Good data are not enough: Understanding limited information use for climate risk and food security management in Guatemala

Abstract

Climate extremes are one of the main drivers of acute food insecurity. In Guatemala, acute food insecurity reaches alarming levels when the usual dry period during the bimodal rainy seasons is extended or starts earlier than expected. Drought has a slow-onset which theoretically leaves sufficient lead-time for addressing impacts on food security. In practice, emergency response to drought is often reactive and arrives late, starting when the crisis is already evolving. Climate services and food security information systems are key ingredients for integrated climate risk and food security management worldwide. In Guatemala, stakeholders broadly agree on the useful-ness of this type of information for decision-making and direct significant efforts towards improving information availability and quality. But the impact of agro-climatic and food security information on decisions is ad hoc or not systematic. Through a mix of qualitative, ethnographic, and participatory methods, we investigated why this situation occurs. We found that different aspects lead to this phenomenon: the impact of drought on food security is mediated by different socio-economic, political, and institutional factors that tend to differ strongly between regions or even communities across the country. This puts special requirements on information provision for decision-making. Information use patterns can be explained by technical, data-related aspects as reliability, timeliness, or accessibility. But only by considering the institutional and organiza-tional context we get a complete understanding on what frames the information-use patterns in climate and food security management in Guatemala. Our research shows that investments in technical aspects of data provision and infrastructure for increased climate and food security management need to address institutional and organizational challenges in order to be effective.

Full text

Climate Risk Management 30 (2020) 100248
Available online 8 October 2020
2212-0963/© 2020 Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Good data are not enough: Understanding limited information use
for climate risk and food security management in Guatemala
Anna Müller
a
,
*
, Claudia Bouroncle
b
, Ada Gayt´
an
d
, Estuardo Gir´
on
c
,
Andrea Granados
e
, Vesalio Mora
a
,
e
,
f
, Fernando Portillo
d
, Jacob van Etten
a
a
The Alliance Bioversity-CIAT, Costa Rica Country Ofce c/o Campus CATIE, Turrialba, Costa Rica
b
Universitat Polit`
ecnica de Val`
encia, Val`
encia, Spain
c
Centro Agron´
omico Tropical de Investigaci´
on y Ense˜
nanza (CATIE), Turrialba, Costa Rica
d
Acci´
on contra el Hambre, Guatemala City, Guatemala
e
Universidad de Costa Rica, San Jos´
e, Costa Rica
f
Ministerio de Agricultura y Ganadería, Siquirres, Costa Rica
ARTICLE INFO
Keywords:
Seasonal hunger
Drought
Climate risk management
Food security
Institutional analysis
ABSTRACT
Climate extremes are one of the main drivers of acute food insecurity. In Guatemala, acute food
insecurity reaches alarming levels when the usual dry period during the bimodal rainy seasons is
extended or starts earlier than expected. Drought has a slow-onset which theoretically leaves
sufcient lead-time for addressing impacts on food security. In practice, emergency response to
drought is often reactive and arrives late, starting when the crisis is already evolving. Climate
services and food security information systems are key ingredients for integrated climate risk and
food security management worldwide. In Guatemala, stakeholders broadly agree on the useful-
ness of this type of information for decision-making and direct signicant efforts towards
improving information availability and quality. But the impact of agro-climatic and food security
information on decisions is ad hoc or not systematic. Through a mix of qualitative, ethnographic,
and participatory methods, we investigated why this situation occurs. We found that different
aspects lead to this phenomenon: the impact of drought on food security is mediated by different
socio-economic, political, and institutional factors that tend to differ strongly between regions or
even communities across the country. This puts special requirements on information provision for
decision-making. Information use patterns can be explained by technical, data-related aspects as
reliability, timeliness, or accessibility. But only by considering the institutional and organiza-
tional context we get a complete understanding on what frames the information-use patterns in
climate and food security management in Guatemala. Our research shows that investments in
technical aspects of data provision and infrastructure for increased climate and food security
management need to address institutional and organizational challenges in order to be effective.
1. Introduction
Seasonal climatic conditions and climate variability can severely affect food and nutrition security. Effective and dynamic man-
agement of climate risks for food and nutrition security needs improved data and information for decision-making (Coughlan De Perez
* Corresponding author.
E-mail address: a.muller@cgiar.org (A. Müller).
Contents lists available at ScienceDirect
Climate Risk Management
journal homepage: www.elsevier.com/locate/crm
https://doi.org/10.1016/j.crm.2020.100248
Received 25 June 2019; Received in revised form 31 July 2020; Accepted 4 October 2020

Climate Risk Management 30 (2020) 100248
2
and Mason, 2014; Enenkel et al., 2015b; Pulwarty and Sivakumar, 2014). Many organizations around the world are developing and
improving information for food security management, yet often the information is on a coarse scale to inform donors for humanitarian
response and assistance (Mock, Morrow, & Papendieck, 2013). Different efforts are attempting to overcome these shortcomings to
better inform national and local decision-making.
In Guatemala, periods of acute food and nutrition insecurity related to climatic patterns occur frequently with signicant impact on
small-scale farm families (Food Security Information Network, 2018). In 2019, more than 3 million Guatemalans were in a critical food
security situation (Food Security Information Network, 2020). Multiple actors are investing in the improvement of information
provision for better climate and food security management and emergency response. Several organizations collect, collate, interpret,
and provide agro-climatic and food security information. But we have observed that despite the substantial effort on improving in-
formation provision, this information has a limited and unsystematic inuence on decision-making and policy implementation for
climate risk and food security management. Here, we attempt to explain this paradox.
One possible explanation of the gap between availability and use of information is the mismatch with the needs of information
users, requiring a more tailored approach (Enenkel et al., 2015a; Moss et al., 2016). The determining factors of climate information use
for decision-making in different sectors as agriculture or water management have been relatively well understood (Lach and Rayner,
2017; Lemos et al., 2012; Rayner et al., 2005; Saylor Mase and Prokopy, 2014; Soares and Dessai, 2016). Food and nutrition security
are complex, embracing multiple dimensions. Adequate management and response need contextualized analysis to consider causal
interdependencies (Mock et al., 2013). This includes information on climatic aspects, but also on agricultural and socioeconomic
indicators. In the literature explaining climate information use for decision-making, little attention has been paid to the use of climate
and food security information for managing food insecurity.
Patterns of information use can be explained through characteristics of the information itself (accuracy, credibility, salience,
timeliness) and the process of production of the information (Lemos, 2015). But a sole focus on technical explanations might fall short
in explaining climate and food security information use as it does not consider the role of context, institutions, organizational culture
and norms ins shaping decision-makers behavior (Flagg and Kirchhoff, 2018; Saylor Mase and Prokopy, 2014; Vogel and OBrian,
2006). For example, Rayner et al. (2005) use an institutionalist perspective to explain why US water managers hardly use weather
forecasts, demonstrating how institutions and organizational culture inuence information use. The role of contextual and institutional
factors shaping information use for climate risk management is gaining increased attention but requires more in-depth research to
understand important dimensions in different settings and cases (Flagg and Kirchhoff, 2018). Such understanding is needed as in-
vestments in better food security information systems might have little effect, unless they contribute to contextual or institutional
changes that will foster information use in decision-making. To examine patterns of information use for climate risk and food security
management, we conducted an in-depth, multi-year case study in Guatemala. We used multiple sources of qualitative and quantitative
data to build a picture of information usage patterns referring to two different frameworks, one referring to the institutional de-
terminants of information use in decision-making (Thompson and Wildavsky, 1986) and the other referring to different dimensions of
data quality (Wang and Strong, 1996). This allows to consider data-related and contextual and institutional aspects.
The paper is structured as follows. We start with a presentation of the data collection and analysis approach in chapter 2, explaining
the different sources and how we triangulate the different data and information. This is followed by two contextual chapters that give
an insight into the food security situation in relation to climatic factors in Guatemala and into the information management strategy of
Table 1
Details on data collection activities.
Type Information Respondents/Participants
1. Survey, convenience sample Information preferences, information use, decisions that
should be supported by information using different food
security scenarios
33 respondents from different public sector organizations
and NGOs from municipal and central level
2. Semi-structured interviews,
purposeful and snowball
sampling
Information use at the municipal level, decision to be
supported, rating and evaluation of the existing information
22 respondents from public sector organizations at the
municipal level
3. Semi-structured interviews,
purposeful and snowball
sampling
Stakeholder-based evaluation of existing information
products
40 decision-makers from public and international
organizations, central and municipal level (see Bouroncle
et al., 2019; Bouroncle et al., 2017)
4. Semi-structured interviews,
purposeful and snowball
sampling
Understand decision-making processes, budgeting and use
of forecasts for public planning in selected municipalities if
the Guatemalan dry corridor
45 decision-makers and technical staff from municipal
organizations, mainly MAGA, Mayors and Community
Council members
5. Workshops Decision-making cycles, co-design and prototyping of
information products and information system,
Around 80 participants in 4 workshops, mainly from
SESAN central and municipal level, but also other public
institutions
6. Roundtables Discussions on which data are needed, problems of the
existing data system, barriers to data use
From public institutions, NGOs and international
organizations
7. Meetings Personalized meetings among project partners and SESAN
to understand problems and needs, to coordinate activities|
SESAN delegates, mostly central level
8. Secondary data and reports Main source was www.reliefweb.int, depository for reports
from ofcial governmental and non-governmental sources
on crises and disasters.
A. Müller et al.

Climate Risk Management 30 (2020) 100248
3
the government in relation to this in chapter 3. In chapter 4, we describe with a concrete example, how food security responses are
inuenced (or not) by the information available to decision maker. We then analyze decision making and information use in more
detail in chapter 5. The paper closes with concluding remarks in chapter 6.
2. Data collection and analysis
We used an in-depth, longitudinal case study approach that enables us to explore complex linkages in a situation that evolved over
time. Throughout the case study, we used a mix of quantitative and qualitative research techniques according to data availability,
contextual factors, and specic research interest. All information and data used in this research was generated throughout a project on
the development of an agro-climatic and food security information system for the Secretariat for Food and Nutrition Security (SESAN)
from 2015 through 2018. This project was nanced by the CGIAR program on Climate Change, Agriculture and Food Security and was
implemented with the support of the Tropical Agriculture Research and Higher Education Center and Acci´
on contra el Hambre (ACH).
During the project we addressed different research questions (e.g. about the usability of information products (see Bouroncle et al.,
2019) or decision-makers information preferences (unpublished thesis by Granados, 2017). This research uses the different data and
information collected to address our specic research question.
We used a mix of different data collection methods as outlined in Table 1. Most of the data collected during the project period
focused on Guatemala City, the capital, and the Department of Chiquimula. This area forms part of the so-called Guatemalan Dry
Corridor, where climate-related acute undernutrition is especially of concern. Firstly, much of the data were collected through
participatory observation during the project activities as eld trips, stakeholder workshops, and project meetings.
Secondly, more structured data came from one survey and three rounds of semi-structured interviews to get a more systematic
understanding of information use and needs, and to evaluate existing information products. During the course of the project and the
different research activities, we interviewed around 140 central and municipal level decision-makers and technical staff from the
public sector, NGOs and international donors relevant for food security (see Table 1).
Thirdly, we used secondary data and ofcial reports from the government and non-governmental actors to complement our data
collection efforts. For both, qualitative and quantitative data we applied a non-random sampling approach as the idea of a case study is
to dive deep into purposefully selected few cases in order to explore patterns (Gobo, 2004).
We applied a mix of stratied purposeful and snowball sampling to select respondents. Systematic selection criteria were developed
and included for example professional occupation, geographical location, and experience in the topic.
To analyze the data, we applied simple descriptive techniques to the quantitative data. We applied content analysis to the qual-
itative data, for some of the contents applying computer-assisted qualitative data analysis using the R package RQDA (Chandra and
Shang, 2017). Content analysis allows to identify concepts, patterns or relationships that describe the phenomenon under study (Elo
and Kyng¨
as, 2008). We used document analysis to analyze secondary, mostly unstructured data coming from ofcial reports (data type
8 in Table 1). We used a deductive approach (going from the general to the more specic) parting from the model of data quality
developed by Wang and Strong (1996), explained in chapter 5, to explore patterns of data quality. We used an inductive approach for
exploring the role of contextual factors as we did not identify an adequate conceptual framework.
2.1. Context: Food security and drought in Guatemala
Seasonal hunger is a recurring phenomenon in Guatemala and the level of acute food insecurity, especially among children under 5
years, is alarmingly high (Food Security Information Network, 2018). Acute undernutrition is related to the occurrence of an extended
dry spell that coincides with a critical phase for crop development in staple grain production (FAO and ACH, 2012). This so-called
canícula occurs between the two rainfall peaks in the bimodal annual rainfall pattern and is related to the ENSO-cycle (Maga˜
na
et al., 1999). During the El Ni˜
no phase of the ENSO cycle, the onset and duration of the dry-spell changes signicantly: it may start
earlier and last longer than normally expected by farmers. Since small-scale producers rely much on their own grain production, dry
periods directly affect food security, but also indirectly through the potential impact on food prices or the availability of off-farm
employment. Extended dry spells occur with increasing frequency and over an expanded area in recent years (Acci´
on contra el
Hambre, 2015a).
Drought has a slow onset, which in theory would leave sufcient lead time for preparation and response. Acute food insecurity
usually does not come by surprise. Seasonal climate forecasts of drought based on ENSO are increasingly reliable and make it possible
to anticipate drought occurrence (Palmer and Weisheimer, 2012). We also observed an increase in information available that could
support the proactive management of climate-related food security problems.
In spite of the increasing skill of seasonal climate forecasts, it is still difcult to predict the onset, duration, and exact location of
droughts (Enenkel et al., 2015a). A clear denition or typology of drought is still lacking in Guatemala and many other contexts. Even
though the relationship between drought and food security is clear, the impacts of different types of drought on socio-economic in-
dicators and food security is not straightforward (Enenkel et al., 2015a) and is mediated by several factors as for example availability of
off-farm labor, food prices, out-migration and remittances (Hedlund, 2007). So, despite the long lead time, the effect of drought on
food security is surrounded by a degree of uncertainty that makes decision-making far from straightforward.
2.2. Context: Food security information management in Guatemala
Over the last ten years, Guatemala has seen signicant effort from public and international actors to improve the regulatory
A. Müller et al.

Climate Risk Management 30 (2020) 100248
4
framework for food security and climate risk management in Guatemala. In 2008, the National Food and Nutrition Security System
(SINASAN) was established by law as the institutional framework to strategically organize and coordinate all food and nutrition se-
curity related activities by public, private and non/governmental actors in the country (Gobierno de Guatemala, 2008). The main
objective of SINASAN is to impulse interventions to eradicate food insecurity and implement the national food and nutrition security
policy. As shown in Fig. 1, the SINASAN integrates different administrative levels: the National Council for Food and Nutrition Security
(CONASAN) is the governing body, comprised of the ministers of all Ministries related to food and nutrition security and the Vice-
president (Gobierno de Guatemala, 2008). The CONASAN impulses policies and actions supporting food and nutrition security in
Guatemala. SESAN is the coordination body of the SINASAN and responsible for technical planning and coordinating and articulating
all national and international actors working in food and nutrition security present in the country. There are food and nutrition security
councils at the department (CODESAN), municipal (COMUSAN) and community level (COCOSAN), although the law doesnt specify
their functions within the SINASAN. The system has a hierarchical structure and follows a linear model where information ows from
the local to the national level and the decisions are made in a top-down manner. The system has a weak integration of actors outside the
public sphere and does not foresee these interactions for information exchange and decision-making.
The 2008 Law states different objectives to reach food and nutrition security. The objectives are dened around the four pillars of
food and nutrition security: food availability, access to food, utilization, and stability over time. As one mechanism for monitoring,
evaluation and early warning, the law foresees the development of a National Food Security Information System (SIINSAN) that is
coordinated by SESAN. Different public institutions and organizations are supposed to feed information into the system. A main
contributor outside the SESAN is the Ministry of Public Health (MSPAS). The data coming from these different actors is openly
accessible at www.siinsan.gob.gt. The platform integrates the different monitoring schemes for acute and chronic undernutrition, for
pregnant women and children under two in public health centers and monitoring of the food and nutrition related activities. Most of
the information available on the platform are performance indicators (e.g. how many health centers are operating; how many doses of
supplements are available etc.). Pron´
osticoSAN, the food security outlook generated by SESAN in coordination with a network of
government and non-governmental actors, is available for download.
Recognizing the information problem for early response to acute undernutrition in Guatemala and responding to its legal obligation
to implement an early warning system, SESAN adopted in 2014 a community-based monitoring and early warning system based in
sentinel communities (Secretaría de Seguridad Alimentaria y Nutricional, 2014). The method was initially developed by ACH (2013).
SESAN and different international NGOs implemented the system in prioritized municipalities. Local food security committees
(COCOSAN) were supposed to provide information on relevant indicators related to seasonal hunger, such as rainfall, damage and loss
in staple crops, undernutrition, or off-farm labor opportunities (Acci´
on contra el Hambre, 2013). This information was integrated into
the SIINSAN to support decision-making at different levels. Eventually, the system lost government support after a political change in
2016 and was discontinued in practice. In 2018, SESAN adopted an improved community information system with national coverage
that was co-developed with the support of our project. The system provides local information to decision-makers at the municipal and
national level.
3. Drought and food security response, 2014–2015
To understand the complexities that surround decision-makers when it comes to food security management, we reconstructed the
response of the Guatemalan government to the 2014–2015 droughts and resulting food insecurity. We analyzed ofcial reports from
the Government, international organizations like FAO or World Food Programme, and NGOs like Oxfam or ACH (see data type 8 in
Fig. 1. Model showing information ow and decision-making in the SINASAN.
A. Müller et al.

Climate Risk Management 30 (2020) 100248
5
Table 1).
In 2014, after several years of irregular rainfall, Guatemala experienced an early-onset canícula with an unusual long duration,
resulting in a poor agricultural harvest (Acci´
on contra el Hambre, 2015b). In August 2014, the government realized a country-wide
agriculture loss and damage assessment (Gobierno de Guatemala, 2014). More than 20% of the surveyed families experienced par-
tial or total losses. In more than 50% of the communities, farmers did not have enough inputs (mainly seeds) to produce in the next
season (Gobierno de Guatemala, 2014).
Guatemala declared a state of emergency (Diario de Centro Am´
erica, 2014). This allowed the government to access emergency
funds and triggered a response by the international community. Although the unusual canícula was the event that triggered the alarms,
the deterioration of the food security situation was related to a more complex interplay of factors. In 2014, food grain reserves were
almost exhausted due to low harvests in the previous years. Central America was experiencing a serious crisis in the coffee sector due to
the coffee rust epidemic, which reduced coffee harvest by 10–16% and thus reduced the employment rate in the coffee sector (Avelino
et al., 2015). Off-farm employment in the coffee sector is a relevant coping strategy for communities in the lean period (FAO and ACH,
2012). Missing off-farm employment affected the already weak household economy in the drought-affected regions (World Food
Programme, 2015). At the same time, the Guatemalan state was not able to provide basic health care. Many public health centers were
not operating at all or with low resources, unable to fulll their functions in providing proper attention to acute malnutrition cases
(ACH, 2015a). The Guatemalan government had to admit difculties in funding the attention of the food security crisis (Prensa Libre,
2014).
In 2015, persistent El Ni˜
no conditions suggested there was a high probability of an extended canícula, and the food security situation
was not expected to improve (Secretaría de Seguridad Alimentaria y Nutricional, 2015). After the 2014 crisis, families did not have the
possibility to recover and many had depleted their food stocks and household and farm assets (WFP, 2015). An estimated 70% of the
farmers were not able to undertake the investment needed for sowing in the following season (Aguilar et al., 2014).
The reduced availability of off-farm employment opportunities continued, while prices of basic food items rose (WFP, 2015). The
WFP estimated 1.3 million people to be affected by the extended drought, 54% were food insecure (WFP, 2015). Among subsistence
farmers, loss and damage in grain production was estimated to reach up to 70% for the second consecutive year (Equipo Humanitario
de País Guatemala, 2015). The Government of Guatemala was calling for international support to overcome the food security crisis
(Gobierno de Guatemala, 2015).
Despite the urgency of the situation, MAGA did not present any ofcial data on the impact of the drought in 2015 (ACH, 2015b).
These data are essential for declaring a state of emergency, which opens mechanisms to redirect national and international funds for
emergency response (Müller et al., 2019). Guatemala did not declare a food security emergency. According to government ofcials we
talked to, this was due to the limited availability of government funds. The government seemed afraid to raise expectations it could
then not full. The crisis was occurring during a politically very critical time. In 2015, presidential elections were to be held and the
country was in turmoil due to a big corruption scandal that involved high government ofcials. The attention of the chronically
undernanced public sector was diverted to actions for electoral effect.
Comparing 2014 and 2015 shows that the government response is not predictable from the status or forecasts of climate and food
security. The 2015 crisis was not less serious than in 2014 but did not trigger the same response. In other types of emergencies,
emergency protocols leave little room for political and institutional improvisation. In the case of drought-induced acute malnutrition,
the institutional response is more complex and less predictable, and the lack of closely dened procedures provides more opportunities
for political and institutional maneuvering (Müller et al., 2019). This highly negotiated character of government food security re-
sponses in Guatemala is important to explain the patterns of information use in food security management in the next sections.
4. Decision making and information use
To get a clearer picture about the decision-making context, we started asking which decisions key actors in the area have to make in
terms of climate risk and food security management. These decisions comprise a broad eld of technical, management and budgetary
decisions that differ according to the administrative level (see Table 2). Most respondents across all administrative levels stated to
Table 2
Main decisions at different administrative levels.
Frequency of responses per administrative level
Main decisions, categories Central Department Municipal
Training and capacity building for technical assistance 1 0 0
Design and manage interventions and prevention strategies 6 9 9
Technical recommendations for agriculture 7 10 14
Divulgation of information 6 10 6
Institutional and budget planning 6 4 0
Geographical priorization 7 16 11
Distribute non-governmental funds, request international assistance 0 3 2
Food and Nutrition Security Evaluations 3 10 7
Number of respondents at administrative level 22 25 27
Total number of respondents 73
Aggregated results from survey 1 and 3, Table 1, multiple answers were possible
A. Müller et al.

Climate Risk Management 30 (2020) 100248
6
make decisions around technical recommendations and geographic prioritization. A key technical decision would be the recom-
mendations on sowing dates considering climatic forecasts, like the predicted onset and duration of the canícula.
One key decision on geographic prioritization is about which communities to cover in what they call in Guatemala a barrido
nutricional, a representative nutrition survey. This is a (nearly) exhaustive survey to detect children with acute undernutrition. This
type of survey is necessary as not all malnourished children are captured by statistics: they are not taken to health centers and reporting
systems are decient. This survey provides more reliable information for triggering a decision, e.g. to prioritize specic communities
for interventions that needs reliable and representative information to assure affected communities are attended.
We describe decision making and data use patterns in Table 2 using descriptive results from two different surveys (Survey 1 and 3 in
Table 1). The main difference between the administrative levels refers to aspects of institutional and budgetary planning decisions.
This is very prominent at the central level and plays no role at the municipal level, reecting the centralized governance structure in
Guatemala. Many civil servants we interviewed at the local administrative level stated that they have limited decision-making power,
e.g. on the allocation of public funds. The central government makes most budgetary decisions. Other important decisions mentioned
are the design and management of food security interventions and the divulgation of information.
In another survey we asked several stakeholders at the municipal level how their decisions would change if the available infor-
mation would better t their needs (see survey 2 in Table 1). Out of a sample of 22 respondents, 12 (55%) told us their decisions would
be more specic in terms of what to do and where to act. 12 respondents (55%) said decisions would improve if there would be a
follow-up and monitoring of the outcome of the decisions. 9 respondents (40%) stated that decisions would improve if they can be
made in a timelier fashion. For 6 respondents (27%), an improved decision would be a more trustworthy and solid. A lower fraction of
the respondents perceived that decisions and their implementation should be better coordinated among actors at the local level (5
respondents or 23%) and should respond to a strategic planning framework (3 or 14%).
Despite a common agreement on the usefulness of agro-climatic and food security information for decision-making, we found
evidence of the unsystematic and ad hoc use of information in decision-making among the respondents. Often, it was not clear from the
responses how concrete decisions are linked to concrete types of information. A semi-structured survey with 40 respondents revealed
that more than 20% did not use any climate and food security information product at all or could not specify which information they
use for their decision-making. Those who used the information did use it for example for prioritizing areas for emergency interventions
(e.g. to distribute food or vouchers) or to dene the sowing date and recommend varieties. At a more strategic level, the information is
used to develop yearly work plans or to decide on budget and stuff needs for future time periods. Without the adequate information at
hand, these decisions may not have enough substance. Interviewees mention that they criticize or not support decisions that have been
done based on what they assume inadequate data.
4.1. Data quality
In order to understand the unsystematic and ad hoc use of information for decision-making, we rst discuss our ndings in the light
of a framework of data quality and its different dimensions (Wang and Strong, 1996). The concept provides a good (but, as we show, an
insufcient) framework to describe information use patterns in our study context.
Data quality refers to “data that is t for use by data consumers” (Wang & Strong, 1996, p 6.). This implies a user-centered
perspective. Based on a study of consumer perceptions, (Wang and Strong, 1996) identify four dimensions of data quality: intrinsic,
contextual, representational data quality and accessibility of data. Intrinsic data quality consists of data accuracy and objectivity,
believability and reputation. Contextual data quality means that data quality is assessed according to the context of the data use and
includes relevancy, timeliness and appropriateness, among others. Representational data quality describes aspects like easy under-
standing and interpretation and consistent representation. And nally, accessibility is about the importance of unrestricted access of
data.
To shed light on the topic of data quality we refer to the different qualitative data collected during the project period, coming from
semi-structured surveys and participatory observation (See data collection type 4 – 7 in Table 1).
4.1.1. Intrinsic data quality
Scale mismatch: One of the main factors explaining mentioned by interviewees to explain their use of information for decision-
making is the incorrect temporal and spatial scale. The coarse nature of most of the available agro-climatic and food security infor-
mation is usually sufcient for central programmatic or policy decisions affecting the whole country. Decisions affecting the municipal
level, however, where the execution and implementation takes place, requires localized information. The incidence of drought and its
impact on food security can vary signicantly between communities and depends on multiple factors. To prioritize the distribution of
scarce resources (personnel, budget, aid), it is crucial to have information on the situation in the different communities. Most of the
ofcial information products including SESANs food security information bulletin “Pron´
osticoSAN” only make a broad differentiation
between regions in their data analysis and recommendations. The spatial scale of information is an important aspect that has been
identied as a barrier to climate information use for decision-makers in Guatemala (Bouroncle et al., 2017) and elsewhere (Rosas et al.,
2016). During our study we observed that in different municipalities, locally produced information about rainfall, damage and loss in
grain production or cases of children with acute undernutrition, was produced and used for decision-making. But this remained un-
systematic and ad hoc, depending on the motivation of an individual decision-makers and did not necessarily involve ofcial or
systematically collected information.
Incomplete data: Respondents and stakeholders often mentioned that the incompleteness of ofcial information. In 2014–2015,
MSPAS suffered from severe budget problems with health centers shutting down in many regions of the country. The situation in
A. Müller et al.

Climate Risk Management 30 (2020) 100248
7
MAGA was similar: the extension system was understaffed and underfunded. These two institutions are the main players to generate
drought and food security information at the local scale. The crisis resulted in inconsistent reporting. For example, during the crisis in
2014, only around 70% of the health centers where reporting to the SIINSAN (Secretaría de Seguridad Alimentaria y Nutricional,
2014). Many stakeholders, especially from NGOs, were reluctant to use ofcial information for their decision-making.
Data representativeness: In response to the problem of spatial scale and the need of the decision-makers to have localized infor-
mation, SESAN worked towards lling this information gap through the establishment of a system of sentinel site communities
providing localized information to the SIINSAN. But this information remained unused and tells another important story about why
decision-makers refuse to use information.
Decision-makers mainly at the central level questioned constantly whether the selected sentinel communities were producing
representative data. This is a common critique of sentinel-site information systems. Although, per denition, sentinel sites are not
meant to produce representative data but rather represent a cost-effective way of producing proxy indicators to monitor trends
(Doledec, 2014). In the Guatemalan case, stakeholders were criticizing that there were no formal selection criteria for sentinel
communities in place. As the coverage of the public health and extension system in the rural areas was low, stakeholder feared that
only those communities were selected as sentinel sites that had good coverage with public services and thus lower rates of food
insecurity. This could introduce an observational bias in the information, assuming that acute malnutrition is lower in communities
covered by public assistance. Stakeholder were thus reluctant to use the information. Decision-makers from the local level on the other
hand did nd the information useful. But they have little decision-making power within the SINASAN system and as soon as the in-
formation leaves the context where it has been generated it loses its usefulness.
The problem of observational bias in sentinel surveillance for food security is well-known (Grellety et al., 2013) and regular
rotation of sites, stratied random sampling or of clearly dened selection criteria is recommended. Overcoming this problem seemed
tricky: public organizations and NGOs select communities because of logistical reasons, security issues or path-dependency. This
makes it difcult to convince decision-makers of rotation or sampling strategies. For dening selection criteria, the use of secondary
data is necessary, where we return to the problems mentioned already: wrong scale, incompleteness, inaccurate or outdated data.
Reliability and perceived reliability of the information: Stakeholders complained about the low reliability of the data, especially locally
generated data on food security and production indicators. In the case of local sentinel site information, decision-makers questioned
extension-agents or community members’ ability to produce meaningful information, as they are often not well-educated. The in-
formation was perceived as unreliable.
In stakeholder meetings that we attended, decision-makers especially from NGOs, were constantly questioning ofcial data. One
example frequently mentioned was data about the level of acute undernutrition among children under 5, reported by MSPAS. During
2014 and 2015, several municipalities of the southern coast region demonstrated the highest prevalence. The region does not form part
of the dry corridor. Stakeholder at the central level had serious concerns about the reliability of this data and thus did not act upon this
information. They argued they knew that the employees of the health center of that region were not skilled in detecting acute un-
dernutrition and the centers had a very bad data management. Thus, the numbers would not reect the real situation of seasonal
hunger hot spots in the country. Decision-makers preferred to trust their own impressions or experience rather than the ofcial data.
4.1.2. Contextual data quality
Data bloat: The complex relationship between drought and acute food insecurity makes it difcult to get the “right” information for
decision-making. As participatory observant in multiple stakeholder meetings, we could witness long and controversial discussions
around the selection and denition of the information that is needed to support decision-making. SESANs intention to establish a
localized information system within the SIINSAN was partly hampered by the fact that every technical extension agent in the eld
reported different variables with different measurements without continuity across reporting periods. Some would report the number
of undernourished kids in a community, others the percentage share; some would report days without rain and others reported damage
in crop production as a proxy for the presence of drought. This left decision-makers with a lot of information that was difcult to
analyze and use for somebody not directly involved in the data collection process. The missing agreement on which indicators would
best support decision-making and the missing clearness about how this information should support decision-making, resulted in what
we perceived as data bloat.
Timeliness of the information: The timeliness of the information, an aspect of contextual data quality, is another critical factor for
decision-making. Decision-makers were complaining that the publication of relevant information is regularly delayed or not published
at all, so the information is no longer relevant, or the decisions have been made without consulting the information. In the case of the
information on climate-related damage and loss in production provided by MAGA, for example, interviewees were complaining that
the information was not published on a regular basis so it was difcult to include it into routine decision-making.
4.1.3. Representational data quality
Unprocessed information: Representational data quality relates to format and meaning of data and is important for user’s ability to
interpret the information. Agro-climatic and food security information in Guatemala is often only available in a raw and unprocessed
format. It is difcult to contextualize the information and relate it to a specic decision-making context. The food security and health
data available through the SIINSAN platform are basically raw data and does rarely provide any analysis, conclusions or recom-
mendations, nor does it provide user-friendly data presentations. This makes it hard to analyze the data in a meaningful and decision-
supportive way. This is also reected in ofcial information products that do not present the information in a user-friendly way. A
systematic evaluation of information products in Guatemala revealed that they mostly show technical language difcult to understand
for a lay person, with a lot of text and graphs of low quality without explanation (Bouroncle et al., 2017). Forecasts and agronomic
A. Müller et al.

Climate Risk Management 30 (2020) 100248
8
information are presented with little supporting analysis that could guide stakeholders in using the information for their context-
specic decisions.
Probabilistic nature of information: Information users expressed difculties in using and interpreting probabilistic information, like
the seasonal ENSO forecast. We found strong resistance towards using probabilistic information among various decision-makers as the
perception was that probabilistic information is unreliable . We saw some differences in the attitude towards forecasts between the
central and the local level: whereas probabilistic climate forecasts are widely referred to at the central level (which does not mean that
they are used for decision-making), skepticism at the local level is much more widespread. We perceived that probabilistic information
in the Guatemalan context is not presented in a way that allows the decision-makers to easily understand and contextualize the in-
formation (Bouroncle et al., 2017).
4.1.4. Accessibility data quality
Access to information is an important factor for use as in Guatemala access to information technologies is often restricted. We saw
clear differences in access to food security information: at the central level, the access to ofcial information coming from SESAN,
MAGA or MSPAS is more widespread than at the local level. Most information products relevant for drought and food security
management are issued and circulated at the central administrative level and are not broadly known, circulated or used at the local
level (Bouroncle et al., 2019). When talking to civil servants at the local level we often heard that they had to report regularly on the
food and nutrition security situation but they seldom receive information from the central administrative level. Several interviewees
from the municipal administration of the dry corridor explained that they often refer to unofcial information for their decision-
making, e.g. when planning their eld activities. Sources might be farmers or members of women groups in the communities. At
the local level, ofcial and unofcial information is often distributed word-of-mouth in meetings or among colleagues. There are no
well-established information distribution channels between the central and the local level. The unsystematic use of information or the
use of unofcial information or eld impressions could introduce signicant bias in drought and food security management, leaving for
example population in need unattended.
4.2. Institutional factors inuencing information use
We outlined in detail how different technical dimensions of data quality inuence information use by decision-makers for drought
and food security management in Guatemala. During our research project we gained a deeper insight into how a focus on technical
aspects of data alone does not solve the underlying problems of (non-)use: institutions, contexts and norms also shape strongly how
information is used for decision-making. Without considering these factors, improvements in data quality and technical dimensions
will likely not lead to the full potential of data use in decision-making. We discuss and underpin this argument in the following.
We interpret our ndings referring to institutional theory in a neo-Durkheimian tradition explaining data use as part of institutional
processes and incentive structures (Douglas, 1986; Thompson and Wildavsky, 1986). The organizational context can have a signicant
inuence on the information culture in an organization and in how individual members manage, use and exchange information (6
et al., 2004). This perspective suggests that for increasing information use in decision-making in a given context we do not only have to
think about reducing the technical barriers to usage (Lach and Rayner, 2017; Rayner et al., 2005). Context as a shaping factor of
information use and its role in the era of increased information and digitization is starting to gain more attention in research. The
political environment at the meso-level, the characteristics and dynamics that surround individual behavior at the macro level and the
characteristics of the individual user play an important role in information use for decision-making (Flagg and Kirchhoff, 2018). In the
following we show how attitudes, bureaucratic structures, political instability, political interference and fatalism are important factors
to consider when explaining information use in our case-study context.
Attitudes towards climate change, food security and information use. At the institutional and individual level, among the interviewed
decision-makers we found a broad consensus on that climate variability and extremes are signicantly and increasingly affecting food
security. Interviewees were also quite certain that agro-climatic and food security information would allow them to improve their
decisions in terms of timeliness, credibility, effectiveness and targeting. This is important as personal attitudes towards the role of
climate and climate change seem to be a signicant determining factor especially for climate-information use (Flagg and Kirchhoff,
2018). These attitudes seem to support information use among individual decision-makers.
Bureaucratic structures and centralization. Decision-making structures in ministries, the public administration and the SINASAN are
hierarchical and centralized. Information usually ows upstream and is seen more as an instrument to prove compliance with
bureaucratic obligations than a tool to reach impact in climate risk management and food security. This culture of satisfying
bureaucratic needs can be a strong factor inuencing information use (Byrne and Sahay, 2007; Luís Mosse and Byrne, 2005). Decision-
maker criticized the missing focus on the impact of the decisions: information in the SINASAN is collected to fulll obligations and not
to reach an intended impact on food security on the ground. The SINASAN for example reports the stock of medicine in the health
centers, the number of children attended or the number of supplements distributed, but there is no measure on how effective this is. If
decisions are made and implemented, there is hardly a follow-up, not to speak of monitoring and evaluation, of the effects of the
decisions. This in turn reects what Thompson and Wildavsky (1986) describe as a bureaucratic-hierarchical approach to information
use: the outcome of information use is secondary. Information is produced and used to prove that bureaucratic procedures are
followed.
Budgeting and planning in public institutions in Guatemala is a centralized process. Decentralization efforts in Guatemala are stagnant.
Ofcial directives and decisions come from above with little scope for actors at the lower administrative level (municipalities,
department dependencies) or individuals to make decisions according to the specic situation in the locality (Bossuyt, 2013; Willis
A. Müller et al.

Climate Risk Management 30 (2020) 100248
9
et al., 1991). Ministries dene the budget at the central level. Decision-makers at the local level do not have any decision-making
power to make budget adjustments, if they see e.g. a tense food security situation coming. Funds for drought-related food insecu-
rity can only be liberated after the Government has ofcially declared an emergency (Müller et al., 2019). The situation is similar for
municipal authorities. Legally, municipalities are not able to allocate funds or adjust the budget considering information on a drought
forecast or data indicating difculties in agricultural production and food security. Again, they are only able to release or reallocate
funds once the Government declares an emergency. One mayor stated that he would need a 100% certainty that the event would occur
to be able to consider it in the yearly operational budget. When municipalities plan their yearly operational budget, we often heard that
the information that is used to prioritize communities e.g. for infrastructure projects, comes from informal sources. Decision-makers
rely on their own impression or have a trusted person in a community or institution they contact. This reects the ad hoc and person-
dependent character of information use and does not provide the context for pro-active, crisis-avoiding climate risk and food security
management. Whereas the institutional structure is centralized and hierarchical and the collection and ow of the information from
the local to the central level is regulated, the use by decision-makers for response and the dissemination of the information at different
administrative levels in contrary is characterized by a lack of organizational protocols and procedures, and seems to be ad hoc and
person-dependent. This leads to the paradox that although there is the “right” information available, decision-makers might not know
how to apply and use the information in their decision-making context (Müller et al., 2019). Although highly bureaucratic and
centralized, the system has no clear authority that establishes in a reliable way universal technical standards and protocols for data
collection, analysis, use and response to manage climate risks and food security.
Constant renegotiation and political change. Political decision-making in the area of drought management and food security is sur-
rounded by suspicion of inappropriate use of resources and being mainly guided by political interests and not by data and information.
Also, insecurities about (future) political priorities seem to have a strong impact. This can inuence the continuity of efforts to improve
the technical dimension of information use problems: projects aiming at overcoming certain constraints, e.g. through generating local
information or harmonizing indicators, might lose political support abruptly, as we experienced ourselves. Relevant information might
not be used as actors are reluctant to make decisions due to a political instable situation. At the macro-level we observed a constant re-
negotiation of efforts to improve information for food security, nothing seems to be constant in the face of the political uncertainty.
With every change in political leadership within SESAN or a government change, we had basically to justify and renegotiate every step
and major decision of our project. Again, we see this pattern of ad hoc and unsystematic information policies and low levels of
formalization of procedures in the context of highly bureaucratic structures (Thompson and Wildavsky, 1986). Government in-
stitutions are inuenced by how national politics function in Guatemala. Charles W. Anderson (1962) observed that the political
sphere in Guatemala and other Central American countries is conceived as a space in which factions that compete for power reach
arrangements through constant negotiation. The sources of power of these different factions are highly diverse and based on access to
resources or control over social forces and processes. This is far from the ideal of a representative democracy and the constant
renegotiation of power pervades government institutions as well.
Interference in information generation and use. Drought and food security information has a political dimension and decision-makers
information use is likely to be inuenced by the (perceived or actual) political interference in relevant information. After a drought
impacts production, a key information for decision-makers is the list of affected farmers to plan distribution of emergency support (in-
kind or vouchers for food and agricultural inputs). The compilation of beneciary lists for public subsidies is one area where political
loyalty seems to play a stronger role than actual drought damage and loss in subsistence production (Asociaci´
on de Investigaci´
on y
Estudios Sociales, 2012; Universidad Rafael Landívar and Universidad de San Carlos de Guatemala, 2013). The before mentioned
relative absence of protocols for data generation, analysis and use, opens the door to political interferences at different stages: (Müller
et al., 2019).
Fatalism and failure avoidance. We perceived that the combination of these contextual factors led to a strong attitude of fatalism
among decision-makers and stakeholders, paired with a dominant culture of failure avoidance especially in the public sphere. We often
heard that improving decision-making through sound agro-climatic and food security information would anyway not make any dif-
ference as the chronically underfunded public sector was not able to respond or response would be subject to political interference and
delayed through democratic procedures. Individual decision-makers felt powerless to act upon relevant information because of
existing structures and power balances. But fatalism expressed itself also as a feeling of predetermination and the inuence of un-
controllable external forces (Díaz et al., 2014). We heard from different decision-makers especially at the local level that God in-
uences the impact of climate on food security and there is nothing to do about it.
We also perceived a strong tendency towards failure avoidance in individual decision-making, e.g. fear that the individual would
contradict ofcial decisions or held accountable if a decision was based on a forecast but the forecast did not come true. In many public
institutions in Guatemala, staff uctuation due related to the political instability is very high (Müller et al., 2019). Positions often
depend on (political) loyalty which leads to a high employment insecurity among many civil servants. Combined with strictly hier-
archical structures in the public sphere, this leads to a strong failure avoiding and risk reducing behavior and inaction among decision-
makers to reduce accountability for (possible) erroneous decisions. Structures incentivizing failure avoidance instead of motivating
early actions are a signicant obstacle to information-based early action (Ewbank et al., 2019). Thompson and Wildavsky (1986)
describe this strategy of information use as individualistic survival in fatalistic organizational settings.
5. Discussion and conclusions
Climate services and food security information are considered important ingredients for drought related food security manage-
ment. Nevertheless, in Guatemala the use of information for decision-making in drought and food security management is
A. Müller et al.

Climate Risk Management 30 (2020) 100248
10
unsystematic and ad hoc. Our research aims at explaining patterns of information use in the management of climate-related, seasonal
food insecurity in Guatemala. We show that the relationship between extended midsummer-drought and acute food insecurity is
complex and not straightforward, which makes it difcult for decision-makers in Guatemala to contextualize available information
into their considerations for on-the-ground risk management, early actions and crisis response.
Our results show that main decisions that are or should be supported by information are geographical prioritization of in-
terventions, technical recommendations for agriculture and institutional planning. The relative importance of these decisions differs
according to the administrative level. At the central level, the decisions are more of strategic and at the local level of operational
character, reecting the centralized structure of the public administration in Guatemala. This structure leaves little scope for decision-
makers at the local level to include climate and food security related information into forward-looking planning decisions. We show
that information use for decision making is often unsystematic. We argue that one of the main reasons for this pattern is the missing t
between the available data and information and users’ needs in terms of characteristics like scale, reliability, accessibility, repre-
sentativeness or timeliness of information. We argue that these factors alone cannot fully explain patterns in information use in our
case-study setting.
We look beyond the different technical dimensions of data quality and show that information use patterns are closely aligned with
the institutional and organizational context surrounding decision-makers. In Guatemala, centralized structures in institutional plan-
ning and budgeting, undened procedures as well as public lack in response capacity and fatalism determine actor’s willingness to
include relevant information into their decision-making.
Although many factors related to the different dimensions of data quality and the institutional context outlined above seem to
inhibit information use in decision-making, it is not that food security and agro-climatic information is not used at all for decision-
making. We saw a broad consensus among the people we talked to and observed during different our research that information is
necessary to improve food security and drought management and to realize the shift from reactive crisis response to proactive
management. There are interesting cases and anecdotal evidence on decisions made based on information, for example of successful
changes in the sowing date or the timely attention of a community at risk of acute food insecurity. What we experienced is an un-
systematic and ad hoc way of using agro-climatic and food security information, and often decision-makers refer to unofcial, more
“trustful” sources, e.g. their colleague or own eld impressions, to base their decisions on. This mirrors Anderson’s observation of the
importance of conanza (trust), meaning mutual empathy based on personal relationships, in the Guatemalan political system
(Anderson, 1962). Contextual and institutional factors as described dominate and shape incentives to information use for decision-
making.
These ndings have important implications. They imply that an investment in improving the delivery and scientic value of food
security information will not automatically lead to better decisions and outcomes, if the institutional context is not providing a
supportive environment. Thus, efforts to improve data and information for decision making should go hand in hand with supporting
institutional and governance processes, otherwise institutional incentives might undermine the effort. Studies in other contexts come
to similar results and conrm our statement (Sandefur and Glassman, 2015; Srinivasan et al., 2011). Technical barriers to information
use might also be related to contextual factors and thus need a more integrated approach to problem solving.
In the case of Guatemala, we found it was useful for decision-makers to have more guidance in terms of decision-support structures
through the development of information use and response protocols that consider the scope of decision-making power of an actor for
example at the municipal level. The integration of digital tools for data collection and processing can address different technical
obstacles, for example by reducing time to collect and upload data, automate analysis and visualization. But digital tools also hold the
promise to reduce entry points for political interference in public information. By integrating stakeholders in all stages of the design
process, the participatory co-design of information systems can be a useful approach to assure the sustainable insertion of solutions into
a decision-making context. Collaboration with key stakeholders from the public sector, NGOs and civil society is crucial in order to
understand challenges to information use and to assure that the development of new technical solutions or the improvement of existing
mechanism goes hand in hand with an effort of creating or supporting the institutional enabling environment. The process required to
achieve meaningful change in this context, however, is not limited to digital design, but also involves institutional design.
Information use for decision making is a every time more urgent topic. With the rise of digital tools in many developing countries to
support decision making, it is of researchers and practitioners’ responsibility to assure the match between data and decision-making.
Future research should go beyond case-study evidence and develop more systematic studies to how information can make a difference
in decision-making.
Declaration of Competing Interest
This research was made possible by the nancial support of the Inter-American Institute for Global Change Research (grant number
CRN3107) and was implemented as part of the CGIAR Research Program on Climate Change, Agriculture and Food Security (CCAFS),
which is carried out with the support of the CGIAR Trust Fund and through bilateral funding agreements. For details please see https://
ccafs.cgiar.org/donors. The views expressed in this document are those of the authors and cannot be taken to reect the ofcial
opinions of these organizations. There is no conict of interest.
Acknowledgements
We would like to acknowledge the valuable collaboration of the Secretaría de Seguridad Alimentaria (SESAN) in Guatemala
throughout the project period. We thank two anonymous referees for the useful and constructive comments on this manuscript.
A. Müller et al.

Climate Risk Management 30 (2020) 100248
11
References
6 et al., 2004 6, P., Bellamy, C., Raab, C., 2004. Data sharing and condentiality: spurs, barriers and theories. In: Paper given at the Political Studies Association
Conference, University of Lincoln, April 5–8.
Acci´
on contra el Hambre, 2015a. Evidencias para entender el impacto humanitario de las sequías de 2014 y 2015 en el Corredor Seco de Guatemala. Informe de
Situaci´
on, septiembre 2015. Guatemala.
Acci´
on contra el Hambre, Famine Early Warning System Network, USAID; 2015b. Nutrition, Mortality, Food Security, and Livelihood Survey based on SMART
Methodology. Guatemala. URL http://www.fews.net/sites/default/les/documents/reports/Guatemala SMART Survey - August 2015.pdf (accessed 2.8.18).
Acci´
on contra el Hambre, 2013. Guía para la Implementaci´
on de Sitios Centinela en Centroam´
erica. Guatemala. URL https://reliefweb.int/sites/reliefweb.int/les/
resources/Guia para la Implementacion de Sitios Centinela en Centroamerica.pdf (accessed 5.10.19).
Anderson, C.W., 1962. Central American Political Parties: a functional approach. West. Polit. Q. 15, 125. https://doi.org/10.2307/446102.
Asociaci´
on de Investigaci´
on y Estudios Sociales, 2012. Metodologías aplicadas en el Programa de Entrega de Fertilizantes -PEF- en Guatemala 2000-2012. URL http://
54.208.222.19/cgi-bin/koha/opac-detail.pl?biblionumber=524 (accessed 11.23.18).
Avelino, Jacques, Cristancho, Marco, Georgiou, Selena, Imbach, Pablo, Aguilar, Lorena, Bornemann, Gustavo, L¨
aderach, Peter, Anzueto, F., Hruska, Allan J,
Morales, Carmen, Cristancho, M., Avelino, J., Georgiou, S., Imbach, P., Aguilar, L., Bornemann, G., Hruska, A.J., L¨
aderach, P., Morales, C., 2015. The coffee rust
crises in Colombia and Central America (2008–2013): impacts, plausible causes and proposed solutions. Food Sec 7, 303–321. https://doi.org/10.1007/s12571-
015-0446-9.
Bossuyt, J., 2013. Overview of the Decentralisation Process in Latin America: Main achievements, trends and future challenges (No. 148), ECDMP Discussion Paper
Series. URL www.ecdpm.org/dp148 (accessed 5.10.19).
Bouroncle, C., Gir´
on, E., Imbach, P., Müller, A., P´
erez, S., Portillo, F., Van Etten, J., 2017. Oferta y demanda de informaci´
on para la gesti ´
on de las sequías en el
Corredor Seco de Guatemala: ¿cu´
al es la percepci ´
on de los tomadores de decisiones?, CCAFS Working Paper Series. URL https://cgspace.cgiar.org/rest/
bitstreams/113944/retrieve (accessed 5.19.17).
Bouroncle, C., Müller, A., Mendez, D.C.G., Rios, D., Imbach, P.A., Gir´
on, E., Portillo, F., Boni, A., Ettenvan, J., Ramirez-Villegas, J., 2019. A systematic approach to
assess climate information products applied to agriculture and food security in Guatemala and Colombia. Clim. Serv. 16 https://doi.org/10.1016/j.
cliser.2019.100137.
Byrne, E., Sahay, S., 2007. Participatory design for social development: a South African case study on community-based health information systems. Inf. Technol. Dev.
13, 71–94. https://doi.org/10.1002/itdj.20052.
Chandra, Y., Shang, L., 2017. An RQDA-based constructivist methodology for qualitative research. Qualitative Market Res. 20, 90–112. https://doi.org/10.1108/
QMR-02-2016-0014.
Coughlan De Perez, E., Mason, S.J., 2014. Climate information for humanitarian agencies: some basic principles. Earth Perspect. 1 https://doi.org/10.1186/2194-
6434-1-11.
Diario de Centro Am´
erica, 2014. Decreto Gubernativo Número 3-2014.
Díaz, D., Blanco, A., Bajo, M., Stavraki, M., 2014. Fatalism and well-being across hispanic cultures: the social fatalism scales (SFS). Soc. Indic. Res. 124, 929–945.
https://doi.org/10.1007/s11205-014-0825-1.
Doledec, D., 2014. Analysis of the Sentinel Site Nutrition Surveillance System in Mozambique. Washington D.C. URL https://www.fantaproject.org/countries/
mozambique/sentinel-site-nutrition-surveillance-system.
Douglas, M., 1986. How Institutions Think. Syracuse University Press, Syracuse, New York.
Elo, S., Kyng¨
as, H., 2008. The qualitative content analysis. J. Advanced Nurs. 62, 107–115. https://doi.org/10.1111/j.1365-2648.2007.04569.x.
Enenkel, M., See, L., Bonifacio, R., Boken, V., Chaney, N., Vinck, P., You, L., Dutra, E., Anderson, M., 2015a. Drought and food security – improving decision-support
via new technologies and innovative collaboration. Glob. Food Sec. 4, 51–55. https://doi.org/10.1016/j.gfs.2014.08.005.
Enenkel, M., See, L., Karner, M., ´
Alvarez, M., Rogenhofer, E., Barald`
es-Vallverdú, C., Lanusse, C., Salse, N., 2015b. Food security monitoring via mobile data collection
and remote sensing: results from the Central African Republic. PLoS One. https://doi.org/10.1371/journal.pone.0142030.
Equipo Humanitario de País Guatemala, 2015. Crisis humanitaria Sequía 2015 Reporte de Situaci´
on No. 3 (28 Agosto 2015). URL https://reliefweb.int/sites/
reliefweb.int/les/resources/Reporte de Situacion No 3.pdf (accessed 5.8.18).
Ewbank, R., Perez, C., Cornish, H., Worku, M., Woldetsadik, S., 2019. Building resilience to El Ni˜
no-related drought: experiences in early warning and early action
from Nicaragua and Ethiopia. Disasters 43, S345–S367. https://doi.org/10.1111/disa.12340.
Flagg, J.A., Kirchhoff, C.J., 2018. Context matters: context-related drivers of and barriers to climate information use. Clim. Risk Manag. 20, 1–10. https://doi.org/
10.1016/j.crm.2018.01.003.
Food and Agriculture Organization of the United Nations, Acci´
on contra el Hambre, 2012. Estudio de caracterizaci´
on del Corredor Seco Centroamericano. Guatemala.
URL https://reliefweb.int/report/guatemala/estudio-de-caracterizaci´
on-del-corredor-seco-centroamericano.
Food Security Information Network, 2018. Global Report on Food Crises 2018. URL https://docs.wfp.org/api/documents/WFP-0000069227/download/?_ga=2.
98905374.1828289637.1542733896-1286683626.1542733896 (accessed 11.20.18).
Food Security Information Network, 2020. Global Report on Food Crises 2020. URL https://www.wfp.org/publications/2020-global-report-food-crises (accessed
29.7.2020).
Gobierno de Guatemala, 2015. Gobierno de Guatemala solicita apoyo internacional para enfrentar efectos de la canícula - Guatemala | ReliefWeb. URL https://
reliefweb.int/report/guatemala/gobierno-de-guatemala-solicita-apoyo-internacional-para-enfrentar-efectos-de-la-can (accessed 5.8.18).
Gobierno de Guatemala, 2014. Informe Da˜
nos Ocasionados por la Canícula Prolongada 2014.
Gobierno de Guatemala, 2008. Política Nacional De Seguridad Alimentaria y Nutricional.
Gobo, G., 2004. Sampling, representativeness and generalisability. In: Qualitative Research Practice, SAGE publications: London, Thousand Oaks, New Delhi. pp. 435-
456.
Granados, A., 2017. Preferencias de los usuarios de informaci´
on a nivel institucional, en t´
erminos de indicadores de vigilancia agroclim´
atica y de seguridad
alimentaria y nutricional, bajo tres diferentes escenarios de riesgo, en Guatemala. Universidad de Costa Rica, unpublished tesis.
Grellety, E., Luquero, F.J., Mambula, C., Adamu, H.H., Elder, G., Porten, K., 2013. Observational bias during nutrition surveillance: results of a mixed longitudinal and
cross-sectional data collection system in Northern Nigeria. PLoS One 8. https://doi.org/10.1371/journal.pone.0062767.
Hedlund, K., 2007. Slow-onset disasters: drought and food and livelihoods insecurity–Learning from previous relief and recovery responses, Active Learning Network
for Accountability and Performance in Humanitarian Action (ALNAP). URL http://www.alnap.org/resource/5243.aspx.
Iv´
an Aguilar, Saskia Caruis, Miguel ´
Angel García, Roberto Cabrera, Bernard McCaul, 2014. Worst drought in 40 years puts more than 2 million people in Central
America at risk - European Commission. URL http://ec.europa.eu/echo/eld-blogs/stories/worst-drought-40-years-puts-more-2-million-people-central-america-
risk_en (accessed 5.7.18).
Lach, D., Rayner, S., 2017. Are forecasts still for wimps? J. Southwest 59, 245–263. https://doi.org/10.1353/jsw.2017.0013.
Lemos, M.C., 2015. Usable climate knowledge for adaptive and co-managed water governance. Curr. Opin. Environ. Sustain. 12, 48–52. https://doi.org/10.1016/j.
cosust.2014.09.005.
Lemos, M.C., Kirchhoff, C.J., Ramprasad, V., 2012. Narrowing the climate information usability gap. Nat. Clim. Chang. 2, 789–794. https://doi.org/10.1038/
nclimate1614.
Luís Mosse, E., Byrne, E., 2005. The role of identity in health information systems development: a case analysis from Mozambique. Inf. Technol. Dev. 11, 227–243.
https://doi.org/10.1002/itdj.20014.
Maga˜
na, V., Amador, J.A., Medina, S., 1999. The midsummer drought over Mexico and Central America. J. Climate. https://doi.org/10.1175/1520-0442(1999)
012<1577:TMDOMA>2.0.CO;2.
A. Müller et al.

Climate Risk Management 30 (2020) 100248
12
Mock, N., Morrow, N., Papendieck, A., 2013. From complexity to food security decision-support: Novel methods of assessment and their role in enhancing the
timeliness and relevance of food and nutrition security information. Glob. Food Sec. 2, 41–49. https://doi.org/10.1016/j.gfs.2012.11.007.
Moss, R.H., Jacobs, K., Moser, S., Buizer, J., 2016. Assessing decision support systems and levels of condence to narrow the climate information “usability gap”. Clim.
Change 135, 143–155. https://doi.org/10.1007/s10584-015-1549-1.
Müller, A., Mora, V., Rojas, E., Díaz, J., Fuentes, O., Giron, E., Gaytan, A., van Etten, J., 2019. Emergency drills for agricultural drought response: a case study in
Guatemala. Disasters 43, 410–430. https://doi.org/10.1111/disa.12316.
Palmer, T.N., Weisheimer, A., 2012. On the Reliability of Seasonal Forecasts, ECMWF Seminar on Seasonal Prediction. URL https://www.ecmwf.int/sites/default/
les/elibrary/2013/11531-reliability-seasonal-forecasts.pdf (accessed 4.26.18).
Prensa Libre, 2014. Faltan recursos para enfrentar efectos por sequía – Guatemala, 20 de Septiembre 2014 [WWW Document]. URL https://reliefweb.int/report/
guatemala/faltan-recursos-para-enfrentar-efectos-por-sequ (accessed 5.7.18).
Pulwarty, R.S., Sivakumar, M.V.K., 2014. Information systems in a changing climate: Early warnings and drought risk management. Weather Clim. Extrem. 3, 14–21.
https://doi.org/10.1016/j.wace.2014.03.005.
Rayner, S., Lach, D., Ingram, H., 2005. Weather forecasts are for wimps: Why water resource managers do not use climate forecasts. Clim. Change 69, 197–227.
https://doi.org/10.1007/s10584-005-3148-z.
Rosas, G., Gubler, S., Oria, C., Acu˜
na, D., Avalos, G., Begert, M., Castillo, E., Croci-Maspoli, M., Cubas, F., Dapozzo, M., Díaz, A., Van Geijtenbeek, D., Jacques, M.,
Konzelmann, T., Lavado, W., Matos, A., Mauchle, F., Rohrer, M., Rossa, A., Scherrer, S.C., Valdez, M., Valverde, M., Villar, G., Villegas, E., 2016. Towards
implementing climate services in Peru. The project CLIMANDES. Clim. Serv. 30–41 https://doi.org/10.1016/j.cliser.2016.10.001.
Sandefur, J., Glassman, A., 2015. The political economy of bad data: evidence from African Survey and Administrative Statistics. J. Dev. Stud. 51, 116–132. https://
doi.org/10.1080/00220388.2014.968138.
Saylor Mase, A., Prokopy, L.S., 2014. Unrealized potential: a review of perceptions and use of weather and climate information in agricultural decision making.
Weather. Clim. Soc. 6, 47–60. https://doi.org/10.1175/WCAS-D-12-00062.1.
Secretaría de Seguridad Alimentaria y Nutricional; 2014. Estrategia operativa para la implementaci´
on del sistema de vigilancia y alerta temprana en SAN, basado en la
metodología de Sitios Centinela.
Secretaría de Seguridad Alimentaria y Nutricional; 2015. Pron´
osticode Seguridad Alimentaria y Nutricional. Período: Junio a Agosto 2015. URL http://www.siinsan.
gob.gt/siinsan/wp-content/uploads/Monitoreo_y_Evaluacion/Pronostico/INFORMES2015/f1ff072e5af5e4525f2f4b0a679a2924.pdf (accessed 5.13.19).
Soares, M.B., Dessai, S., 2016. Barriers and enablers to the use of seasonal climate forecasts amongst organisations in Europe. Clim. Change 89–103. https://doi.org/
10.1007/s10584-016-1671-8.
Srinivasan, G., May Rasura, A., Subbiah, A.R., 2011. Climate information requirement for community-level risk management and adaptation. Clim. Res. Clim Res 47,
5–12. https://doi.org/10.3354/cr00962.
Thompson, M., Wildavsky, A., 1986. A cultural theory of information bias in organizations. J. Manag. Stud. 23, 273–286. https://doi.org/10.1111/j.1467-6486.1986.
tb00954.x.
Universidad Rafael Landívar, Universidad de San Carlos de Guatemala, 2013. Evaluaci´
on del Programa de Fertilizantes del Ministerio de Agricultura, Ganadería y
Alimentaci´
on (MAGA). URL https://www.url.edu.gt/publicacionesurl/FileCS.ashx?Id=40201.
Vogel, C., OBrian, K., 2006. Who can eat information? Examining the effectiveness of seasonal climate forecast and regional climate-risk managemen strategies. Clim.
Res. 33, 111–122.
Wang, R.Y., Strong, D.M., 1996. Beyond accuracy: what data quality means to data consumers. Source J. Manag. Inf. Syst. 12, 5–33.
Willis, E., Garman, C. da C.B., Haggard, S., 1991. The politics of decentralization in Latin America. Lat. Am. Res. Rev. 34, 7–56.
World Food Programme, 2015. Initial Analysis of the Impact of the Drouht on Food Security in Guatemala, El Salvador, and Honduras. URL http://documents.wfp.
org/stellent/groups/public/documents/ena/wfp277948.pdf?_ga=2.169760097.1134511981.1518082981-1736649751.1518082981 (accessed 2.8.18).
A. Müller et al.