Modeling Performance of Micro and Small Manufacturing Enterprises [602425]

1
Modeling Performance of Micro and Small Manufacturing Enterprises
Conditional to entire constraints in Northern Ethiopia, Case Study in Adigrat
City Tigray, Ethiopia.

ABSTRACT
Background: Certainly , the thoughtfultricky in urban areas of emerging countries like Ethiopia
is scarcity and deficiency which could be expressed in terms of joblessness and low revenue .
Hence, Adigrat is one of the city in northern Ethiopia suffering from same problem. The
overwhelming solution that could be taken is enhancing micro and small industries of the which
in turn are essential catalyst and tools for growth, jo b creation and s ocial progress to generate
more equitable income distribution.
Objective: The aim of this study was to model performance of micro and small -scale
manufacturing industries conditional to entire constraints in Adigrat city.
Methods and tools: The research has been designed in cross -sectional format using stratified
sampling techniques in six different manufacturing enterprises (strata). Total of 127
respondents, 122 out of target population 648 including has been taken using systematic
sampling techniques.A fter collecting primary data through well -administered questionnaires
multiple linear regressions was modeled .
Results and findings: Out of six categorical and one continues predictors included in multiple
linear regression model , except lack of raw materi als and lack of demand, all the rest predictors
have significant influence on monthly revenue which in turn affect t he growth performance of
MSEs. The overall regression model is statistical ly significant ( 𝑃−𝑣𝑎𝑙𝑢𝑒≤0.001) and 72.3%
of the variability is explained by model.
Conclusions: Generally, it can be concluded that , experience of the enterprises , lack of initial
capital, lack of market linkage , lack of access to latest technology, and power interruption are
the major constraintson performance of manufacturing enterprise in Adigrat city.
Keywords: Capital; Growth Performance ;MSE ;Multiple linear regressions, Revenue ;

1
INTRODUCTION
In many developing countries, micro and small -scale manufacturing enterprises account for the
majority of firms and a large share of employment, mainly consisting of small firms with one
person working alone or with unpaid family members. Self -employment is a central element in
these economies .
In fact , the thoughtful tricky in urban areas of emerging countries like Ethiopia is scarcity which
could be expressed in terms of joblessness and low revenue. Hence, Adigrat is one of the city in
northern Ethiopia suffering from same problem. The irresistible solutio n that could be taken is
enhancing micro and small enterprises of the city due to the reasons that MSEs and the small
business sector are essential catalyst and tools for growth, job creation and social progress to
generate more equitable income distributi on. Thus, micro and small enterprises are expected to
play a significant role in the national economic development, particularly, in the creation of
employment opportunities and poverty reduction. However, in Ethiopia their contribution is
very low when c ompared to other developing countries. This may be due to different reasons;
for instance, in developing countries, lack of capital is seen as the real factor that is limiting the
expansion of businesses at least at the micro -enterprise level of operation (Abay et al.,
2014). Keeping this in mind, the federal government of Ethiopia ha d articulated its first micro
and small enterprise development strategy in 1997.
As an experience, m icro, small and medium enterprises (MSMEs) hold a strategic position in
the Indonesian economy. It can even be said that they are the backbone of the economy,
constituting 98 percent of all business units in Indonesia. MSMEs have the ability to use
production resources efficiently, create job opportunities, and improve income dis tribution. In
2006, the output of MSMEs reached almost 58% of the total Gross Domestic Product (GDP)
and absorbed 82% of the total workforce; more than 73 million people (Biretawi, 2013 ).
Although the Tanzanian financial system comprises a host of both formal and informal
organizations, adequate finance is not being provided to allow for the development of the sector.
In both the 1991 and 1995 informal sector surveys, lack of capital was ci ted as pressing need of
the MSE sector operators. The surveys further indicated that working capital (necessary for
business growth) was the most needed followed by investment capital (for starting up new

2
business). It was established from the surveys that only 5% could obtain credit from Banks,
NGOs, and other financing institution, but the rest (95%) were from own saving (66%)
borrowing from friends and relatives (12%) and assistance from friends and relative (17%) in
(Biretawi, 2013 ).
Asgedom (2013) noted that the enterprises face difficulties in getting loans from the financial
institutions due to high collateral requirements and the plant has suffered from inadequacy of
machineries and lack of modern technology . MSEs are constrained by shortage of skilled labor
force, lack of business premises, poor product quality, and shortage of raw material, lack of
business information and training and burden of some bureaucratic and high taxation.
Several questions raise in the mind of authors in context of the constraints that mainly influence
the performance of micro and small -scale manufacturing industries in emerging countries. One
may ask himself that, w hat are the key challenges affecting the growth of micro and small
manufacturing enterprises in Adigr at city?How do the government and other stakeholders
contribute in MSEs to play their role in poverty alleviation? Therefore, these kinds of questions
are expected to be answered at the end of the study.
The development of small and micro enterprises is hi ndered by various problems among the
problems lack of capital for commence, operation and expansion can be mentioned. Small
enterprises even at the initial stages just after starting operation do require adequate promotional
services; however, most of them cannot afford to purchase due to capital shortage (Walmobo,
1996)
Moreover, f inance is an engine for both new entry to the sector and existing MSEs operators.
There are several studies that focused on the financial aspect of MSEs through his study in
MSE s’ profitability, showed that the profitability of micro and small enterprises mainly depends
on the initial capital, the low the initial capital results low productivity and in turn to low earning
per employ and vice versa. Moreover, Walmobo (1996) small enterprises especially at the initial
stages just after starting operation do require adequate promotional services however, most of
them cannot afford to purchase due to capital shortage. Different research evidenced that the
small firms start their busin ess with their own savings supplemented by borrowing from friends
and relatives. Since most of the operators/owners are poor, they start their business with very

3
little capital. A few meet their capital requirements through informal credit mechanisms, whic h
exist within their community, but rarely from the formal sector institutions (Abay et al., 2014).
In developing countries, MSEs operators do not seem to try out market niches for improves or
slightly diversified products. Moreover, since they mostly copy what neighbors and friends
produced, market and profit margins are limited (Walomm, 1996) micro and small enterprises
entrepreneur faces problems in the matter of marketing his product. Due to want to adequate co –
operative or other marketing facilities an d intelligence often suffers from remunerative price of
their goods in the open market (Burn, 2001). In free market economy, especially with stiff
competition marketing is a key factor for the success of small business (Gebretinsae, 2003)
Furthermore, new technologies improve efficiency, enable greater production, and are a source
of profit for SMEs. Technological capabilities benefit SMEs in several ways. They enhance
SME efficiency, reduce costs, and broaden market share, both locally and globally. Same
scholars argue that a small business that adopts greater levels of technological sophistication
can be expected to grow more rapidly than a similar firm that does not (Asgedom, 201 3).
The remained part of this paper is organized as follows. Section 2 introduces the methods and
framework which are going to be used in the whole paper. Section 3 talks about the results of
multiple linear regressions and general discussion of the article. F inally, Section 4 presents
conclusions and recommendations of the study.
METHODOLOGY OF THE S TUDY
TARGET POPULATION
The target population of this study has covered from two kebeles’ micro and small
manufacturing enterprises and from the MSE governmental officials and head municipal of the
city. The reason why the researcher has drawn the target population from two kebel es such as 02
and 04 was that purposively based on number of operators and diversity of businesses in each
kebeles. In addition to this core, persons from MSE governmental officials and head of the city
municipal were selected purposively for interview, a s they are the key informants to the research.

4
RESEARCH DESIGN
Cooper and Schindler (2008) refer to research design as the plan and structure of investigation so
conceived as to obtain answers to the research questions. Moreover, Kothari (2004) contends tha t
the research design describes the arrangement of conditions for collection and analysis of data,
bringing together the relationship and rationale of the study as a means to achieve the research
objectives using empirical evidence obtained economically.In summary, a research design is a
master plan that specifies the methods and procedures for collecting and analyzing the needed
information.
Accordingly, t he cross -sectional research design has been conducted using stratified sampling
techniques in six diff erent manufacturing enterprises (strata). Therefore, out of the total number
of 648 target population, 122 respondents have been taken using systematic sampling techniques.
Both qualitative and quantitative data were included using primary data source in order to
modelthe predetermined predictors on performance of MSEs . The well -administered
questionnaires were distributed to the respondents in systematic way to minimize the subjective
biases. Finally, the relevant and most widely used statistical mode ls so known as multiple linear
regressions for continues response was applied to assess the relationship between the response
variable monthly revenue , which is the main indicatorof MSE and associated constraints so-
called predictor s.
SAMPLE SIZE DETERMIN ATION
Taking sample size have a great advantage over census in different studies to eliminate the
wastage of time as well as cost and resources; therefore, the investigators always take a sub -part
of populations which must represents the information of popu lations as whole. In order to have
adequate information to achieve the reasonable results and conclusions it must to have enough
sample size which represents the information of populations as whole. According to the sample
size formula published by Dillman (2007), the sample size will be determined (quoted by Fissuh,
et al., 2016) .

5
n=N∗P(1−P)
N−1 ∗ d
Zα2 2
+P(1−P) , (1)
where n is total sample, size needed for desired level of precision N is total size of population P
is chance that any respo ndent will answer a question the same as any other respondent d is
acceptable amount of sampling error (margin error) Zα2 is statistic associated with the
confidence level.
To have maximum sample size P=0.5, d=0.08 or 8% and α=0.05 or 5% level of
signific ance was taken and therefore, Zα2 will be equal to 1. 96 according to standard normal
table. The desired total sample size n will be proportionally allocated in to strata sample sizes. In
Accordingly, to allocate sample size for each stratum the general f ormula will be used as
follows.
nh=nNh
N . (2)
Notations: n=the overall sample size: n= nhh
i=1,
nh= The sample size in each stratum ,
Nh= The total number of populations in the hth stratum and N is total size of study population.
METHODS OF DATA COLL ECTION
In any research, different sources of data are implemented to respond the research questions
correctly. For this study, the researcher has used both primary and secondary data sou rces. The
main source of primary data was micro and small enterprises survey .
One important of statistical survey is questionnaire method. The simplest definition of
questionnaire is that of a group or sequence of question. Thus, f or this study ,well-administered
questionnaire wasdistributed to 122 respondents to collect relevant information from the sample
taken from target population. Furthermore, the review of relevant documents or files, annual

6
abstract and literatures was conducted to have adequa te information about main constraints
influencing on the performance of MSEs growth . In order to elicit information up on subject from
respondents, similar questionnaire for all respondents was prepared. both open and close -ended
questionnaire have been incl uded.
METHODS OF DATA ANAL YSIS
On the analysis stage, both quantitative and qualitative data were analyzed. Some relevant
statistical analysis was implemented based on the characteristics of data. In this study to observe
and overview, the general features of the data on each variable some descriptive as well as
inferential statistics are employed. Descriptive statistics is a part of statistics that describes or
explains the characteristics of sample data without generalizations or drawing conclusions about
the characteristics of populations (Fissuh, et al., 2016) . The techniques most of in used for
describing the characteristics of the sample and the major study variables are displayed in the
form of frequency distribution, percentage, proportion, and diagra mmatic representation such as
bar chart, pie chart etc. Inferential statistics is statistical method deals with making inference or
conclusion about population based on data obtained from a limited number of observations that
come from the population. Infe rential statistics consists of estimation and hypothesis testing
(Fissuh, et al., 2016) . Then with techniques of statistical analysis multiple linear regression was
conducted in the research paper. Inferential statistics consists of estimation and hypothes is
testing. Therefore, to model the performance of micro and small enterprises which is measure by
continues response (Monthly Revenue of MSE) conditional to predetermined constraints multiple
linear regression was proposed.
MULTIVARIATE LINEAR REGRESSION MODELS
Multiple linear regressions are the most common form of linear regression analysis. As a
predictive analysis, the multiple linear regressions are used to explain the relationship between
one continuous dependent variable and one or more independent variables (Agresti, 1990 , quoted
by Fissuh, et al. ,2016 ). Hence, in this paper, the response variable monthly revenue of firm
versus some predefined predictors or challenging factors including one continues predictor
experience (life time) of the firm and 6 categorical factors/predictors such as, technology, power

7
interruption, access to row materials, initial capital, lack of demand and Market linkage.
Therefore, the variables are defined in the table below.
Table 3.1 Variable descriptions and creating dummy variables for multiples lin ear regression
Continues variables
Name of Variables Measurement Scale
Monthly Revenue of the Firm (Yi) Ratio Scale/continues
Experience (Life time of Firm) (X1i) Ratio scale/continues
Categorical Predictors
Name of
Variables Measurement
Scale Category levels (K-1) Dummy/Indicator Variables K
indicates number of categories
Technology used
in the given firm Ordinal 1=Latest, 2=Outdated,
3=Manual (Reference
level) X2i= 1,if latest
0,otherwise
X3i= 1,if Outdated
0, Otherwise

Sequence of
Power
Interruption Ordinal 1=Frequently, 2=Rarely,
3=Not at all (Reference
level) X4i= 1,if Frequently
0,Otherwise
X5i= 1, if Rarely
0, Otherwise

Raw Materials Nominal 0= No, 1=Yes X6i= 1, if Yes
0,Otherwise
Initial Capital of
given firm Ordinal >50,000=1, 20,001 –
50,000=2, <20,000=3 X7i= 1, if>50,000
0, Otherwise
X8i= 1,if 20,001−50,000
0, Otherwise
Demand Nominal 0= No, 1= Yes X9i= 1, if Yes
0,Otherwise
Market Linkage Ordinal 1=High, 2= Fair, 3= Low X10i= 1, if High
0, Otherwise
X11i= 1, if Fair
0, Otherwise

8
Based on the defined variable in table 3.1, the following multiple linear regression model is
formulated for to predict the relationship between monthly revenue, which is the one of the
indicator of MSE growth and predictor/independent variables (Challenging factors)
Yi=βo + β1X 1i+β2X 2i+⋯+βkX ki+∈i (3)
Where, Yi dependent variable (Monthly Revenue), βi′s are the estimated coefficients or
parameters, X ki are fixed known predictors or independent variables and ∈i ~N μ,σ2 random
errors k=1,2,…,11 &𝑖=1,2,…,𝑛=122.
General concept: Linear Regression is a process that allows you to make predictions about
variable “Y” based on knowledge you have about variable “X”. The Correlation Coefficient is a
single summary number that tells you whether a relationship exists between two or more
variables, how strong that relationship is and whether the relationship is positive or negative
(direction of relati onship). The Coefficient of Determination is a single summary number that
tells you how much variation in one variable is directly related to variation in another variable.
The Standard Error of Estimate is a single summary number that allows you to tell how accurate
your predictions are likely to be when you perform Linear Regression (Agresti, 1990 ; quoted by
Fissuh, et al.,2016 ).
RESULTS AND DISCUSSION
RESULTS
DATA DESCRIPTIONS AN D GENERAL CHARACTERI STICS
This study generally deals with modeling theperformance of micro and small enterprises which
is mainly rated by monthly revenue of the micro and small -scale manufacturing enterprise
holders in Adigrat city, Tigray, Ethiopia. The study is done based on a number of literature
dealing with MSEs’ major constraints . Cross sectional, study design was applied with careful
investigation about main constraints influencingperformance of MSEs. Particularly, the study is
conducted in manufacturing enterprises in Adigrat city. Generally, after the careful and
intenti onal collection of data, the investigator has come up with well -organized methodological
and empirical results.

9
Results of this study are based on the objectives and analysis of the sample answers of the
respondents from MSEs operators and management bod ies. The result incorporated both
descriptive and inferential statistics accompanied by empirical analysis and are organized in
respective tables and graphs. The careful interpretations, discussions and conclusions were
conducted.
Table 4. 1 Percentage of MSE owners who got support from NGOs and stalk holders
Type of Support Frequency Percent Valid % Cumulative %
Valid Training 17 13.9 13.9 13.9
Financial 11 9 9 23
Information 11 9 9 32
Others 3 2.5 2.5 34.4
No support 80 65.6 65.6 100
Total 122 100 100
Sources: Own Survey data (2017)
Table 4. 1 signified the proportion of support for MSE owners from NGOs and stakeholders. The
majority (65.6%) of the respondents said they did not get any support. Approximately 14% of
them said they got training support whereas the small and equal proportion that is 9% of MSE
owners were given financial and information support.
Thus, the result in turn elaborates the gap of the government and other concerned sectors in
supporting the motivated new business starters in capital, in job creativity, in technical skill
improvement trainings and motivations or incentives. Hence, based on this fact, it is possible to
conclude that there has been very serious problem to setup a new business in Adigrat City. This
is due to inadequate support from the government, and NGOs and stalk holders.
Therefore, it is recommended that the government and other concerned sectors should take
careful look and attention to solve these serious challenges of several motive individuals at
starting enterprises.

10

Figure 4. 1 initial capital of the enterprise (Sources: Own Survey data, 2017)
According to figure 4. 1, the pie chart clearly illustrates that the percentage of the initial capital of
the individual enterprises. Therefore, the chart shows that more than half (63.11% ) of the
respondents started their business with low capital which is below 20,000 ETB. However, the
rest less than half individual enterprise beginners, started with the capital of between 20,001 and
50,000ETB; that is about 29.51% of total respondents an d very few of them about 7.38% of the
respondents were started their business with initial capital above 50,000 ETB. Therefore, it is
true to say that, most of micro and small enterprise beginners in Adigrat City were started their
business with very low c apital. This fact in turn implies that in recent economy of the country, it
is not easy to start the manufacturing business with the mentioned insignificant amount of
capital.

11

Figure 4. 2 types of Enterprise by sector (Sources: Own Survey data, 2017)
In fact, the Ethiopian government is giving special attention to the growth -oriented sectors
specially manufacturing. Based on this fact the survey result figure 4. 3 shows that the majority
of the enterprises accounted 43.44% in Adigrat city has been parti cipating in metalwork, 25.41%
in woodwork 17.21% in textile and 10.66% in Handcraft. It is relatively better performance,
whereas the figure indicated that leather production and agro -processing industries has taken
small number of contribution accounting 2.46% and 0.82% respectively in manufacturing
participation in the city. This clearly implies that there is a problem on surplus product in rural
area and gap between urban and rural area linkage. Consequently, the government must focus in
these aspects t o overwhelm the serious challenges.

Figure 4. 4 Growth pattern of Enterprises (Sources: Own Survey data, 2017)

12
Figure 4. 4 and table 4. 2 shows the level of micro and small enterprise growth in Adigrat city. As
it is depicted in the pie chart the majority around half of the micro and small enterprise in
Adigrat city shows slow rate of growth level or stagnant which, accounted 50.8%, which is also
followed by decline level accounted for 14.7 % of the total. The remaining 34.4 % of the
respondents replied tha t their business level show fast growth. Thus, the result of survey study in
table 4. 2 indicates that, 102(83.6%) of the operators were micro level and 20(16.4%) of the
operators were small level at the time of establishment. However, according to summary result
in table 4. 2, the current status shows that among 20 small enterprises 5 of them turn down to
micro level and some of the operators were grown up from micro to small enterprises yet the
number is not significant change as more than half 61.5% of th e enterprise are still in micro
level.
This shows that most of the micro and small enterprise in Adigrat city are not growing at an
expected rate. The respondents are also asked the main reason for show decline and stagnant
growth level of their business a nd they mentioned major problem like lack of access to credit,
less amount of startup capital, lack of working place fit to the growth level of their business and
lack of technical skill as the major causes.
Table 4. 2 of enterprises at the time of establishment and current status
Current Status
Micro Small Total
Status at Time of
Establishment Small Count 5 15 20
Table N % 4.1% 12.3% 16.4%
Micro Count 70 32 102
Table N % 57.4% 26.2% 83.6%
Total Count 75 47 122
Table N % 61.5% 38.5% 100.0%
Sources: Own Survey data (2017)
With the introduction of the new economic policy in Ethiopia, which is designed along the line
of a market economy and the implementation of the reform programs, a more conducive business
environment is being created . Even though, there is a general positive environment for the

13
development of MSE, since the strategies and support programs are not fully or adequately
implemented, there are still problems that constrain the growth of the MSE sector.
The major problems encountered by the Adigrat MSE operators in the process of running and
expanding their activities is presented in figure 4. 5.

Figure 4. 5 the major problems of micro & small enterprises (Sources: Own Survey data, 2017)
The major Source for any business is finance. The availability of financial resources thus highly
determines the productivity and development of any economic activity. According to the survey
findings illustrated in figure 4. 5, lack of working capital was among the most pressing problems
that small enterprises identified as limiting expansion of their business.
The figure shows 35.25% of the interviewed small business operators in Adigrat city responded
that lack of capital has been the first problem to their operation. Likewise, 18.03 % of MSEs
have been suffering from high interest rate.
Beside to the fact that, the operators’ opinion and actual observation, the cluster shade provided
by the government have been stayed very limited for manufacturing small and micro enterprises.
The chart also shows that about 17.21 % of the entrepreneurs reported that lack of working
premises/ places was the third major problem.

14
The other problem, which is identified by the operator, was the poor supply of infrastructure
facilities. According to the figur e 4.5, the infrastructure problem was list among the stated
problems contributing about 11.48 % out of the total challenges according to survey result
reported by the respondents referring that the infrastructure facility problem was one of serious
problem s of the business enterprises. In most cases, the small and micro enterprises
establishments are not well furnishing with the basic infrastructure facilities. In line to this, the
accessibility of latest technology also was mentioned as the one of the prob lem faced by the
operators indicating 11.48% of the MSES in Adigrat city have been using poor technology, this
implies that the profit of the operators is influenced to grow up. Indeed, without technology there
is no development so it is impossible to grow n up easily.
Furthermore, the limited market for any sort of goods and services has an automatic effect on the
profitability of these business activities. The figure shows that 6.56% of the business operators
responded that lack of market is the problem of their business activities,
MODEL RESULTS
4.1.2.1. Multiple Linear Regressions
Table 4. 3 Multiple Linear Regression model for Revenue of Enterprises versus Independent
Factors
Predictors Coefficient
(𝛃𝐢) SE (𝛃𝐢) T P-value 95% CI
Lower Upper
Constant – 𝛃 𝟎 7660.810 2139.411 3.581 0.001 3421.001 11900.620
Exp. of MSE -𝛃 𝟏 291.850 91.236 3.199 0.002 111.041 472.658
Tec. Latest -𝛃 𝟐 4115.336 1760.305 2.338 0.021 626.824 7603.848
Outdated -𝛃 𝟑 990.263 956.146 1.036 0.303 -904.594 2885.120
POI Frequent -𝛃 𝟒 -6880.063 1774.629 -3.877 0.000 -10396.96 -3363.165
Rarely -𝛃 𝟓 -6557.674 1624.894 -4.036 0.000 -9777.834 -3337.515
RAM Yes-𝛃 𝟔 928.722 918.358 1.011 0.314 -891.248 2748.693
INC >50,000 -𝛃 𝟕 8664.013 1659.435 5.221 0.000 5375.401 11952.624
20,001 -50-𝛃 𝟖 908.473 911.813 0.996 0.321 -898.527 2715.474

15
Note: – the t-test of significant relationship between response variable monthly revenue of MSE
and individual parameter/coefficient of the predictors or challenging factors.
Null hypothesis: Monthly Revenue of MSE has no significant relationship with each predictor.
𝐇𝟎: 𝛃𝒊=𝟎,𝐟𝐨𝐫 𝐢=𝟎,𝟏,…,𝐤,𝐤=𝟏𝟏
Alternative hypothesis : Monthly Revenue of MSE has significant relationship with individual
predictors 𝐇𝐀: 𝛃𝒊=≠𝟎,𝒇𝒐𝒓 𝒊=𝟏,…,𝒌,𝒌=𝟏𝟏
Yi =7660 .810 +291 .85X1+4115 .336 X2+990 .263 X3−6880 .063 X4−6557 .674 X5
+928 .722 X6+8664 .013 X7+908 .473 X8−872 .157 X9+2910 .833 X10
+123 .772 X11
The result for multiple linear regressions to monthly revenue of MSE versus some predictors
(challenging factors) is displayed in table 4.6. One continues and six categorical predictors with
respective dummy variables were included in the model. Consequ ently, out of total 11
predictors, five of them, such as outdated technology users (P -value=0.303), access of raw
materials (P -value=0.314), INC (20,001 -50,000) holders (P -value=0.321), lack of demand (P –
value=0.295) and fair market linkage owners (P -value =0.874) were found that, statistically
insignificant. This sense indicates lack of evidence to reject the null hypothesis that says there is Demand Yes-𝛃 𝟗 -872.157 828.027 -1.053 0.295 -2513.112 768.798
Market
Linkage High -𝛃 𝟏𝟎 2910.833 1270.485 2.291 0.024 393.030 5428.637
Fair-𝛃 𝟏𝟏 123.772 777.241 0.159 0.874 -1416.537 1664.080
Model Summary
Correlation Coefficient(r) Coefficient of
Determination ( 𝑅2) Adjuste
d (𝑅2) Standard Error of the
Estimates
0.851 0.723 0.696 3759.50534
ANOVA (Analysis of Variance) For overall Test
Model Sum Squares
(SS) DF Mean Square(MS) F P-Value(Sig.)
Regression 4066582092.1 11 369689281.102 26.156 0.000
Residual 1554726842.3 110 14133880.385
Total 5621308934.4 121

16
no relationship between Monthly Revenue of MSE and the mentioned factors at 5% level of
significance. Whereas, the rest six predictors, such as experience (lifetime) of enterprise or firm
(P-value = 0.002), latest technology user (P -value=0.021), both frequently and rarely power
interruption problem owners (P -value <0.001), above 50,000 initial capital holders (P -value
<0.001) and high market linkage owners (P -value = 0.024) have statistically significant
relationship with monthly revenue of MSE. Therefore, there is enough evidence to reject the
null hypothesis stated above at 5% level of significance.
Moreover, except the predictor variables power interruption ( β4=−6880 .063 S.E.=
1774 .629 ,β5=−6557 .674 S.E.=1624 .894 indicated statistically significant negative
relationship with monthly revenue of MSE, and lack of demand (β6=-872.157 (S.E. = 828.027 )
indicated statistical ly insignificant negative relationship with monthly revenue of MSE, all the
rest variables are positively related with response variable monthly revenue of MSE. Therefore,
the negative coefficient or parameter for power interruption indicated that, the mon thly revenue
of MSE is 6880 .063 times lower in the firms that face frequent power interruption than that of
firms without any power interruption when all the rest -challenging factors kept constant.
Likewise, the firms with rare power interruption has 6557 .674 times low amount of monthly
revenue of MSE relative to the firms without any power interruption holding all the rest
challenges unchanged. Consequently, the result has shown that power interruption has
statistically significant influence on the monthly revenue of MSE (p -value<0.001) for both levels
relative to the reference level no power interruption.
Moreover, the coefficient for experience of MSE ( β1=291 .850 (S.E. = 91.236 )) indicated that,
the amount of monthly revenue of MSE increases by rate of 291.85 per a unit increase in
experience of MSE if all the other variables kept constant. Similarly, the parameter for latest
technology users ( β2=4115 .336 (S.E. = 1760.305 )) refers, the amount of monthly revenue of
MSE is 4115.336 times higher in technology users than that of comparison groups manually
involved assuming all the rest variables hold constant. Likewise, the monthly revenue of above
50,000 initial capital holders ( β4=8664.013 (S.E. = 1659.435)) is 8664.013 times higher than
that of the firms below 20,000 holders if all the rest variables hold constant. Finally, the
parameter for high market linkage owners ( β10=2910 .833 (S.E.=1270 .485 )) indicated that,

17
the amount of monthly revenue of MSE for high market linkage owners is 2910.833 times higher
than that of low market linkage owners if all the rest kept constant.
Generally, the variable with the large test statistic value, which is t -values accompanied with
small p -values, indicates the statistically significant relationship between resp onse variable
monthly revenue and respective predictors. Likewise, 95% confidence intervals displayed at the
right -side column of table 4.6, could provide similar conclusions.
For instance, the 95% confidence intervals for statistically significant parame ters of predictors
such as experience of MSE, latest technology users, firms faced frequent power interruption,
firms faced rare power interruption, above 50,000 initial capital holder and high market linkage
owners are (3421.001, 11900.2620), (111.041, 47 2.658), ( -10396.96, -3363.165), ( -9777.834, –
3337.515), (5375.401,11952.624) and (393.030, 5428.637) respectively. Hence, all the
mentioned intervals excluded zero inside the intervals indicating the rejection of null hypothesis
stated in the above in favo r of the alternative one. However, the 95% confidence intervals of
parameters for the rest statistically insignificant predictors have included zero in the intervals
indicating that there is no evidence to reject the null hypothesis. For further details, look at table
4.6.
Note: The Overall F -test and Good fit test of the multiple linear regression models
Null hypothesis : Monthly Revenue of MSE has no significant relationship with all predictors.
H0: β0=β1=β2=⋯=βk=0, for i=0,1,…,k, k=11
Alternative hypot hesis: Monthly Revenue of MSE has significant relationship at least with one
predictor. HA: βi≠0,at least for one i, i=0,1,…,k,k=11
Moreover, the result of ANOVA for multiple regression displayed in table 4.6, indicated that the
overall multiple regression model is statistical significant with (F =26.156, p -value<0.001).
Furthermore, correlation coefficient, which is the measure of strength of relationship and
determination of coefficient, which is measure of variability, explained by the model or by the
predictors/independent variables are displayed in table 4.6. Accordingly, correlation coefficient

18
(r=0.851) indicated that there is strong positive relationship between response variable
monthly revenue and predictors (challenging factors) included in the model.
Similarly, the determination of coefficient ( R2=0.723) shows, the 72.3% of the variability is
explained or caused by the data in model. In short, 72.3% of the response/dependent variable
monthly revenue of MSE is explained by the given predi ctors/independent variables included in
the model. This also measures the goodness of fit of the model. Therefore, it indicated that the
regression model is moderately good fit.
Gradually, the values appeared in the bracket with coefficients everywhere in this paper are
standard errors and are measures of precision, accuracy and consistency of the estimated results
for prediction. The small values of standard errors tell us the estimations are good for prediction
and large standard errors tell us the revers e. Furthermore, the standard errors are very important
to calculate confidence intervals and estimate coefficients in the model.
Finally, multiple linear regression models for response variables which is the monthly revenue of
MSE has shown that there is statistically significant association among the dependent variables
monthly revenue of MSE and some challenging factors.
DISCUSSIONS
Based on different well -organized literatures and analysis that were included in this thesis, some
discussions and review of works are organized as following.
The aim of this study was to model the performance of micro and small manufacturing
enterprises which was mainly represented by monthly revenue conditional to predetermined
constraints so -called predictors in North par t of Ethiopia Adigrat city. To do so , questionnaires
were d rafted and distributed to respondents to find out the main constraints associated to the
performance of MSEs then in turn to suggest the solutions for thoughtful problems . The study
was organized b y descriptive survey study and 122 micro and small enterprise individual holders
were included in the study. The data obtained were analyzed by using descriptive method and
relevant models to measure the performance of micro and small enterprises. According to several
reviews of previous works and realities in the ground, the obvious and well-known measuring
indicators of growth are, profit or revenue, employment size, supply, quality of product and total

19
assets. Howe ver, in our case the monthly revenue was taken as the indicator response of
performance of MSEs.
In contrast to this, Abay et al. (2014) in their recent work have taken employment size as the
main measure of MSEs growth. However, it is obviously true to sa y that, measuring the MSE
growth with monthly revenue may have reasonable advantage. Therefore, after deciding the main
indicators of the MSEs’ growth as one continuous variable, which is monthly revenue the
relevant statistical modelmultiple linear regres sion has been applied to measure the significant
relationship between responseand predetermined constraints . The prevalence of the performance
in MSEs was set in terms of Monthly revenue , whether it has been performing or not based on
the respective experi ences or life time of the firm.
Generally, the overall models for both models such as ANOVA for multiple regression ( p-
value<0.001 ) is found to be statistically significant and the results of latter model in lines to the
results of binary logistic regressi on in Abay et al. (2014).
Finally, this paper was concerned only withone indicatorof MSEs growth out of several
indicators. Therefore, this work can open the direction for the future work that can measure the
MSEs growth by extending to several alternative growth indicators.
CONCLUSION AND RECOM MENDATIONS
CONCLUSION
The main aim of the study is to model the performance of MSEs conditional to thepredetermined
constraints or predictors. Hence, after passing throughrelevant analyzing techniques to show the
relationship between MSEs performance and associated constraints or predictors, we came up
with summarized results and discussions.
The result has shown that, the majority (65.6%) of the respondents have been working witho ut
any NGOs support. and (63.11%) of the respondents started their business with low capital
which is below 20,000 ETB. in Addition to this when we see the source of initial capital of the
enterprises covering 92.62% of the total enterprises were from pers onal saving , relative support
and loan from the relatives; Likewise, even though there is somehow a support especially in

20
training from government, yet the result magnifies the gap of the government and other
concerned sectors in supporting the motivated new business starters in capital, in job creativity,
in technical skill improvement and motivations or incentives. Hence, this fact is leads to
conclude that there has been very serious problem to setup a new manufacturing business in
Adigrat City.
Based on the fact the result shows that the majority of the enterprises accounted 68.85% in
Adigrat city has been participating in metalwork and woodwork. Whereas the survey Indicated
that leather production and agro -processing industries has taken mingles numbe r, this clearly
implies that there is a problem on surplus product in rural area and gap between urban and rural
area market linkage. From this fact, it can be possible to conclude that the government bodies are
not concerned to motivate micro and small en terprises to join the sectors. Consequently,
according to the national strategy, the government must focus in these aspects to overwhelm the
serious challenges.
Beside the mentioned conclusions above, the results of the models elaborated similar
conclusion s. According to the multiple linear regression model, out of six categorical and one
continues predictors, except two such as lack of raw materials and lack of demand, all the rest
predictors have significant influence on monthly revenue which in turn affe ct the growth
performance of MSEs. Furthermore, the model shows that power interruption has negatively
significant influence on growth of MSEs and demand has negatively insignificant association
with monthly revenue whereas all the rest predictors have po sitive association with monthly
revenue. The overall regression model is statistical significant ( 𝑃−𝑣𝑎𝑙𝑢𝑒≤0.001). Similarly,
the determination of coefficient ( R2=0.723) shows, the 72.3% of the variability is explained by
model. In short, the given pr edictors included in the model explain 72.3% of the variability in
monthly revenue of MSE. This also measures the goodness of fit of the model. Therefore, it
indicated that the regression model is moderately good fit.
Generally, the findings have shown tha t, although few numbers of enterprises have shown
improvement in growth pattern, majority of the enterprises are still stable as stagnant and some
of them have shown the decline pattern. The main reason mentioned by the respondents are lack

21
of amount of st artup capital, lack latest technology, power interruption and lack of technical skill
as the major causes.
RECOMMENDATION
According to the major findings of this study the MSE sector faces a lot of challenges. In order
to overcome the challenges, it needs effort from government, NGOs and all other stakeholders,
therefore, the concerned bodies should make attention on the development of MSEs and avoid
constraints that hinder the growth performance of MSEs. Therefore, the following
recommendations are drawn b ased on the findings and conclusion.
With regard to infrastructure, the study resultsindicated that power interruption is one of the most
challenging factor to manufacturing enterprise growth in Adigrat city. Government and stake
holders could promote and facilitate the growth of MSEs by accessing necessary infrastructure
facilities like electric c ity, road, water etc
Market linkage is another constraint of MSEs operators in the study area. Hence, government
should protect micro enterprises from large enterprises in order to get market for their
commodities, by providing subsidies like tax exemption until they stand by their feet in addition
to this local government, NGOs and other stakeholders are expected to search market alternatives
to link MSEs with organizations who need MSEs products. Moreover, the local government is
recommended to create mar ket linkage for micro and small manufacturing enterprises with large
industries, with farmers and so on.
Based on the findings, the supports given by the governments and NGOs have been yet
unsatisfactory; therefore, it is recommended that the government an d NGOs should take careful
look and attention to solve these serious challenges of several motive individuals at starting
enterprises.

22

REFERENCE
Abay, H.H, Tessema, F.G. and Gebreegziabher, A.H. ( 2014 ). External Factors Affecting the
Growth of Micro and Small Enterprises (MSEs) in Ethiopia:A Case Study in Shire
Indasselassie Town, Tigray. European Journal of Business and Management , 6(34):
www.iiste.org
Agresti, A. (1990). Categorical Data Analysis . New York: Wiley.
Asgedom, D. B. (2013). Challenges and Opportunities of Micro and Small Manufacturing
Enterprises the Case of Easter Zone of Tigray. International Journal of Science and
Research (IJSR) ISSN (Online): 2319 -7064, Pat Richardson.
Biretawi, T. (2013). The Role of Micro and Small Enterprise in Poverty Alleviation: The Case of
Hawzen Town in Eastern Zone of Tigray. Ethiopian civil service University.
Cooper, D. R. & Schindler, P. S. (2008). Business research methods. (10th Ed) McGraw -Hill
Dillman, D.A. (2007). Mail and interne t surveys: The tailored design method (2nd ed.). Hoboken,
NJ: John Wiley & Sons.
Fissuh, G.H., Fissuh, Y.H., Negash, Z. and Girmay, H. (2016). Assessment of the Effective total
quality management onperformance at selective colleges in Tigray regionaladmini strative
State, Ethiopia. IOSR Journal of Business and Management (IOSR -JBM), 18(8) .Ver. IV,
56-84, e -ISSN: 2278 -487X, p -ISSN: 2319 -7668; www.iosrjournals.org
Gebretensea, H. (2003). Entrepreneurship. Mekelle; central Press
Kothari. C.R. (2004). Research Methodology, methods and techniques (2nd Ed). New Delhi: New
Age International.
Walmobo, W.E. (1996). small business development in Ethiopia concept practice and prospect, 4th
symposium on HRD for business, A.A, 2-18.