The Adoption of The Logical Framework in International Development Projects A Survey of Non-Governmental Organizations

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 11

Impact Assessment and Project Appraisal

ISSN: 1461-5517 (Print) 1471-5465 (Online) Journal homepage: www.tandfonline.com/journals/tiap20

The adoption of the logical framework in


international development projects: a survey of
non-governmental organizations

Ruggero Golini, Paolo Landoni & Matteo Kalchschmidt

To cite this article: Ruggero Golini, Paolo Landoni & Matteo Kalchschmidt (2018) The
adoption of the logical framework in international development projects: a survey of non-
governmental organizations, Impact Assessment and Project Appraisal, 36:2, 145-154, DOI:
10.1080/14615517.2017.1354643

To link to this article: https://doi.org/10.1080/14615517.2017.1354643

Published online: 28 Jul 2017.

Submit your article to this journal

Article views: 4207

View related articles

View Crossmark data

Citing articles: 1 View citing articles

Full Terms & Conditions of access and use can be found at


https://www.tandfonline.com/action/journalInformation?journalCode=tiap20
Impact Assessment and Project Appraisal, 2018
VOL. 36, NO. 2, 145–154
https://doi.org/10.1080/14615517.2017.1354643

The adoption of the logical framework in international development projects:


a survey of non-governmental organizations
Ruggero Golinia , Paolo Landonib and Matteo Kalchschmidta
a
Department of Management, Information and Production Engineering, Università degli Studi di Bergamo, Dalmine, Italy; bDepartment of
Management and Industrial Engineering – Politecnico di Torino – Corso Duca degli Abruzzi, Torino, Italy

ABSTRACT ARTICLE HISTORY


Non-governmental organizations (NGOs) play a fundamental role in International Development Received 30 January 2017
but their projects often lack efficiency and effectiveness. A key aspect is the proper use of project Accepted 31 May 2017
management tools and methodologies. The Logical Framework (LF) has been suggested as a KEYWORDS
key tool in this sector. However, the extent and impact of its use by NGOs is still not known. Logical Framework;
Based on almost 500 questionnaires collected from project managers working on International international development;
Development (ID) projects in NGOs around the world, we show that the LF is broadly adopted, NGO; impact assessment
especially for large projects and ‘soft’ projects. Moreover, we demonstrate that when used in the
various phases of a project’s life-cycle, the LF contributes significantly to the achievement of
superior project impact.

1. Introduction namely the Logical Framework (LF). The LF is a matrix


(normally a 4 × 4 matrix) that summarizes the project’s
Compared to emergency projects, which provide goals, activities, assumptions, indicators, and sources of
immediate assistance to populations hit by wars or nat- verification in order to measure and report the achieve-
ural disasters (Durning 2014; Gore and Fischer 2014), ment of objectives (Coleman 1987; Gasper 1997; Akroyd
International Development (ID) projects aim to long-term 1999). An example of the LF is reported in Figure 1.
improvements for the beneficiaries in terms of health, Despite its importance and widespread use by gov-
education or living conditions in general. ID projects ernmental organizations (Landoni and Corti 2011), its
attract relevant flows of public funds: in 2013 about $ level of adoption by NGOs is still unknown. Moreover,
135 billion have been poured into international develop- as we will explain in the literature review, this tool has
ment by DAC member countries.1 However, also private been subject to several criticisms, so that its actual effec-
funding, even if difficult to be estimated has a relevant tiveness should be further investigated. Accordingly, on
share. For instance, in 2010, private funding for human- the basis of a large-scale survey, the goal of this paper
itarian assistance was estimated to be about one-third is to shed light on the following aspects: the extent to
of the total ($ 5.8 billion on a total 18.22). Nevertheless, which the Logical Framework is adopted by NGOs; the
empirical and anecdotal evidence (e.g. TED Talks3) and factors that induce its adoption; and its impact on project
recent researches (Lovegrove et al. 2011; Zimmermann performance.
and Smith 2011; Ika and Saint-Macary 2012) show that The rest of the paper is structured as follows. First, we
the success rate and long-term impact of such projects is provide a review of the literature on the LF, highlighting
often below the expectations. This glitch is often associ- the gaps and thus formulating our research questions.
ated to a lack of a professional managerial approach and Next, we illustrate the methodology used to answer
understanding of the context (Ahsan and Gunawan 2010; them, and then set out and discuss the results. Finally,
Ika 2012), especially in NGOs (Newcomer et al. 2013). As a we draw the main conclusions and highlight the princi-
consequence, several authors have called for additional pal limitations and future developments of this research.
research (Youker 1999; Golini and Landoni 2014), in par-
ticular regarding project appraisal tools and approaches
2. Literature review
able to adapt the best practices in project management
with the complex context these projects have to deal ID projects have specific characteristics such as the pres-
with (Ebrahem 2003; Khang and Moe 2008). ence of a large number of different stakeholders in terms
In response to this call, in this paper we focus on one of interests, power and culture (Youker 1999; Saad et al.
of the most specific tools developed for ID projects, 2002; Diallo and Thuillier 2005; Zhai et al. 2009; Steinfort

CONTACT Ruggero Golini ruggero.golini@unibg.it


© 2017 IAIA
146  R. GOLINI ET AL.

Project Description Indicators Sources of Verifications Assumptions gathered must be reviewed and, if necessary, revised.
Overall Objectives After completion of the first column (Hierarchy of
Purpose
Objectives), the second step consists in identification
of the fourth column of the matrix (Assumptions).
Results
Assumptions are usually identified during the project
Activities planning phases through analysis of stakeholders,
objectives and tasks. When the project description and
Figure 1. Example of logical framework. assumptions have been drafted, the next step is to iden-
tify indicators that may be used to measure and report on
the achievement of objectives (column 2) and the sources
2010) that cope generally with instable environments of information (column 3). The indicators describe the
(Youker 1999; Diallo and Thuillier 2004; McCarthy and project’s objectives in operationally measurable terms
Zakrajšek 2007; Khang and Moe 2008; Zhai et al. 2009; Ika (quantity, quality, time).
et al. 2012) and with the presence of intangible project Use of the LF can provide several advantages in the
outputs (e.g. improvement in living conditions or educa- various phases of a project by remedying many of the
tion) (Khang and Moe 2008; Ahsan and Gunawan 2010; limitations of other tools (Gasper 1997; Bell 2000). In the
Steinfort 2010; Ika et al. 2012) making these projects first phase (initiation), it helps to achieve a better problem
particularly challenging to be managed and evaluated, analysis and definition of objectives through systematic
especially in their social impact (George 2001; Esteves et analysis of problems and by placing the project within
al. 2012; Golini et al. 2017). a broader context. Next, it can be used in the planning
In 1970, Baum introduced a specific approach for ID phase to set objectives, establish measurement systems,
projects based on the ‘project cycle’, and specific tools and identify and manage risks. Furthermore, it can be
(e.g. the Logical Framework) were developed in the used in the execution and control phase to monitor pro-
following years, see e.g. Coleman (1987). In recent dec- gress and manage scope changes and risks. Finally, it can
ades, the project cycle approach has been adopted by be used in the concluding phase to check, measure, and
development agencies because it offers a structured report the project’s achievements. The LF can also pro-
framework of techniques that enables people to work vide several organizational advantages. First, it helps to
together while bringing the project’s objectives into define different levels of management responsibility and
focus (Ahonen 1999; Landoni and Corti 2011). However, gives greater accountability to the project manager for
while the project cycle concept is shared also by busi- the risks and the results. Second, it allows better stake-
ness-oriented project management guidelines (e.g. the holder involvement and communication by its simple
PMBOK Guide® by the Project Management Institute), and clear format. Given all these advantages, the LF is
some of the tools, like the Logical Framework, have extensively adopted not only in governmental agencies,
remained confined to the international development but also in NGOs (Golini et al. 2015). However, despite its
field. This has led to the scant integration of standard many advantages, the LF has been also significantly crit-
project management tools with the Logical Framework icized (Gasper 1997; Crawford and Bryce 2003; Couillard
that causes issues in the management of ID projects et al. 2009; Ika and Lytvynov 2011). The main limitations
(Golini and Landoni 2014). To gain better understand- reported by the literature are: oversimplification, deter-
ing of these issues, in this paper we focus on the Logical ministic and time-invariant logic, lack of flexibility, lack
Framework, its diffusion, and its actual contribution to of accountability, unclear terminology, and lack of inte-
project performance. gration with other project management tools. In particu-
The origins of the Logical Framework in the context lar, the LF is considered to be sometimes too simple to
of ID projects can be traced back to 1969, when two represent the complexity of a project, and it may omit
US-based consulting firms (Fry Associates and Practical some important elements (such as cultural aspects or
Concepts, Inc.) developed it for the United States Agency the objectives of all the stakeholders). Secondly, risks or
for International Development (USAID) (Solem 1987). variability in the variables (either positive or negative) are
Thereafter, the LF was adopted by many other gov- not included and the underlying logic is rather determin-
ernmental agencies around the world (Landoni and Corti istic. Another issue is the difficulty of defining the objec-
2011). Today, there are many versions and variations in tives: ID projects are usually complex, i.e. many internal
the terminology used, but the Logical Framework main- and external variables must be considered. Hence, the
tains its typical structure of a matrix that breaks a pro- objectives are particularly difficult to identify and formal-
ject down into its component parts in order to facilitate ize. Moreover, given the high number of stakeholders
management (Cracknell 2000). One of the best-known involved, with their complex relationships and differ-
versions of the tool is probably the one adopted by the ent interests, the project’s objectives may be mixed or
European Commission. conflict with individual objectives. Similarly, changes in
Preparation of the LF is an iterative process. As new projects may give rise to a difference between ‘official’
parts of the matrix are completed, information previously and ‘actual’ objectives.
IMPACT ASSESSMENT AND PROJECT APPRAISAL  147

Given the complexity of ID projects, it is likely that, Ketokivi and Schroeder 2004; Zotteri and Kalchschmidt
in the course of a project, several changes will occur in 2007; Kalchschmidt 2012), and on this basis we defined
terms of activities and in terms of the desired outcomes. specific variables. Specifically, we considered three main
The LF appears to be rigid to the objectives and activities groups of variables: first NGO objectives, then project
stated at the outset because it lacks a ‘time perspective’. characteristics, and finally the external context.
Furthermore, the LF has been criticized because it does Therefore, our second research question concerned
not always straightforwardly identify a causality effect the drivers that induce greater adoption of the LF:
among its levels. Moreover, relationships may change RQ2: How do NGO’s objectives, project characteristics,
during the project’s execution. Also, unclear terminology and contextual factors influence the adoption of the
can cause problems: many users may find it difficult to LF?
understand the differences among the various levels (e.g. Finally, the existence of advantages and disadvan-
purpose, goal, and outcome). Another issue is the lack of tages to the LF raises the question of whether it posi-
integration with other project management tools, since tively contributes to a project’s performance. Hence our
PCM and LF do not substitute traditional project man- last research question was:
agement tools (e.g. work breakdown structure, Gantt RQ3: How does the adoption of the LF affect perfor-
chart). A final limitation is the lack of accountability: the mance (impact) of the project?
LF does not show who is responsible for what. The three research questions can be summarized in
All these potential problems may induce NGOs to use the theoretical framework shown in Figure 2. The fig-
the LF only when necessary or compile it because it is ure highlights that the main purposes of our research
mandatory, but not using it as a key project management were to verify the level of adoption of the LF (RQ1), to
tool, thus missing its potential. Because of that, beside identify what the related drivers were (RQ2) (i.e. context
the analysis of how spread the LF is among NGOs, we variables, project variables, and project objectives), and
aim to understand its extent of adoption as a key tool in to determine how the performance indicators were
the different project phases. Therefore, the first research affected (RQ3).
question is:
RQ1: To what extent is the LF adopted as a key project
management tool by the NGOs to manage ID projects? 3. Methodology
The LF is considered one of the critical tools needed To address these questions, we adopted a survey
to fulfill the project objectives and ensuring long-term approach. This approach was consistent with our
sustainable impact. Still, the simultaneous presence of research goals since we could leverage on already
pros and cons in the use of the LF, led us thinking that defined constructs and we were interested in a widely
its adoption can vary according to different situations. diffused methodology. We decided to design a survey
From this point of view, we took a contingent approach tool to cover project managers working for NGOs in the
on analyzing the extent of the LF’s adoption. Specifically, development sector throughout the world and thus cre-
there is no theoretical contribution stating what contex- ate a global data-set. The questionnaire was designed to
tual elements may influence LF adoption, but this could obtain quantitative, well-comparable, and manageable
be very relevant for research and practice. By under- answers. Besides English, the survey was also translated
standing the drivers, in fact, it is easier to identify the into Spanish (using a double parallel translation proce-
contexts in which the LF is less adopted and which can dure). Therefore, our results were not restricted to a spe-
be enabling factors. For this reason, we drew mainly on cific geographical area that might have specific cultural
the literature that has considered contingency theory and context biases. This improved the generalizability
in other fields of research (e.g. project management, of the results. To control for the quality of the results,
operations management, organizational theory) (Powell we selected only respondents with at least two years of
1995; Delery and Doty 1996; Sousa and Voss 2001, 2008; expertise in managing projects.

Figure 2. Variables included in this study and their relationship.


148  R. GOLINI ET AL.

Project managers operating in NGOs were contacted Table 1. Descriptive NGO statistics.
directly using information found in public databases Tot. Income
(mainly the Directory of Development Organizations4) Continent % N. Employees % (USD) %
and associations of NGOs (e.g. MESA de articulation,5 Africa 26 1–20 54 <1.000 15
America 11 21–50 17 1.000–10.000 6
CONCORD6). It is important to remind that the organi- Asia 20 51–100 10 10.000–100.000 12
zations listed in these databases do not necessarily have Europe 40 101–500 9 100.000–1.000.000 25
Oceania 2 >500 5 >1.000.000 26
ID projects in their portfolio. Because of this, we clearly Missing 1 Missing 5 Missing 16
stated at the beginning of the questionnaire that the Total 100% Total 100% Total 100%
survey was aimed to organizations dealing with ID pro-
jects and, to make it sure, we also introduced a control
question in the survey. Table 2. Descriptive project statistics.
Contacts with NGOs’ networks were very important Average
duration
to increase the size of the sample and to avoid sample Principal aim % (months) % Average size (USD) %
biases. NGOs were contacted by email and phone, when Delivery and 2 0–6 14 <100.000 38
possible, to increase the response rate. Contacts were integration
of tools
searched and managed globally to avoid limiting our Construction 6 7–12 29 100.000–200.000 24
considerations to a specific area or country. The survey of infra-
was administered in 2012 and the data-gathering pro- structure
Development 57 13–18 9 200.000–500.000 17
cess lasted for about four months. Eventually, 496 organ- of services
izations provided information useful for the research. Awareness 22 19–24 18 500.000+ 19
raising
Around 30,000 email and phone contacts were made to Other 12 24+ 24
construct the sample, and a response rate of 1.6% was Missing 1 Missing 6 Missing 2
Total 100% Total 100% Total 100%
obtained. The response rate is quite low, even if aligned
with other works published using the same approach
(e.g. Martinsuo et al. 2006). One of the reasons for the low In the following subsections, we present the meas-
response rate can be also related to the fact that, since the ures used for each group of variables (already shown in
PM discipline is quite new for NGOs, the level of interest Figure 2).
in our study might have been quite low. Given this low
response rate, we contacted about 50 non-respondents
3.1. Logical framework adoption
by telephone, but we could not identify a reason for the
non-responses (e.g. lack of time, lack of interest in the The focal variable in our study was adoption of the LF. To
research, not dealing with ID projects, the NGO secre- evaluate this, we employed two measures related to the
tary did not forward the email to the project managers). extent and depth of use. As a matter of fact, the litera-
To support the validity of our sample, we also checked ture states that not only should the LF be used in every
possible biases due to late respondents and contacts project (extent of adoption) but also that its adoption
established through different channels (e.g. email, tele- should be pervasive throughout the project (i.e. it has to
phone), finding that our sample was not affected by such be a core tool, not just a formal one) (Gasper 1997, 2000;
biases. Further analyses indicated that the low response is Smith 2000; Couillard et al. 2009). The two measures are
mainly since many of the organizations contacted did not theoretically independent of each other because a PM
actually manage ID projects (about half of the contacts). can always employ the LF only because it is mandatory
Table 1 presents the main characteristics of the NGOs (e.g. the tender or the sponsor requires it), but in this case
for which the surveyed PMs worked. its adoption may be formal and superficial.
Table 1 shows that the majority of the NGOs in our As said, the extent of adoption measure was the
sample were European. However, without considering number of times that the LF was employed in projects.
Oceania, a considerable global distribution was achieved, To measure this amount, we asked the percentage of
as envisaged in our initial goals. Most of the NGOs were projects in which they adopted the LF on a 1–5 scale
small-sized organizations, with fewer than 20 employees (where 1 = 0%; 2 = 1–25%; 3 = 26–50%; 4 = 51–75%;
and incomes that grew proportionally with size. 5 = 76–100%). As the difference between 1 and 2 in
Besides information about NGOs, the survey also pro- the scale is different from the other points (3, 4, 5), we
vided data on other descriptive variables concerning the merged 1 and 2, so the final scale is from 1 to 4.
projects directly managed by the interviewees (Table 2). The depth of relates to the how the LF was used and
Most the latter (72%) had managed fewer than 10 pro- its importance in guiding the project. To obtain this
jects in the previous two years. Moreover, 40% of the measure, we used the average of a set of indicators
projects in our sample were of small scale, with a high measured on a 1–5 Likert scale (1: Strongly Disagree – 5:
variance in terms of duration. The principal aim of most Strongly Agree) representing the importance of the LF
of them (57%) was the development of services (educa- in the various phases of a project (Table 3). These indica-
tion, finance, health care, etc.). tors were averaged into a single construct (the statistical
IMPACT ASSESSMENT AND PROJECT APPRAISAL  149

Table 3. Measures for LF adoption (all items are measured on a the social desirability effect. We then performed the
1–5 scale). Harman single factor test: by means of a factor analy-
LF adoption Cronbach’s Std. sis that took account of all the variables included in the
measures Items alpha Average dev study (LF adoption, performance and drivers), we found
LF depth of LF helps in the initial 0.741 3.94 0.852
adoption part of the planning
that they did not pertain to a single factor.
phase
LF provides guidance for
the entire project 4. Results and discussion
LF is frequently updated
during the project 4.1. Adoption of the LF as a key PM Tool (RQ1)
LF is used in the evalua-
tion phase at the end of
the project
Inspection of the statistics in Table 3 provides an imme-
LF is used in the reports diate answer to our first research question. The LF is a
and project meetings widely adopted tool, with 93.4% of our interviewees stat-
LF extent of Percentage of projects – 3.47 0.055
adoption adopting LFa ing that they used it. Moreover, the level of adoption was
a
Scale: 1 = 0–25%; 2 = 26–50%; 3 = 51–75%; 4 = 76–100%. very high, since 60% of the respondents reported using
the LF in 75% or more of the managed projects (extent
correctness of the procedure was supported by the very of adoption measure) and with a mean of 3.9 out of 5
high Cronbach’s alpha). in terms of depth of use. This analysis shows that the
situation has changed since the year 2000, when Abbasi
and Al-Mharmah (2000) pointed out the low diffusion of
3.2. Drivers and performance the LF among NGOs and confirms the results of Golini
In accordance with the literature, the drivers that may et al. (2015). A possible explanation for the LF’s wide-
induce a greater or lesser adoption of the LF were spread use relates to the criticism made by Couillard et
grouped into three categories: context variables, project al. (2009), Crawford and Bryce (2003) and Gasper (1997)
variables, and NGO objectives (Table A1 in Appendix). concerning its oversimplification and scant integration
Context variables were elements that described the con- with the other project management tools. This criticism
ditions in which projects were managed. They included was not really shared by the project managers in our
factors like the size of the NGO, the level of development sample. Using additional questions, we confirmed that
of the country in which the NGO operated, and the over- there was relatively little agreement with the criticism
all variability of the context. Because the PM and his/her of oversimplification (mean: 2.43 out of 5) and, on the
project was the unit of analysis, we included the total contrary, the respondents underlined that the LF is quite
income and size of the NGO among the context charac- easily integrated with other project management tools
teristics. Project variables referred to the characteristics (mean: 3.71 out of 5).
of projects, such as size, the number of projects typically
managed by the NGO, the number of active stakeholders, 4.2. Drivers of LF adoption (RQ2)
etc. Finally, NGO objectives concerned the attention of
the NGO to both long-term and short-term objectives. Despite the LF’s average high adoption, a substantial
Next, in line with Golini et al. (2015) a factor analysis number of PMs still do not use it to any great extent.
was used to divide the performance indicators consid- Hence, the second research question concerned the fac-
ered into two groups: the first related to time and cost tors that explain a higher adoption of the LF. Three kinds
performance, the second to the strategic project results of factors were considered: NGO objectives, project char-
(e.g. satisfaction of the local community) (Table A2 in acteristics, and the context (Table 4). We related these
Appendix). The project managers were asked to assess factors to the LF’s adoption (both extent and depth). In
these indicators on a 1–5 Likert scale (1 – ‘Very low’, 5 – Table 4, the significant factors, with a p-value <0.05, are
‘Very high’). highlighted in gray.
To answer RQ2 and RQ3, we performed several regres- Starting with the context variables, to be noted is that
sion analyses to consider the extent and depth of adop- context variability relates to the depth of use of the LF.
tion of LF and internal (i.e. the NGO’s attitude to focus on This result seems reasonable, given that in highly vari-
management and operational parameters) and external able contexts it is necessary to review and, if necessary,
project objectives (i.e. the NGO’s attitude to focus on adapt the project objectives. Hence, the logical frame-
long-term and strategic impacts). For each regression, work needs to be used and updated in all the project
we checked that multicollinearity was not present by phases. Furthermore, the size of the NGO negatively
looking at the variance inflation factor (always below affects the LF’s extent of adoption. This result seems
2.75) and the condition indexes (always below 7.24). We rather counterintuitive and requires further investiga-
also checked that the residuals were normally distrib- tion. Regarding project variables, note that the average
uted. We also controlled for common method bias. We size of the project impacts positively on the extent of
first guaranteed the respondents’ anonymity to avoid adoption. This may be related to the fact that larger
150  R. GOLINI ET AL.

Table 4. Research model (principal aims: 1: delivery and integration of tools, 2: construction of buildings and infrastructure, 3 devel-
opment of services, 4: awareness raising). Income and size were measured on a logarithmic scale.

*Sig. < 0.10; **Sig. < 0.05; ***Sig < 0.01.

projects can have greater requirements – for instance, Table 5. Relationship between LF and performance indicators.
by the donors – also in terms of the tools to be used.
However, the project size does not impact on the depth
of adoption, so that there is the risk of a formal, rather
than substantial, use of the LF, even in larger projects.
Also, to be noted is that NGOs involved in construction
projects (principal aim = 2) use the LF less in terms of
depth. In fact, construction projects have more tangible
and clearer objectives than the others. As the literature
reports, these are ‘hard’ projects (Crawford and Pollack
2004), in contrast to ‘soft’ projects, which require more
careful definition of objectives in relation to the stake-
holders. This may explain the lesser adoption of the LF by
this group. Communication of the results is significantly
related to a greater extent and depth of use of the LF.
In fact, as highlighted in the literature, use of the LF in
the various phases of the project, and especially in the
concluding one, allows better communication with the
project stakeholders. Finally, the focus on internal objec-
tives induces a higher level of depth of adoption, even if
with a relatively low effect (sig. <0.10).
*Sig. < 0.10; **Sig. < 0.05; ***Sig < 0.01.
4.3. Impact of LF on performances (RQ3)
internal objectives, the better the performance achieved
The last research question regards the extent to which
in the corresponding performance area (i.e. time–cost or
LF adoption impacts on project performance. To answer
strategic). What is most important is that the LF’s depth
this question, we performed a regression similar to the
of adoption has an impact on strategic performance,
previous one (Table 5). In this case, we ran two regression
while the extent of adoption on time–cost performance.
models to check the effects on strategic and time–cost
This result has several implications. Firstly, the LF has a
performance. We also left the drivers as control variables
positive or neutral effect on performance, but not a detri-
in the regression to avoid spurious effects in the meas-
mental one. This supports the view that the LF is a robust
ure of the LF (Baron and Kenny 1986). Compared with
and useful tool. Secondly, the adoption of the logical
the previous analysis, we omitted results communication
framework seems to ensure better operational results,
because it was not significant and created some multi-
while only a widespread use throughout the project can
collinearity issues.
lead to superior strategic performance. This highlights
First to be noted is that the control variables are sig-
the importance of providing training to NGOs to show
nificant only for the NGO objectives. As expected, the
how the LF can be a core tool to be used and updated
greater the importance given by the NGO to external or
during the various stages of the project. A result, which
IMPACT ASSESSMENT AND PROJECT APPRAISAL  151

is almost a warning, is that the LF is used by NGOs that This work provides also some suggestions to man-
are internally focused (Table 4), but its positive effects agers in NGOs actively involved in ID projects. First, we
are on strategic performance (Table 5). This means that suggest considering properly the purpose for which they
NGOs have probably not yet mastered the use of the LF, decide to adopt the LF. Besides being a useful tool, in the
so that its adoption does not yield the intended results. case of complex and large projects it may be not effec-
tive as managers may expect since main stakeholders
(i.e. donors) may expect or require more complex and
5. Conclusions
structured tools. Thus, we suggest to critically evaluate
Despite their importance and their poor performance in case by case whether to use or not the tool. Second, in
terms of project management, International Development case LF is considered a proper tool, we suggest adopt-
projects have received limited attention in the literature ing it extensively also for communicating both internally
for what concerns methods and tools. Our research, and externally the project goals and progress. Our results
thanks to a survey of almost 500 PMs around the world, suggest in fact to managers that an extensive use of LF
sheds some light on this topic with a focus on the Logical can favor coordination and control of the project, with
Framework. What emerges from the analysis is that the positive impacts on project’s performance. It is however
situation has changed from the one depicted by the LF important for managers to properly communicate the
diffusion analysis performed by Abbasi and Al-Mharmah benefits of the tool and to train people to its effective
(2000). The LF is now widespread among NGOs (93.4% of use, in particular to adopt it in conjunction with other
the respondents stated that they have used it), with a very project management systems.
high level of adoption in terms both extent (i.e. number Finally, we wish to highlight some limitations of this
of projects in which the LF is used) and depth (i.e. the LF study. Firstly, we collected a global sample but with
is used throughout the project). While an evaluation of limited focus on the differences among practices in
the extent of adoption was also presented by Golini et al. different countries. This comparison was not possible
(2015), in this paper we added the depth of adoption that since the sample did not allow comparison of what
represents an important and original contribution. This NGOs in different countries are doing. We argue that
higher level of adoption may be explained by another this analysis would be beneficial for NGOs willing to
result of our survey: the project managers did not share benchmark their operations, and that future research
the criticisms commonly made of the LF (oversimplifica- should address this issue. Secondly, the context varia-
tion and scant integration with other project manage- bles considered were derived from other studies and
ment tools). On the contrary, they considered the LF easy were not verified in the field. A case-based analysis
to integrate with other project management tools. As to would likely be of help in determining whether other
the reasons why the LF is adopted, it is used more in con- contextual variables are important in this case, thereby
texts characterized by high variability that require review making the theoretical model presented here more
and adaptation of the plan and the project’s objectives. complete and reliable.
Moreover, the LF is associated to a greater willingness to
communicate with the project stakeholders. This means Notes
that organizations willing to communicate better their
1. 
DAC Members’ Net Official Development Assistance
results are facilitated by the LF that results in an effec-
i n 2 0 1 3 ( h t t p s : / / w w w. o e c d . o r g / d a c / s t a t s /
tive tool for disclosing the achieved results. Interestingly, documentupload/ODA%202013%20Tables%20
we found that the average size of projects has a positive and%20Charts%20En.pdf ).
impact only on the extent of use of the LF. This may be 2. 
h ttps://www.globalhumanitarianassistance.org/
because larger projects are subject to higher expectations wp-content/uploads/2012/04/Private-funding-an-
by donors, for instance, in terms of the tools to be used. emerging-trend.pdf.
3. 
Ernesto Sirolli, September 2012, https://www.ted.com/
Instead, the project’s size does not impact on the depth
talks/ernesto_sirolli_want_to_help_someone_shut_
of adoption, indicating the risk of a formal, rather than up_and_listen?language=it. David Damberger, Arpil
substantial, use of the tool, even in larger projects. Finally, 2011, https://www.ted.com/talks/david_damberger_
the analysis shows that LF has a positive or neutral effect what_happens_when_an_ngo_admits_failure.
on performance, supporting the view that the LF is a use- 4. 
https://www.devdir.org.
5. 
https://mesadearticulacion.org/.
ful tool for NGOs. However, this result also shows that its
6. 
https://www.concordeurope.org/.
impact could be higher, or that other instruments could
be introduced to increase project success. This highlights
the importance of supporting the diffusion of the tool Acknowledgments
and knowledge about it, but also the need for further The authors gratefully acknowledge financial support
research on its use and limits. received by the PMI Sponsored Research Program.
152  R. GOLINI ET AL.

Disclosure statement Esteves AM, Franks D, Vanclay F. 2012. Social impact assessment:
the state of the art. Impact Assess Project Appraisal. 30:34–
No potential conflict of interest was reported by the authors. 42.
Gasper D. 2000. Evaluating the ‘logical framework approach’
towards learning oriented development evaluation. Publ
Funding Adm Dev. 20:17–28.
Gasper DR. 1997. Logical frameworks: a critical assessment:
This work was supported by ID Project Management: a survey
managerial theory, pluralistic practice. ISS Working Papers-
of nongovernmental organizations.
General Series.
George C. 2001. Sustainability appraisal for sustainable
development: integrating everything from jobs to climate
ORCID change. Impact Assess Project Appraisal. 19:95–106.
Ruggero Golini http://orcid.org/0000-0002-8632-2793 Golini R, Corti B, Landoni P. 2017. More efficient project
Paolo Landoni http://orcid.org/0000-0001-5079-1101 execution and evaluation with logical framework and
Matteo Kalchschmidt http://orcid.org/0000-0001-7997-8535 project cycle management: evidence from international
development projects. Impact Assess Project Appraisal.
35:128–138.
References Golini R, Kalchschmidt M, Landoni P. 2015. Adoption of
project management practices: the impact on international
Abbasi GY, Al-Mharmah H. 2000. Project management practice development projects of non-governmental organizations.
by the public sector in a developing country. Int J Project Int J Project Manage. 33:650–663.
Manage. 18:105–109. Golini R, Landoni P. 2014. International development projects
Ahonen P. 1999. General considerations and user experiences by non-governmental organizations: an evaluation of the
of project cycle management in a small donor-transitional need for specific project management and appraisal tools.
country context. Impact Assess Project Appraisal. 17:97–105. Impact Assess Project Appraisal. 32:121–135.
Ahsan K, Gunawan I. 2010. Analysis of cost and schedule Gore T, Fischer TB. 2014. Uncovering the factors that can
performance of international development projects. Int J support and impede post-disaster EIA practice in developing
Project Manage. 28:68–78. countries: the case of Aceh Province, Indonesia. Environ
Akroyd D. 1999. Logical framework approach to project Impact Assess Rev. 44:67–75.
planning, socio-economic analysis and to monitoring and Ika LA. 2012. Project management for development in Africa:
evaluation services: a smallholder rice project. Impact Assess why projects are failing and what can be done about it.
Project Appraisal. 17:54–66. Project Manage J. 43:27–41.
Baron RM, Kenny DA. 1986. The moderator-mediator variable Ika LA, Diallo A, Thuillier D. 2012. Critical success factors for
distinction in social psychological research: conceptual, World Bank projects: an empirical investigation. Int J Project
strategic, and statistical considerations. J Pers Soc Psychol. Manage. 30:105–116.
51:1173–1182. Ika LA, Lytvynov V. 2011. The “management-per-result”
Bell S. 2000. Logical frameworks, Aristotle and soft systems: a approach to international development project design.
note on the origins, values and uses of logical frameworks, Project Manage J. 42:87–104.
in reply to Gasper. Publ Adm Dev. 20:29–31. Ika L, Saint-Macary J. 2012. The project planning myth in
Coleman G. 1987. Logical framework approach to the international development. Int J Manage Projects Bus. 5:5–5.
monitoring and evaluation of agricultural and rural Kalchschmidt M. 2012. Best practices in demand forecasting:
development projects. Project Appraisal. 2:251–259. tests of universalistic, contingency and configurational
Couillard J, Garon S, Riznic J. 2009. The logical framework theories. Int J Prod Econ. 140:782–793.
approach-millennium. Project Manage J. 40:31–44. Ketokivi MA, Schroeder RG. 2004. Strategic, structural
Cracknell BE. 2000. Evaluating development aid: issues, contingency and institutional explanations in the adoption
problems and solutions. Thousands Oaks, CA: Sage. of innovative manufacturing practices. J Oper Manage.
Crawford L, Pollack J. 2004. Hard and soft projects: a framework 22:63–89.
for analysis. Int J Project Manage. 22:645–653. Khang DB, Moe TL. 2008. Success criteria and factors for
Crawford P, Bryce P. 2003. Project monitoring and evaluation: a international development projects: a life cycle based
method for enhancing the efficiency and effectiveness of aid framework. Project Manage J. 39:72–84.
project implementation. Int J Project Manage. 21:363–373. Landoni P, Corti B. 2011. The management of international
Delery JE, Doty DH. 1996. Modes of theorizing in strategic development projects: moving toward a standard approach
human resource management: Tests of universalistic, or differentiation? Project Manage J. 42:45–61.
contingency, and configurational performance predictions. Lovegrove N, Gebre B, Lee T, Kumar R. 2011. McKinsey-Devex
Acad Manag J. 39:802–835. survey results: practitioners see need for new approaches
Diallo A, Thuillier D. 2004. The success dimensions of to system-wide reform. McKinsey-Devex. http://www.devex.
international development projects: the perceptions of com/en/news/mckinsey-devex-survey-results-practitioners-
African project coordinators. Int J Project Manage. 22:19–31. see/77026
Diallo A, Thuillier D. 2005. The success of international Martinsuo M, Hensman N, Artto K, Kujala J, Jaafari A. 2006.
development projects, trust and communication: an African Project-based management as an organizational innovation:
perspective. Int J Project Manage. 23:237–252. drivers, changes, and benefits of adopting project-based
Durning B. 2014. Benefits of coupling environmental management. Project Manage J. 37:87.
assessment and environmental management to aid disaster McCarthy J, Zakrajšek E. 2007. Inventory dynamics and business
risk reduction and management. J Environ Assess Policy cycles: what has changed? J Money Credit Bank. 39:591–613.
Manage. 16:1450029. Newcomer K, Baradei LE, Garcia S. 2013. Expectations and
Ebrahim A. 2003. Accountability in practice: mechanisms for capacity of performance measurement in NGOs in the
NGOs. World Dev. 31:813–829. development context. Publ Adm Dev. 33:62–79.
IMPACT ASSESSMENT AND PROJECT APPRAISAL  153

Powell TC. 1995. Total quality management as competitive Steinfort P. 2010. Understanding the antecedents of project
advantage: a review and empirical study. Strateg Manag J. management best practice-lessons to be learned from
16:15–37. aid relief projects [PhD]. Melbourne: School of Property,
Saad M, Cicmil S, Greenwood M. 2002. Technology transfer Construction and Project Management, RMIT University.
projects in developing countries–furthering the Project Youker R. 1999. Executive point of view: managing international
Management perspectives. Int J Project Manage. 20:617– development projects-Lessons learned. Project Manage J.
625. 30:6–7.
Smith P. 2000. A comment on the limitations of the logical Zhai L, Xin Y, Cheng C. 2009. Understanding the value of project
framework method, in reply to Gasper, and to Bell. Publ Adm management from a stakeholder’s perspective: case study
Dev. 20:439–441. of mega-project management. Project Manage J. 40:99–109.
Solem RR. 1987. The logical framework approach to project Zimmermann F, Smith K. 2011. More actors, more money, more
design, review and evaluation in AID: genesis, impact, ideas for international development co-operation. J Int Dev.
problems, and opportunities, AID. Working Paper No. 23:722–738.
99. Washington Center for Development Information & Zotteri G, Kalchschmidt M. 2007. Forecasting practices:
Evaluation Agency for International Development. empirical evidence and a framework for research. Int J Prod
Sousa R, Voss CA. 2001. Quality management: universal or Econ. 108:84–99.
context dependent? Prod Oper Manage. 10:383–404.
Sousa R, Voss CA. 2008. Contingency research in operations
management practices. J Oper Manage. 26:697–713.

Appendix

Table A1. Measures for the drivers.


Group Variable Measure
Context variables Context variability 1–5 Likert scale variables measuring the agreement with the following sentences (1: Strongly disagree; 5:
strongly agree):
1.The context of my projects is characterized by a high variability (prices, laws, availability of personnel,
suppliers, …
2.During the execution/implementation phase my projects are significantly changed (due to unexpect-
ed constraints, local needs, etc.)

Correlation of 0.391 (sig. 0.000)


Size of the NGO The NGO’s number of employees
Country development Binary variable identifying NGOs based in developed (0) or developing countries (1)
Project variables Average project The average duration (in month) of the projects managed by the interviewee
duration
Average size of the The average size (in USD) of the projects managed by the interviewee
project
Number of projects The number of projects managed by the interviewee in the last two years
Average number of The average number of stakeholders involved in the managed projects
stakeholders
Principal aim of the Nominal variable identifying the type of the project:
project 1.Delivery and integration of products and tools (e.g. medical, IT, energy equipment)
2.Construction of buildings or infrastructures
3.Development of social, economic, health, and environmental services (education, finance, health-
care, …)
4.Awareness raising and campaigning (human rights, environment, …)
5.Other

Results communication 1–5 Likert scale variable measuring the agreement with the following sentence: At the end of the project
we evaluate and communicate the results to the stakeholders (1: Strongly disagree; 5: strongly agree)
NGO objectives External objectives The NGO’s attitude to focus on long-term and strategic impacts
Likert scale items indicating the importance of the following objectives:
1.Obtain long-term project impact (outcome/goal)
2.Stakeholder/partners involvement
3.Extension of the project’s ownership to the local community
4.Economic sustainability after the end of the project
5.Satisfaction of the local community
Cronbach’s alpha = 0.874
Internal objectives The NGO’s attitude to focus on time and cost. Average of 1–5 Likert scale items indicating the importance
of the following objectives:
1.Comply with the budget
2.Comply with the expected time
Correlation: 0.681 (sig. 0.000)
154  R. GOLINI ET AL.

Table A2. Project performance measures.


Type of performance Performance measures Factor loadings Cronbach’s alpha
Time–cost performance Comply with the budget 0.887 0.772
Comply with the time schedule 0.862
Strategic performance Obtain long-term project impact (outcome/goal) 0.507
Stakeholder/partners involvement 0.616
Extension of the project’s ownership to the local community 0.814 0.752
Economic sustainability after the end of the project 0.645
Satisfaction of the local community 0.739
Minimum eigenvalue: 1.13
Total variance explained: 73%

You might also like