alternative Diffusion Mo
ff
dels
Alternative Models
Internal Infl uence
fl
Mixed Infl uence
fl
External Infl
fluence
H : Internal infl uence
fl
-
t = .89
t = 1.10
0
(p=.438)
(p=.349)
H : Mixed infl uence
fl
t = -0,05
-
t = .394
0
(p=.962)
(p=.72)
H : External infl uence
fl
t =-.439
t = 0,113
-
0
(p=.69)
(p=.917)
This leaves us the task of determining which of the three alternatives,
if any, is the model that best explains diffusion. In order to determine the
best explanation, the P-test is used, which determines the truth of H in the
0
presence of an alternative model H . In any one of the paired confronta-
1
tions, if is statistically no different from zero, then H is the true model.
0
From the results of the P-test reported in Table 14.6 and given the quite small sample, we cannot decide on a “winning” fi
fitting model. Based on the
values in Table 14.4, we infer that horizontal and v
d ertical channels of com-
munication and persuasion can be identified in the diff
ffusion of personal-
ized e-government in the time frame 2006–2010 in the Netherlands.
5.1.2 Phase
2:
Qualitative Field Work
To further analyze the process of communication, persuasion, and adoption
beyond the issue of relevant channels, we compared experiences and consid-
eration of fi v
fi e “adopters” (municipalities off er
ff ing personalized electronic ser-
vices as of 2008) with experiences and considerations of fi ve
fi “non-adopters.”
5.2 Pressure on Adoption Decisions
Respondents in municipal organizations reported perceived expectations
of citizens as the most important source of infl uenc
fl
e on adoption decisions
regarding personalized e-government services. As one alderman phrased it:
a clamor for service provision, less bureaucracy, transparency: that is
external pressure, as I perceive it. (. . .) Just because society does not
tolerate other kinds of organizational behavior. (Alderman)
Another kind of infl u
fl ence that was mentioned quite frequently was
the existence of benchmarks with which the presence of municipalities is
exposed. As a manager of service provision explained:
To score is felt to be important among municipalities. How often is
your municipality being mentioned in professional j
ournals, are you
Diff
ffusion of Personalized Services 191
Table 14.7 Sources of Pressure
Source of Pressure
Frequency
Citizen demand
121
Benchmarks
88
Legislation
81
National initiatives
80
Peer rivalry
5
in the Top 3. . . . that is considered to be very important. (Manager
of service provision)
The fact that municipalities keep a sharp eye on benchmarks and rankings
sometimes results in somewhat perverse incentives to adopt personalized
services, as one respondent reported.
Our decision to implement personalized service delivery was due to our
low ranking . . . Our alderman wanted to improve our ranking, and we
found out that we could improve our ranking quite easily by implement-
ing a Personalized Internet Page . . . and so we did. (Project manager)
Table 14.7 lists reported sources of pressure, including legislation (not as a direct source of infl uenc
fl
e, but for instance national environmental legislation
that instigates municipalities to issue one permit covering a variety of condi-
tions stemming from various acts) and national outreach activities. Together,
these sources indicate that (institutional) pressure aff ects
ff
adoption in line with
existing literatures on isomorphic pressure on adoption of innovations.
5.3 Organizational Search
One consequence of institutional pressure as reported by respondents is
that municipalities, once confronted with pressure, start scanning their
environments for relevant knowledge and experiences (see also Levinthal
& March, 1981). As one respondent indicated:
One member of our support staff m
ff
ade an inventory of associations
staff m
ff
embers are participating in, and she managed to compile a list
of three or four pages. (Manager of service provision)
Respondents reported that pressure did not directly result in new con-
nections with other organizations, but rather that organizational pressure
resulted in more intensive contact with forums and associations (for instance,
the Public Service Provision Managers’ Association, the Association of Dutch
Municipalities, but also outreach programs like GovUnited) one was already
participating in.
192 Vincent
Homburg and Andres Dijkshoorn
Table 14.8 Organizational Search
Organizational search
Frequency
Forums & outreach programs
65
Companies 62
Alliances of municipalities
23
Table 14.8 lists the type of associations, programs, and alliances that are reported by respondents as sources of ideas, knowledge, and solutions. Forums
Public Sector Transformation Through E-Government Page 33