Chapter Three: Methodology.

This chapter focuses on the model, methods, and techniques that researchers used to collect and
analyze data .it greatly concerns the research model, modeling techniques; data collection
techniques, methods of data analysis ethical issues such as reliability and validity measures.
3.1. Research design
this research design aims to evaluate the impacts of the various factors on household aggregate
consumption within the context of the Vision 2020 umurenge program (VUP) in Rwanda. The
study utilizes both theoretical and empirical frameworks to analyze how household
characteristics and program participation influence aggregate consumption.
3.2. Hypothesis
H1: Participation in the VUP program leads to a significant decrease in household aggregate
consumption.

H2: Location significantly influences household aggregate consumption.

H3: Household characteristics such as size, location, and the age of the household head
significantly influence household aggregate consumption.

H4: The sex of the household head significantly influences household aggregate consumption.

3.3. Data collection methods.

This study utilizes secondary data provided by the National Institute of Statistics of Rwanda
(NISR) (EICV5-VUP Report 2016) , from the integrated household living conditions survey
(EICV5). (EICV5) is a nationally representative survey conducted by NISR, designed to assess
various aspects of household living conditions, socioeconomic status, and welfare across
Rwanda. Secondary data collection involves extracting and analyzing information already
gathered by NISR, focusing on indicators relevant to the Vision 2020 Umurenge program and its
impact on household expenditure. By leveraging EICV5 data, this study ensures comprehensive
coverage and statistical rigor in evaluating the long-term effects of the VUP on household
economic behaviors and welfare outcomes.
 3.3.1. Variables
 Independent Variable:
 Vup participation.
 Member of Household in terms of household size
 Location of Household (Rural or Urban)
 Age of head of household
 Ubudehe category
 Sex of head of household.

 Dependent Variable:
 Household Expenditure: We can utilize various expenditure categories from
EICV5 data relevant to the research question. This might include Aggregate
consumption.
3.4. Data analysis methods.
 Descriptive Statistics: Analyze the distribution of VUP participation and expenditure
variables across different household characteristics (e.g., location, poverty level,
household size).
 Regression Analysis: We can run regression models with household expenditure as the
dependent variable and VUP participation as the independent variable, controlling for
relevant household characteristics like location, poverty level, and household size. This
can help assess the association between VUP and expenditure while accounting for the
influence of other factors.
3.5. Model specification
Model specification is the process in which we choose variables used in a given statistical and
mathematical model building. This research aims to analyze the impact of the vision Umurenge
program (VUP) on household expenditure partners in Rwanda using secondary Quantitative data.
3.5.1. Ordinal least squares regression model (OLS)
Ordinary Least Squares (OLS) regression is a workhorse technique in statistics for understanding
how one variable (independent) influences another (dependent). It achieves this by fitting a
straight line through a scatter plot of the data, minimizing the distance between the line and the
data points. This line represents the overall trend, with the slope indicating how much the
dependent variable changes on average with a one-unit increase in the independent variable. OLS
is particularly useful when the relationship between the variables is assumed to be linear and the
error terms (unexplained variations) behave in a specific way. However, it's crucial to consider
these assumptions and explore alternative regression methods if your data deviates from linearity
or the error terms violate the expected properties. the following are models specified regarding
our studies:

γ i=β 0 + β 1 (Age)+ β 2 (Sex Hhead)i + β 3 (HHcategory)i + β 4 (location)i+ β5 ¿

Where:

 γ i stands for aggregate consumption as the dependent variable.

 Sex Hheadi = 1: Male

= 0: Female

 HHcategoryi = 1: People in 5th category (who are described as poor)

= 0: People in other categories
 Locationi = 1: Rural
= 0: Urban
 Vup participationi = 1: vup participants
= 0: non vup participants
 Chapter Four: Data analysis, results, and interpretation

4.1 Results.
In this section, we delve into a comprehensive analysis of the secondary data collected, from the
National Institute Statistics of Rwanda (NISR) on ELCV5, but also regarding the impact of the
Vision Umurenge Program (VUP) on household expenditure patterns. The findings shed light on
how various factors, such as household size, Vup participation, Ubudehe VUP category, and
urban/rural classification, interact to influence aggregate consumption habits among
beneficiaries. Additionally, the findings highlight differences in spending between urban and
rural areas within the context of VUP interventions, providing valuable insights into the
program's effectiveness in addressing economic disparities and promoting sustainable
livelihoods.

4.1.1. Descriptive statistics

Table 1 Descriptive statistics

Variable Obs Mean Std. Dev. Min Max

aggregate consumption 64314 1719264 2981374.9 26588.279 83437800
age Hhead 14580 45.157 15.641 14 109
Sex Hhead 44392 .245 .43 0 1
HH category 64287 2.631 1.427 1 5
location 64314 .833 .373 0 1
member 64314 5.428 2.159 1 22
Vup Participation 64314 .116 .32 0 1

The descriptive statistics provided above offer key insights derived from analyzing a dataset
comprising 471,952 observations, providing a comprehensive understanding of household
dynamics. Our analysis begins by examining household expenditures, revealing an average
monthly spending of around 668,797.2 on food and 83,972.94 on non-food items, shedding light
on economic behaviors within the dataset. Additionally, the analysis offers valuable insights into
household composition, with an average household size of 6.15 members, providing a nuanced
understanding of family structures and resource allocation. Geographical distribution emerges as
a crucial factor, with a mean location score of 1.85 indicating a predominantly urban setting.
Further exploration uncovers complex relationships within households, alongside diverse
employment scenarios reflected in a mean occupation status of 3.47. Noteworthy is the
significant impact of certain programs, benefiting numerous households and enhancing financial
resilience. Insights into socioeconomic stratification reveal a moderate level (mean of 2.5),
offering context on the broader socio-economic landscape. Additionally, on average Head of
household is 33 years old in VUP activities, and the household's history of receiving income or
loans from VUP is around 3.68 on average not only that the diverse household categories within
Ubudehe VUP (on average are two categories) further underscore the program's inclusive reach
across varied socio-economic profiles.

4.1.2. Multiple regression analysis

Model development:

γ i=β 0 + β 1 (Age)+ β 2 (Sex Hhead)i + β 3 (HHcategory)i + β 4 (location)i+ β5 ¿

Where:

 γ i stands for aggregate consumption as the dependent variable.

 Sex Hheadi = 1: Male

= 0: Female

 HHcategoryi = 1: People in 5th category (who are described as poor)

= 0: People in other categories
 Locationi = 1: Rural
= 0: Urban
 Vup participation = 1: vup participants
= 0: non vup participants
Table 2 multiple regression model

Aggregate Coef. St. Err. t-value p-value [95% Conf Interval] Sig
consumption
Age Hhead 9310.339 1293.033 7.20 0 6775.829 11844.848 ***
Sex Hhead 152220.72 46657.455 3.26 .001 60766.192 243675.26 ***
HHcategory 126003.76 11574.46 10.89 0 103316.35 148691.17 ***
location -2166476.1 49224.733 -44.01 .002 -2262962.8 -2069989.4 ***
Member of HH 284150.01 9151.102 31.05 0 266212.69 302087.33 ***
Vup participation -240718.82 58536.892 -4.11 0 -355458.55 -125979.08 ***
Constant 1131263.8 96979.015 11.67 0 941172.63 1321355 ***

Mean dependent var 1441180.533 SD dependent var 2434527.440

R-squared 0.588 Number of obs 14572

F-test 562.403 Prob > F 0.000

Akaike crit. (AIC) 466900.237 Bayesian crit. (BIC) 466953.345

*** p<.01, ** p<.05, * p<.1

This output presents the results of a regression analysis conducted on variables related to the
VUP Program. The analysis includes the Age of the Household Head, Sex of the Household
Head, Household Category, Location, Household Size, and VUP Participation as independent
variables, with Aggregate Consumption as the dependent variable.

The model's statistical significance is evident from the F-statistic of 562.403, with a p-value of
0.000, indicating a highly significant relationship between the independent variables and
Aggregate Consumption. The R-squared value of 0.588 suggests that approximately 59% of the
variability in Aggregate Consumption is explained by the independent variables in the model.

4.1.3. Interpreting the coefficients:

Age of Household Head: For every additional year in the age of the household head, Aggregate
Consumption is expected to increase by approximately 9,310.339 RWF, holding all other
variables constant.

Sex of Household Head: Households with a male head are associated with an increase in
Aggregate Consumption by approximately 152,220.72 RWF compared to those with a female
head, assuming other factors remain constant.
Household Category: Being in a higher household category (i.e., poorer households) is
associated with an increase in Aggregate Consumption by approximately 126,003.76 RWF,
compared to the reference category, while keeping other variables constant.

Location: Being in a rural location is associated with a decrease in Aggregate Consumption by

approximately 2,166,476.1 RWF, compared to an urban location, holding other variables
constant.

Household Size: Each additional member in the household is associated with an increase in
Aggregate Consumption by approximately 284,150.01 RWF, assuming other factors remain
constant. This effect is highly significant (p < 0.01).

VUP Participation: Participation in the VUP program is associated with a decrease in Aggregate
Consumption by approximately 240,718.82 RWF compared to non-participation, holding other
variables constant. This effect is highly significant (p < 0.01).

Chapter Five: Discussion, conclusion, and Recommendation

5.1. Discussion
Based on the findings from the linear regression equation (22-28 n.d.) the socio-economic
development of women beneficiaries in the Rusiga district is consistently influenced by a
constant factor of 5.206, irrespective of other impacts. Additionally, the variables indicate that
each unit change in VUP public works, VUP direct support, VUP financial services such as
loans, and income from socio-economic activities significantly impact the socio-economic
development of these women, with changes of 0.407 and 0.089 and a standard error of 0.988 in
the model. The research also shows that wages are sometimes used for care work or as
contributions to savings groups. Not only that (Gahamanyi and Kettlewell 2015) , just highlight
that VUP transfers were most commonly spent on food (75 percent of households reporting that
they had bought food), other consumables (reported by 38 percent of the sample), farm
investment, livestock purchase (36 percent), and education (19 percent). Accumulating savings is
an important element of the journey towards graduation. In the 2014 Household Survey, 28
percent of the households that received Direct Support reported having saved part of their VUP
income, while 37 percent of Public Works participants had made some savings. The average
amount saved is RwF 28,468 (US42). These percentages increase when the beneficiaries also
receive a loan under the Financial Services component, to 33 percent and 46 percent,
respectively, (Stanley et al. 2017).

Came up with our analysis of variables related to the Vision 2020 Umurenge Program (VUP) in
Rwanda reveals several significant findings regarding household aggregate consumption. Age of
the household head positively influences consumption, with each additional year correlating to
an increase of approximately 9,310.339 RWF. Households headed by males show significantly
higher consumption compared to those headed by females. Higher household categories,
indicative of greater poverty, are associated with increased consumption, underscoring VUP's
targeted support. Conversely, rural location is linked to decreased consumption by about
2,166,476.1 RWF compared to urban areas. Larger household sizes positively impact
consumption, with each member contributing approximately 284,150.01 RWF. Surprisingly,
VUP participation correlates with decreased consumption by roughly 240,718.82 RWF,
suggesting nuanced economic dynamics despite program benefits. These insights emphasize the
need for tailored policies addressing regional disparities, gender roles, and household economic
conditions to optimize poverty alleviation strategies in Rwanda.

5.2. Limitation
While the analysis of the Vision Umurenge Program (VUP) and its impact on household
expenditure patterns provides valuable insights, several limitations must be acknowledged. The
reliance on secondary data from the National Institute of Statistics of Rwanda (NISR) introduces
potential data quality issues such as inaccuracies or missing values, impacting the reliability of
results. Furthermore, the analysis establishes associations but not causality, and external factors
like economic shifts or policy changes could influence household spending independently of the
VUP Program. Sample representativeness and the lack of qualitative data also limit the
generalizability and depth of the findings, while the simplified regression model may overlook
complex interactions and nonlinear relationships among variables. Future research could address
these limitations by incorporating more diverse data sources, longitudinal analyses, and
qualitative insights to provide a comprehensive understanding of the program's impact on
household expenditure dynamics.
 5.3. Conclusion
In conclusion, this study has provided a comprehensive analysis of the impact of the Vision
Umurenge Program (VUP) on household expenditure patterns in Rwanda. Through meticulous
examination of secondary data obtained from the National Institute of Statistics of Rwanda
(NISR), valuable insights have been uncovered regarding the various factors influencing
household spending behaviors within the context of VUP interventions.

The findings from descriptive statistics illuminated significant aspects of average food and non-
aggregate consumption among households, coupled with demographic and socio-economic
characteristics like household size, location, occupation status, and participation in the VUP
program. These statistics laid a robust foundation for a more detailed exploration through
multiple regression analysis.

The regression analysis uncovered a statistically significant relationship between independent

variables such as monthly non-aggregate consumption, household size, location, occupation
status, HH receives INC Loan, Ubudehe category, age of household, and household category in
Ubudehe VUP, with aggregate consumption as the dependent variable. The coefficients and
interpretations for each variable provided actionable insights into how these factors shape
household spending patterns.

However, it is essential to acknowledge the limitations inherent in this study. The reliance on
secondary data introduces potential data quality issues, and the analysis establishes associations
rather than causality. Additionally, sample representativeness, the absence of qualitative data, and
the simplified regression model may limit the generalizability and depth of the findings.

Moving forward, this study offers several recommendations to stakeholders and policymakers.
These recommendations include enhancing targeting mechanisms within the VUP program,
tailoring interventions based on household characteristics, implementing robust monitoring and
evaluation mechanisms, promoting financial inclusion, advocating for sustainable funding,
encouraging further research, and fostering community engagement.

In conclusion, this study contributes significantly to the existing literature by providing nuanced
insights into how social protection programs like VUP can impact household expenditure
dynamics and promote sustainable livelihoods. It lays the groundwork for future research and
policy initiatives aimed at enhancing economic resilience and reducing poverty among
vulnerable households in Rwanda and similar contexts.

5.4. Recommendation
Based on the analysis of the Vision Umurenge Program (VUP) and its impact on household
expenditure patterns, several key recommendations can be made to various stakeholders and
audiences:

1. Policymakers and Government Authorities:

 Strengthen Targeting: Utilize the findings to improve targeting mechanisms within the
VUP Program, ensuring that resources are directed towards households most in need of
support.

 Enhance Program Flexibility: Consider adapting the program to accommodate diverse

household needs and changing socio-economic conditions, allowing for greater
effectiveness and inclusivity.

2. Program Implementers and NGOs:

 Tailored Interventions: Develop tailored interventions based on household characteristics

identified in the analysis, such as household size, location, and socio-economic category,
to maximize impact.

 Monitoring and Evaluation: Implement robust monitoring and evaluation mechanisms to

continuously assess the program's impact and make data-driven adjustments for better
outcomes.

3. Financial Institutions and Donors:

 Financial Inclusion: Collaborate with the VUP Program to promote financial inclusion
initiatives, providing access to credit and financial services to households, especially
those in rural areas.
  Sustainable Funding: Advocate for sustainable funding mechanisms to ensure the
continuity and expansion of programs like VUP, emphasizing long-term impact and
sustainability.

4. Researchers and Academia:

 Further Research: Encourage further research and analysis to deepen understanding of the
complex interactions between socio-economic factors and household expenditure,
informing future policy and program design.

 Dissemination of Findings: Share research findings and insights with relevant

stakeholders and policymakers to facilitate evidence-based decision-making and policy
formulation.

5. Community Leaders and Civil Society Organizations:

 Community Engagement: Foster community engagement and participation in program

design and implementation, ensuring that the VUP Program reflects the actual needs and
priorities of beneficiaries.

 Advocacy and Awareness: Advocate for increased awareness and understanding of social
protection programs like VUP among communities, promoting uptake and utilization of
available resources.

Overall, the recommendations aim to leverage the analysis findings to optimize the effectiveness,
inclusivity, and sustainability of the VUP Program and similar initiatives, ultimately contributing
to improved livelihoods, reduced poverty, and enhanced economic resilience among vulnerable
households.