Technical efficiency of Macedonian pig farms

8/13/2019 Technical efficiency of Macedonian pig farms

1/8

ORIGINAL SCIENTIFIC PAPER

Agroeconomia Croatica 3:2013 (1) 18-25

Technical eciency of Macedonian

pig farmsMarina Petrovska1, Aleksandra Martinovska-Stojceska1, Bo hlmer2, Ana Kotevska1

1Institute of Agricultural Economics, Faculty of Agricultural Sciences and Food, University of Ss. Cyril and Methodius

in Skopje, bul.Aleksandar Makedonski bb, Skopje, Macedonia ([email protected])2Department of Economics, Swedish University of Agricultural Sciences in Uppsala, Johan Brauners vg 3, Uppsala,

Sweden

ABSTRACTTe economic transition and market globalization processes have triggered structural changes in the Mac-

edonian agriculture and influence on the efficiency and competitiveness in the pig production sub-sector. Tispaper aims to identiy the level o technical efficiency on pig arms in the Republic o Macedonia. Te DataEnvelopment Analysis approach is used to measure the efficiency level, taking into consideration the exactquantity o inputs used in the production in relation to a given quantity o output. Te data is analysed by mak-ing comparative analyses o the managerial behaviour and other non-measurable variables that influence theefficiency. Te results determine what managerial activities influence the efficiency. Tey indicate the type andlevel o inputs that need to be changed, so that arms could reach the same technical efficiency achieved by thebest armers.

Key words: Data Envelopment Analysis, technical efficiency, Macedonian pig arms

ISSN 1333-2422UDK = 631.11 : 636.4 (049.7)

INTRODUCTION

Livestock production is very important or thedomestic consumption in the Republic o Macedo-nia. Te economic transition and market globali-zation processes have triggered structural changesin Macedonian agriculture. Tey had a significantimpact on the whole agricultural sector, includingpig production. Te pork processing industry playsa significant role in the domestic economy, but pigproduction is considered as inefficient and less com-petitive compared to oreign markets (Dimitrievskiet al., 2010). During the transition period many othe previously existing pig arms were shut downand those who continue to operate have changedtheir ownership into private (MAFWE, 2007).

oday, there are only 7 big pig arms lef romthe transition period. Established during 1970s, theperiod when the country was a part o Yugosla-

via, they managed to overcome the transition andown almost 40% o the total number o pigs in thecountry. Te remaining 60% o pigs are owned byindividual producers, mostly small amily holdings,and due to the governmental and IPARD supportto agriculture the number o commercial amily pigarms is constantly growing (MAFWE, 2007).

In recent years, the world aced a number oeconomic, climate and ood crises. As a response,macroeconomic policies and other political deci-sions implicate a concept o the green economy orsustainable development and the need or ast ad-aptation on it (Ocampo, 2012). On the other side,pig producers are acing challenges to meet the newmarket requirements and regulations, and ofenound difficulties to adjust quickly. Moreover, arm-ers lack inormation and knowledge about pro-ducing on the competitive markets and they needormal and inormal education to increase arm


2/8


efficiency (Manevska-asevska, 2012). Accordingto MAFWE (2007) production efficiency is a chal-lenge based on the current inefficient arm manage-

ment practises, ollowed by inadequate technologyand high production costs which additionally in-crease product prices on the domestic market. Inthis sense, it is necessary to pay more attention onmanagerial capacity building and explore activitiesthat influence the increase o the efficiency level.

Te aim o this paper is to determine the levelo technical efficiency on pig arms in Macedonia.Furthermore, the paper analyses the managerial ac-tivities that can be changed with a ocus on the typeand quantity o inputs used in the production in re-spect to the output quantities produced at the end

o the production process. aking into account thatthe managerial activities are key contributors or e-ficient production, changes in their behaviour leadsthe arms to reach the same technical efficiency asthe best armers.

MATERIAL AND METHODS

Many researchers conclude that armers canproduce efficiently by rational use o inputs (De-breu, 1951; Koopmans, 1951; Farrell, 1957). Farm-ers influence on arm technical efficiency by choos-

ing certain amount o inputs to produce the mosteconomically beneficial quantities o outputs(Petrovska, 2011). Measuring technical efficiency isan approach based on solving input and output op-timisation problems (Farrell, 1957). However, thereare other variables that have significant impact onthe efficiency score. Te influencing variables rep-resent sources o inefficiency.

Technical efficiency variables

echnical efficiency scores are estimated by using

non-parametric Data Envelopment Analysis (DEA)approach. Te efficiency indicators are calculatedby using the computer programme DEAP version2.1 (Coelli, 1996). DEAP uses linear programmingto estimate a production rontier unction or a seto decision making units (DMUs), which is used orevaluation o relative efficiency o each unit. Temodel allows a large number o inputs and outputswith different measurement units and gives indi-vidual and multiple efficiency scores or more thanone decision making unit. Efficiency scores com-puted by DEAP lie between 0 and 1. Tose units

that operate on the rontier line ace ull technicalefficiency and have an efficiency score equal to 1.All other units are less efficient and to increase their

technical efficiency they need to make changes inthe production process.

Moreover, production characteristics affect thechoice o the efficiency scale (Manevska-asevskaet al., 2011). Te efficiency scale on which armsoperate is a ratio between a constant and variablereturn to scale. Because pig production as a primaryagricultural production is very sensitive to externalactors such as climate, diseases, market situation,managerial abilities to finalise all activities on time,Variable Return to Scale (VRS) DEA model is moreappropriate or measuring technical efficiency. Due

to the nature o production process, armers can-not influence on the produced quantities and there-ore input oriented DEA is used to measure or howmuch each DMU can reduce the utilised inputs,with output levels held constant. Input orientedtechnical efficiency o each arm that operates un-der VRS DEA is measured as a linear programmerepresented in the ollowing equation:

min,

,

subject to -qi+ Q 0,

xi- X 0,

I1

1 0,

According to Coelli et al. (2005) the equation as-sumes that there are data on N inputs and M out-puts or each o I firms. For the ith firm they arerepresented by the column vectors xi and qi, respec-tively. X is (NxI) input matrix and Q is (MxI) out-put matrix. is a scalar and represents an efficiencyscore that should be estimated or each ith firm. isa (Ix1) vector o constants. A convexity constraint(I11) is used to account VRS which indicated

whether the arm is operating in an area o increas-ing or decreasing return to scale.

For the purpose o this study, the surveyed armsare categorised into three groups: big, medium andsmall. Te analysis is based on direct data collectedconcerning the production activities o pig arms in2010. Due to the small number o pig arms in theRepublic o Macedonia, the data collection was con-ducted with ace to ace interviews with 21 armersrom the whole country.

For the first stage analysis, financial variables

19echnical efficiency o Macedonian pig arms


3/8


were collected in Macedonian currency (MKD) andthen converted into Euro. o receive quantitativeresults or each arm, all variables are normalised

per Livestock Unit (LU). o simpliy the analysis asingle output o total LU produced is considered oreach arm. Also, our inputs are considered as: eed,energy, labour and other inputs. Te other inputsare sum o inputs that have a significant impact ontechnical efficiency: veterinary costs, vaccinationand insemination doses, hygiene and disinectioncosts, ecology cost, cost or transport and insurance.

Sources of inefficiency

Influencing variables are not measurable unitsand thereore cannot be analysed with DEA model.Teir significance is analysed by using obit regres-sion model and a comparative analysis. In general,the emphasis is put on armers behaviour and thewillingness to apply new technologies in order toachieve more technically efficient production.Farmers behaviour is important or an efficientproduction planning process. Tis process includesdecision-making through planning, choosingamong alternatives, implementation o decisionsand their control. Indeed, knowledge influencesthe output increase while using the same amounto inputs (Rivera & Alex, 2008) and choosing the

right production alternatives. Not only ormal, butalso inormal knowledge, such as participation atconerences and workshops and consultations withother armers or experts can contribute in shar-ing the experience and more flexible acceptanceo new technologies and modern market require-ments (Fulton, 1995; Miller, 1994; Millar & Curtis,1997). In that way, good decision-making processescan contribute to increased arm efficiency, whileproduction inefficiency appears because o a lowerlevel o armers education, experience in arming,interpersonal relationship and acceptance o inno-

vations (Kilpatrick et al., 1999; Coelli et al., 2005).

RESULTS AND DISCUSSION

Descriptive statistics is used to evaluate the re-lationship between the use o different technologytypes and the other second stage variables. Tearm accounting variable shows how many armersare providing bookkeeping o the production activ-ities in quantities and prices spent, and marketingincludes the activities provided by armers to repre-

sent their arms, by using internet technology andother marketing sources.

Te level o education variable is the ormal edu-

cation that armers have (no ormal education is es-timated with 1, primary education with 2, second-ary education is equal to 3 and higher education to4). Te inormal education o armers is divided intothree variables: participation in agricultural asso-ciations and the years o armers experience, whileseminars, conerences and workshops are estimatedwith 3, 2 and 1 depending on armers participationofen, rarely or never, respectively. Te descriptivestatistics or the variables used in both first and sec-ond stage analysis is presented below in able 1.

20 Marina Petrovska, Aleksandra Martinovska-Stojceska, Bo hlmer, Ana Kotevska


4/8


Technical efficiency results

Te results show that 14% o the analysed armshave ull scale efficiency, with scale efficiency equalto 1 (SE=1) and operate on the production rontierline. Te remaining arms show variability in tech-nical inefficiency which does not depend on armsize. Most o them, 85% operate on increasing re-turn to scale (IRS) and thus the response should betowards reducing the utilisation o inputs per unito output in order to optimise the arm technicalefficiency. Te overall technical efficiency scores aregiven below in able 2.

Table 2. Summary statistics for technicalefficiency scores (n=21)

TECRS

TEVRS

SE

Mean 0.417 0.895 0.434

SD 0.321 0.196 0.310

Min 0.014 0.500 0.027

Max 1.000 1.000 1.000

Also, the efficiency score gave different valuesin constant and variable return to scale. Te vari-able return to scale approach proves to be more ap-propriate or application in agricultural production

due to the large number o actors influencing armefficiency. According to the results, arms ace veryhigh technical efficiency under VRS representedwith an 89% average. On the other hand, there is abig difference between the two scales, which is dueto the low mean efficiency o an average 43%. ech-nical efficiency under CRS is lower than EVRS orabout 41%, which is confirmed by the theory or es-timating DEA efficiency (Coelli et al., 2005; Coelli,1996).

Despite the high level o average technical effi-ciency under VRS, and considering that there arearms that are only 10% efficient, armers could still

make improvements. Tat way, the most impor-tant thing or the armer to decide is how much heshould reduce the amount o utilised inputs with-out reducing the amount and quality o final prod-ucts. Overall, the results show a 22% inefficiency ineed input and, to increase the efficiency, armersshould reduce eed quantities used in the produc-tion. However, only minimum quantities o theeed input should be reduced considering that it isthe most valuable input or quality pig production.In DEAP labour input is analysed in quantity (thenumber o workers including amily members) andin price unit (total cost spent or labour). Tese two

Table 1. Variables used in the first and second stage analysis (n=21)

Variable Unit Mean SD Min Max

Firststage

variables

Farm revenue EUR/LU 468.20 160.27 206.00 873.99

Total output LU 2,137.41 2,978.89 46.19 10,276.50

Feed quantities kg/LU 1,350.17 574.77 726.35 2,755.54

Price of feed EUR/LU 291.01 125.26 134.64 675.69

Labour No. 16.86 17.28 2.00 43.00

Price of labour EUR/LU 36.54 25.56 2.35 107.63

Price of energy EUR/LU 19.55 21.27 3.12 102.62

Price of materials EUR/LU 32.88 34.05 1.46 104.48

Price of services EUR/LU 38.60 39.97 1.71 122.66

Second

stage

variab

les

Distance to the closest market km 1.70 0.50 6.50

Mortality % 5.55 1 15

Level of education 3.69 3 4

Participation in associations % 42.86 0 1

Seminars 2.28 1 3

Farmers experience % 18.48 2 37

Farm accounting % 52.38 0 1

Marketing % 28.57 0 1

Investment % 57.14 0 1



5/8


approaches do not give any differences in the resultswhich show that there should be no changes in theutilisation o labour. According to the results, the

cost o energy increases arm relative inefficiencyor about 23% and thereore this input should bereduced, as well. Inefficient arms can increase theiroverall efficiency score by spending less or otherinputs including costs o other materials and ser-

vices or a given level o output. Te other inputsincrease the inefficiency or about 42%. Te highestlevel o inputs that has been estimated on some othe arms is 91% or energy and services costs.

Significance of the influencing variables

Tere are many variables that have a significantimpact on technical efficiency o production, such

as: governmental regulations regarding pig breed-ing and animal welare, environmental laws and thetype o production. Governmental regulations sup-

port pig production by laws and all armers are obli-gated to implement. Te influence o environmentalregulations is closely related to the location o arms.Galev & Lazarov (1968) confirmed that the best lo-cation o the arm is to be at least 1 km ar rom themarket, but closer to the main road and slaughter-houses. Smallar distance may cause environmentalproblems in regard to the disposal o manure, butalso in regard to water and air pollution, while larg-er distance leads to increased transportation costs.According to the results, the average distance to theclosest market or big city is 1.7 km. Tere are alsosome small arms that are located 0.5 km and 6.5km away.

Table 3. Relationship between the second stage variables (n=21)

Variable Technology Mean SD Min Max

Feed consumption

(kg/LU)

New 4.66 2.21 2.52 7.89

Combination 3.87 0.45 3.33 4.67

Old 5.14 1.79 3.64 9.39

Mortality (%)

New 3.40 2.67 1.20 8.00

Combination 5.70 5.28 1.20 15.00

Old 6.89 4.51 1.00 13.00

Piglets/sow

(No. of piglets)

New 12.50 3.73 8.00 18.00

Combination 13.50 3.73 9.00 15.00

Old 10.67 1.12 9.00 12.00

Te descriptive statistic analysis show a relationbetween the type o production technology usedand arms sustainability. Te impact o changingthe old technology o production is analysed in re-spect to the mortality, eed consumption and thenumber o piglets per sow in one arrowing. Teaverage mortality is estimated in percentage, whilethe investment variable explains how many arm-ers have a new technology o production includingthose who have changed all production and thosewho have changed only a part o the productionsystem. Te statistical analysis shows that arm-ers, who had decided to change the old technologyo production and to use new production systems,managed to increase the number o piglets per onearrowing. Also, this activity resulted in decreasingthe mortality rate by more than 1.1% and the con-

sumption o eed per live weight by around 1%. Tedescriptive statistic is given in able 3.

Farmers behaviour is very important or mak-ing proper decisions regarding arm operational

activities. Many studies confirmed that by increas-ing knowledge, armers can increase the overall e-ficiency o production (Kilpatrick et al., 1999; Koo-pmans, 1951). Using a descriptive statistic showshow ormal and inormal education o armers in-fluences the decisions or implementing new tech-nology o production (see able 4).

Te results show that armers use all types otechnology, despite their different preerences.However, well educated armers and those whoinvest in increasing their inormal knowledge are



6/8


Table 4. Relationship between the type of technology and knowledge (n=21)

Variable Technology Mean SD Min Max

Formal knowledge

New 3.83 0.41 3 4

Combination 3.83 0.41 3 4

Old 3.33 0.71 2 4

Seminars, conferenc-

es and workshops

New 2.33 0.52 2 3


Old 2.11 0.60 1 3

Associations and co-

operatives

New 1.33 0.52 1 2


Old 1.44 0.53 1 2

Sources of informa-tion

New 1.83 0.41 1 2


Old 1.67 0.52 1 2

flexible to innovations. Tey use new or makechanges to the existing type o production tech-nology. Overall, armers who have a higher edu-cational level are open to innovations and easilyaccept new regulations and technologies. Te in-ormal knowledge is represented by participationon seminars, conerences and workshops, partici-

pation in agricultural associations and coopera-

tives and the source o inormation received byarmers. Here, only participation in associationsand cooperatives does not give a relation to thetype o technology used.

Te environmental variables and armers per-ormances have significant influence in increasingarms efficiency. Te obit regression results are

shown in able 5. Given the results, only one vari-

Table 5. Tobit regression results of the influencing variables (n=21)

Variable Coef. Std. Err. t P>|t| [95% Conf. Interval]

Farm location -0.157 0.047 -3.330 0.008a -0.262 -0.052

Piglets per sow -0.012 0.003 -3.620 0.005a -0.019 -0.005

Mortality -0.234 0.049 -4.740 0.001a -0.344 -0.124

Technology 0.189 0.051 3.670 0.004a 0.074 0.303

Formal education 0.003 0.018 0.190 0.856 -0.036 0.042

Informal education 0.156 0.040 3.940 0.003a 0.068 0.245

Farmers experience -0.012 0.003 -3.390 0.007a -0.019 -0.004

Constant 0.026 0.269 0.100 0.924 -0.573 0.626

astatistically significant at 1%

able ormal education is not significant or the e-ficient production. All other variables that affecttechnical efficiency are statistically significant at 1%.

Especially, arm location, number o piglets persow in one arrowing, percent o mortality o pig-lets and armers experience are significant or low-

ering production inefficiency.

CONCLUSIONS

Relative technical efficiency in input orientationdepends on many variables. In this paper, inputs



7/8


and outputs are analysed by using a rontier pro-duction unction. Te results show a relative tech-nical efficiency score or each arm rom an input

perspective. Only 5 arms operate on a ull scale e-ficiency, while the other pig arms operate on an in-creasing return to scale and they can improve arminefficiency by providing proper production struc-ture. Since it is very difficult to manage the outputamounts because o the nature o production, thebest way to increase the efficiency is or armers todecide on the amount o inputs. Input surplusesshould be reduced or estimated quantities, differ-ent or each arm, without changing quantity andquality o the final products.

Te influencing variables that concern the envi-

ronmental actors and managerial behaviour have asignificant influence in technical inefficiency. Tatis how new policies are emphasising the environ-mental actors as a key issue or sustainable agri-cultural production. Tis modern policy does notallow old production types, since pig production isknown as one o the biggest pollutants o the en-

vironment. However, this trend is still a big chal-lenge in Macedonia and results in additional costsor armers, even i the analysis shows that chang-ing the old production technology is significant orincreasing overall efficiency.

Farmers who have a higher level o educationand invest to improve their knowledge are moreflexible in applying new production technologies.Managerial actors that contribute to more efficientproduction are inormal knowledge and armersexperience.

Tis analysis leads to the conclusion that rationaluse o inputs or a given level o output can improvethe overall technical efficiency by 20%. Proper pro-duction structure and managerial behaviour aresignificant or increasing the technical efficiency oproduction.

REFERENCES

Coelli, J. ., Rao, D. S., ODonnell, J. C., Battese, E.G. (2005). An introduction to efficiency and pro-ductivity analysis. New York: Springer Science.

Coelli, . (1996). A guide to DEAP version 2.1: AData Envelopment Analysis (computer) pro-gram. Center or efficiency and productivityanalysis (CEPA) Working paper 96/8. Australia:University o New England.

Debreu, G. (1951). Te coefficient o resource utili-sation. Econometrica, 19: 273-292.

Dimitrievski, D., Georgiev, N., Simonovska, A.,

Martinovska-Stojceska, A., & Kotevska, A.(2010). Review o the agriculture and agricul-tural policy in FYR Macedonia. In . Volk (Ed.),Agriculture in the Western Balkan Countries:Studies on the Agricultural and Food Sector incentral and Eastern Europe, 57, 145-164. IAMO.

Farrell, M. J. (1957). Te measurement o produc-tive efficiency. Journal o the Royal Statistical So-ciety, Series A, 120: 253-290.

Fulton, A. (1995). Te implications o armers reli-ance on private consultants. Extension net, 2 (5).

Galev, ., Lazarov, S. (1968). Organisation o the

livestock production. Skopje: Faculty o Agricul-ture.

Kilpatrick, S., Johns, S., Murray-Prior, R., Hart, D.(1999). Managing arming: How armers learn.Barton: Rural industries research and develop-ment cooperation.

Koopmans, . C. (1951). An analysis o productionas an efficient combination o activities. In . C.Koopmans (Hrsg.), Activity analysis o produc-tion and allocation (Bd. Monograph 13). NewYork: Wiley.

MAFWE. (2007). National agricultural and rural

development strategy or the period 2007-2013.Skopje: Ministry o Agriculture, Forestry andWater Economy.

Manevska-asevska, G. (2012). Efficiency Analysiso Commercial Grape-Producing Family Farmsin the Republic o Macedonia. PhD thesis. Uni-

versity o Agricultural Sciences. Uppsala: SLU.Manevska-asevska, G., Hansson, H., Latruffe, L.

(2011). Evaluating the potential effectivenesso rural development programme targets onarms in FYR Macedonia - An efficiency studyo grape-growing amily arms. Food Econom-

ics: 1-12.Millar, J., Curtis, A. (1997). Moving armer knowl-

edge beyond the arm gate: An Australian oarmer knowledge in group learning. EuropeanJournal o Agricultural education and Extension,4 (2): 133-142.

Miller, C. L. (1994). Contract arming, agribusinessand global relations in North West asmania.PhD thesis. Brisbane: Griffith University.

Ocampo, A. (2012). Te transition to a greeneconomy: Benefits, Challenges and Risks rom



8/8


a Sustainable Development Perspective: Sum-mary o Background Papers. Report by a panelo Experts. United Nations: Second Preparatory

Committee Meeting or United Nations Coner-ence on Sustainable Development.Petrovska, M. (2011). Efficiency o pig arm produc-

tion in the Republic o Macedonia: Data Envel-

opment Analysis approach. Uppsala: Universityo Agricultural Sciences.

Rivera, W. M., Alex, G. E. (2008). Human resource

development or modernising the agriculturalworkorce. Human Resource Development Re-view, 7 (4): 374-386.

Tehnika uinkovitost makedonskihsvinjogojskih farmi

SAETAKProcesi ekonomske tranzicije i trine globalizacije potaknuli su promjene u makedonskoj poljo-

privredi i izvrili utjecaj na uinkovitost i konkurentnost u svinjogojskom podsektoru. Cilj ovoga rada jeodreivanje razine tehnike uinkovitosti na svinjogojskim armama u Republici Makedoniji. Koritena jemetoda analize omeivanja podataka kako bi se izmjerila razina uinkovitosti, pri emu se uzela u obzirtona koliina inputa u proizvodnju u odnosu na dobivenu koliinu outputa. Podaci su analizirani pomoukomparativne analize menaderskog ponaanja i ostalih nemjerljivih varijabli koje utjeu na uinkovitost.Oni ukazuju na to da se vrsta i razina inputa moraju promijeniti na armama, kako bi se dosegla razinauinkovitosti koji postiu najbolji poljoprivrednici.

Kljune rijei: metoda analize omeivanja podataka, tehnika uinkovitost, makedonske svinjogojske

arme


Documents

Technical efficiency of Macedonian pig farms