Simulating farm structural change dynamics in Thessaly (Greece) using a recursive programming model

Man zia is, S ama is; Rozakis, S elios; Ka anikolas, Paulos; Pe sakos, A hanasios;
Tsiboukas, Kons an inos
A icle
Simula ing a m s uc u al change dynamics in Thessaly
(G eece) using a ecu si e p og amming model
Bio-based and Applied Economics (BAE)
P o ided in Coope a ion wi h:
Fi enze Uni e si y P ess
Sugges ed Ci a ion: Man zia is, S ama is; Rozakis, S elios; Ka anikolas, Paulos; Pe sakos, A hanasios;
Tsiboukas, Kons an inos (2024) : Simula ing a m s uc u al change dynamics in Thessaly (G eece)
using a ecu si e p og amming model, Bio-based and Applied Economics (BAE), ISSN 2280-6172,
Fi enze Uni e si y P ess, Flo ence, Vol. 13, Iss. 4, pp. 353-386,
h ps://doi.o g/10.36253/bae-14790
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/321804
S anda d-Nu zungsbedingungen:
Die Dokumen e au EconS o dü en zu eigenen wissenscha lichen
Zwecken und zum P i a geb auch gespeiche und kopie we den.
Sie dü en die Dokumen e nich ü ö en liche ode komme zielle
Zwecke e iel äl igen, ö en lich auss ellen, ö en lich zugänglich
machen, e eiben ode ande wei ig nu zen.
So e n die Ve asse die Dokumen e un e Open-Con en -Lizenzen
(insbesonde e CC-Lizenzen) zu Ve ügung ges ell haben soll en,
gel en abweichend on diesen Nu zungsbedingungen die in de do
genann en Lizenz gewäh en Nu zungs ech e.
Te ms o use:
Documen s in EconS o may be sa ed and copied o you pe sonal
and schola ly pu poses.
You a e no o copy documen s o public o comme cial pu poses, o
exhibi he documen s publicly, o make hem publicly a ailable on he
in e ne , o o dis ibu e o o he wise use he documen s in public.
I he documen s ha e been made a ailable unde an Open Con en
Licence (especially C ea i e Commons Licences), you may exe cise
u he usage igh s as speci ied in he indica ed licence.
h ps://c ea i ecommons.o g/licenses/by/4.0/
Bio-based and Applied Economics
BAE
© 2024 Au ho (s). Open access a icle published, excep whe e o he wise no ed, by Fi enze Uni e si y P ess unde CC-BY-4.0 License
o con en and CC0 1.0 Uni e sal o me ada a.
Fi enze Uni e si y P ess | www. up ess.com/bae
Bio-based and Applied Economics 13(4): 353-386, 2024 | e-ISSN 2280-6172 | DOI: 10.36253/bae-14790
Ci a ion: Man zia is, S., Rozakis, S.,
Ka anikolas, P., Pe sakos, A., & Tsi-
boukas, K. (2024). Simula ing a m s uc-
u al change dynamics in Thessaly
(G eece) using a ecu si e p og am-
ming model. Bio-based and Applied
Economics 13(4): 353-386. doi: 10.36253/
bae-14790
Recei ed: June 3, 2023
Accep ed: Sep embe 2, 2024
Published: Decembe 31, 2024
Da a A ailabili y S a emen : All el-
e an da a a e wi hin he pape and i s
Suppo ing In o ma ion iles.
Compe ing In e es s: The Au ho (s)
decla e(s) no con lic o in e es .
Edi o : Ma eo Za alloni
Simula ing a m s uc u al change dynamics
in Thessaly (G eece) using a ecu si e
p og amming model
S ama is Man zia is1,*, S elios Rozakis2, Pa los Ka anikolas1,
A hanasios Pe sakos3, Kons an inos Tsiboukas1
1 Depa men o Ag icul u al Economics and Ru al De elopmen , Ag icul u al Uni e si y
o A hens, G eece
2 School o Chemical and En i onmen al Enginee ing, Technical Uni e si y o C e e, Cha-
nia C e e, G eece
3 Depa men o Pe o mance, Inno a ion and S a egic Analysis o Impac , Bio e si y
In e na ional, Rome, I aly
*Co esponding au ho . E-mail: s a.a hens@ho mail.com)
Abs ac . Al hough he policy impac s on a ms accumula e yea by yea , mos a m
decision models ocus on sho - e m decisions, e alua ing policies based on snap-
sho s. S uc u al changes a e g adually buil ; he e o e, a m decision models should
conside he sequences wi hin he pe iod unde s udy. Mul iyea da a om he a able
sec o in Thessaly, G eece, ha e ed a newly de eloped a m-le el ecu si e linea p o-
g amming model mainly o simula e a m s uc u al change dynamics. The p oposed
model inco po a es new e idence on he s a egic decision o a able c op a ms ega d-
ing hei emaining in he p oduc ion sys em and a m expansion. Resul s e eal an
e iden g adual a mland concen a ion in ela i ely la ge a ms, accompanied by a
g adual expansion o he mos p o i able c opping ac i i ies, e i ying he eal-wo ld
su i al s a egy o a ms.
Keywo ds: a m s uc u al change, land use change, ecu si e linea p og amming
model, a able p oduc ion sys em, G eece.
JEL Codes: C61, Q12, Q18.
1. INTRODUCTION
The declining numbe o su i ing a ms o e ime and he inc ease
in a e age a m size gene ally signal he e olu iona y p ocess o s uc u al
change in he ag icul u al sec o o de eloped economies (Plogmann e al.,
2022), implying changes in he a m size dis ibu ions (Zimme mann and
Heckelei, 2012; Sain -Cy e al., 2019).
Ag icul u al economis s ha e shown g ea in e es in desc ibing s uc-
u al change dynamics and unde s anding i s d i e s (Plogmann e al., 2022).
S uc u al change is d i en by a ious economic ac o s (Neuen eld e al.,
2019), en i onmen al ac o s and social d i e s (RIRDC, 2007). Ne e he-
354
Bio-based and Applied Economics 13(4): 353-386, 2024 | e-ISSN 2280-6172 | DOI: 10.36253/bae-14790
S ama is Man zia is e al.
less, some au ho s (Wibo g, 1998; Plogmann e al., 2022)
conside a m economic pe o mance he p ima y d i e
o s uc u al change since i somehow encloses all he
abo e ac o s.
S uc u al change is a no mal e olu iona y p ocess
in an economy (Godda d e al., 1993). O e ime, ising
ag icul u al p oduc i i y enabled he ans e o p oduc-
i e ac o s equi ed o he de elopmen o o he sec o s
o he economy (Balmann and Valen ino , 2016). How-
e e , s uc u al change in he ag icul u al sec o is usu-
ally co ela ed wi h public conce ns, which a e mainly
exp essed h ough public deba es in wo e ms, i s ly
as “dying peasan s” and secondly as “ ac o y a ming”
(Balmann and Valen ino , 2016).
Highligh ing he i s public conce n, his may be
because, gene ally, s uc u al change ha dly leads o
Pa e o Supe io s a es (Balmann and Valen ino , 2016).
F om his pe spec i e, Coch ane (1958) concludes ha
inc eased ag icul u e p oduc i i y posi i ely a ec s only
a limi ed numbe o inno a i e a ms, while mos a m-
e s a e a ec ed nega i ely due o he ollowing d op in
ag icul u al commodi y p ices. Suppose we analyze his
easoning om he poin o iew o public policy; in ha
case, s uc u al change may educe he p oblem con-
ce ning he p o i abili y o emaining a ms bu , on he
o he side, educe he numbe o small a ms and hus
coun e s he equi y goals o public socie y (Finge and
Benni, 2021). Wi hin his con ex , some au ho s con-
side he signi ican ole o public policy in mi iga ing
he consequences o s uc u al change by poin ing ou
ha “much o he public policy agenda has clea ly been
es ablished on a p emise o op imali y o a amily a m
s uc u e” (Godda d e al., 1993: 486). Howe e , imple-
men ing app op ia e policy in e en ions p esupposes
p o iding de ailed in o ma ion (by policy analys s) on
s uc u al change in ag icul u e h ough e idence-based
policy- ele an esea ch o suppo e idence-based ag i-
cul u al policy decision-making.
The Eu opean Common Ag icul u al Policy (CAP)
ma ks essen ial shi s in he con ex whe e a ms ope -
a e, wi h signi ican e o ms a emp ed e e y decade.
Policy impac s on a ms accumula e yea a e yea ,
a ec ing he a m s uc u es and, by ex ension, he
well-being o u al communi ies, c ea ing a ipple e ec
on he local economy. In his amewo k, modeling
he dynamics o s uc u al change adjus men (i.e., he
change o e ime o a m numbe s and a m size dis i-
bu ion) is highly desi able because i can p o ide policy-
make s and s akeholde s wi h possible al e na i e sce-
na ios o s uc u al change adjus men s, bu i is s ill no
widely used in policy analysis (Ciaian e al., 2013; Espi-
nosa e al., 2016). Modeling exe cises such as dynamic
app aisals can suppo policy analys s in o mula ing
public policies o ob ain he “desi ed a m s uc u e”
conside ing he socie al demands o equi y (Finge and
Benni, 2021).
Two main me hodological app oaches inco po a e
s uc u al change in ag icul u e: econome ics and sim-
ula ion models (which aim o analyze a m s uc u al
change endogenously) (Espinosa e al., 2016; Zimme -
mann e al., 2009). Econome ic models include Ma ko
chains (Zimme mann and Heckelei, 2012) and a ious
o he eg ession app oaches (Zimme mann e al., 2009).
Simula ion models include ecu si e p og amming mod-
els (e.g., Wibo g, 1998; Guinde e al., 2005; Henningsen
e al., 2005; O e mann and Ma ga ian, 2014; Djanibe-
ko and Finge , 2018; Mi enzwei and B i z, 2018) and
agen -based models (e.g., Balmann, 1997; Be ge , 2001;
Happe e al., 2008; F eeman e al., 2009; Be e al.,
2011; T oos and Be ge , 2016; Becke s e al., 2018; Sun
e al., 2022; Dona i e al., 2024). As simula ion mod-
els can endogenously cap u e a m s uc u al change,
hey a e conside ed sui ed o analyzing policy changes’
alloca i e and dis ibu i e e ec s on an ag icul u al p o-
duc ion sys em (Guinde e al., 2005; Happe e al., 2008;
Espinosa e al., 2016). Al hough agen -based models such
as Ag iPoliS (Balmann, 1997) a e conside ed by a i-
ous modele s he mos comp ehensi e a emp a ana-
lyzing he impac o policies on s uc u al change (e.g.,
Zimme mann e al., 2009), a e cha ac e ized by g ea e
complexi y (e.g., Zimme mann e al., 2009), and hey a e
e y demanding in e ms o pa ame e isa ion (e.g., Zim-
me mann e al., 2009; Rowan e al., 2011; K emmydas
e al., 2023) and calib a ion (e.g., Zimme mann e al.,
2009). In addi ion, he p e e ence o simple p ocess-
based models1 should no be igno ed (T oos and Be g-
e , 2020). The e o e, while cap u ing s uc u al change
endogenously and p o iding meaning ul insigh s in o
he alloca i e and dis ibu ional e ec s o a ious exog-
enous ac o s, he a m-le el ecu si e p og amming
models can also be manageable ega ding he deg ee o
complexi y and da a equi emen s compa ed o o he
simula ion models such as agen -based models.
Based on he abo e discussion, he main objec i e
o his esea ch is o in es iga e he impac s o policy
expe imen s on a m s uc u al change dynamics in
G eece h ough an endogenous modeling app oach
based on a newly de eloped a m-le el ecu si e linea
p og amming model. While p ima ily aimed a simula -
ing he impac o policy expe imen s on he e olu iona y
p ocess o a m s uc u al change, he p oposed simula-
ion model is also seconda ily used o simula e he e ec
1 P ocess-based models include models such as simula ion models and
sys ems dynamics models.
355
Simula ing a m s uc u al change dynamics in Thessaly (G eece) using a ecu si e p og amming model
Bio-based and Applied Economics 13(4): 353-386, 2024 | e-ISSN 2280-6172 | DOI: 10.36253/bae-14790
on land use change while analyzing i s ela ionship wi h
s uc u al change adjus men .
In he con ex o s uc u al changes, he s a egic
decision o a ms is summa ized h ough he ph ase
“g ow o go” (Plogmann e al., 2022), implying he
aspec s o (i) a m iabili y and (ii) a m g ow h/expan-
sion. Th ough he p oposed modeling app oach, we
in eg a e he a m’s economic pe o mance as he main
d i e o his decision (e.g., Wibo g, 1998; Pa oissien e
al., 2021; Plogmann e al., 2022). In mo e de ail, in addi-
ion o adi ional mone a y alue c i e ia o de e mine
a su i ing/ iable a m, we in oduce a no el iabil-
i y c i e ion, assuming ha a me s may compa e hei
economic pe o mance o socie al consump ion bench-
ma k, in he sense ha he agen (in ou case, eal-
wo ld indi idual a m) mus achie e a minimum le el
o p o i abili y, allowing en y in o he “ a ace” acco d-
ing o “Keeping up wi h he Joneses” (KUJ) p e e ences
(e.g., Ba ne e al., 2010; Lomba do, 2021; Pa oissien
e al., 2021). Rega ding a m expansion, he p oposed
modeling app oach in oduces a u he no el elemen
h ough he concep o ela i e op imal a m g ow h in
equi y o ealloca e/alloca e esou ces be ween neighbo -
ing su i ing a ms.
The p oposed model can also be cha ac e ized as a
One-Way Communica ion Model whe e he in o ma ion
lows om he econome ic model o he ecu si e p o-
g amming a m model (Huang e al., 1980). In pa icula ,
he Au o eg essi e In eg a ed Mo ing A e age (ARIMA)
models a e used o o ecas he alues o he exogenously
de e mined pa ame e s o in e es o conduc ou -o -sam-
ple simula ions. Addi ionally, ARIMA s ochas ic p ocess
es ima es exp ess he agen s’ quasi- a ional expec a ions
ega ding ag icul u al commodi y p ices and c op yields
(Ne lo e and Bessle , 2001; Siegle e al., 2024).
Fo he empi ical applica ion o he p oposed sim-
ula ion model, a ep esen a i e sample o a able c op
a ms (in e ms o a m s uc u e) o he egion o Ka -
di sa (NUTS-3 le el), Thessaly, is chosen. The p io i y o
empi ical applica ion gi en o he a able p oduc ion sys-
em is jus i ied by he ac ha G eek a able a ming is
cha ac e ized by a compa a i ely highe a e o s uc u -
al change conce ning he o he main ypes o a ming
(o he pe manen c ops, o he g azing li es ock) (FADN
Public Da abase).
F om a gene al pe spec i e, wi h his analysis, we
a emp o con ibu e o he deba e on dynamic assess-
men s o he mul idimensional e ec s in he con ex o
policy e o ms. Addi ionally, mo e speci ic con ibu ions
o li e a u e a e exp essed h ough a leas ou ways:
Fi s , we add knowledge by in eg a ing e olu ion-
a y and social psychology elemen s o de ine a a m as
iable based on KUJ p e e ences. Second, we simula e
esou ce ealloca ion based on he c i e ion o ela i e
op imal a m g ow h in equi y as an al e na i e a m
expansion/g ow h c i e ion o adi ional c i e ia such as
he shadow alues o esou ces (e.g., Guinde e al., 2005;
Hennessy, 2007; Espinosa e al., 2016). Thi d, he u ili-
za ion o he ARIMA s ochas ic p ocess o ime se ies
o ecas ing o he alues o he exogenously de e mined
pa ame e s (such as ag icul u al commodi ies p ices,
inpu p ices, and c op yields) is an addi ion o he exis -
ing li e a u e since in simila simula ion models; hese
alues a e mainly de e mined ei he om seconda y
da a sou ces (e.g., Wibo g, 1998; Hennessy, 2007; O e -
mann and Ma ga ian, 2014) o h ough assump ions/
scena ios (e.g., Guinde e al., 2005; Henningsen e al.,
2005; T oos and Be ge , 2016; Mi enzwei and B i z,
2018) o simpli ied end models (e.g., Happe e al., 2008;
Be e al., 2011; Becke s e al., 2018). Fou h, despi e he
g ea impo ance o he a able p oduc ion sys em o he
G eek ag icul u al sec o and he compa a i ely highe
a e o s uc u al change han he o he main p oduc-
ion sys ems, o ou knowledge, a m-le el ecu si e p o-
g amming models ha e no been used o p o ide a “bo -
om-up” simula ion o s uc u al change o G eek a able
p oduc ion sys em.
The es o he pape is o ganized as ollows. Sec-
ion 2 desc ibes he applied me hodology, he da a used
o apply he me hodology, and he policy expe imen s.
The empi ical esul s a e p esen ed in Sec ion 3, Sec ion
4 discusses hem, and concludes.
2. METHODOLOGY AND DATA
2.1. Recu si e p og amming models o impac assessmen
in ag icul u e
Recu si e p og amming models ha e al eady been
in oduced in he 1960s o ep esen dynamic adjus -
men s o p oduc ion capabili ies a he a m le el, and
hen wi h he s udy o Day and Cingo (1978) egional
in e dependence and s uc u al elemen s we e inco -
po a ed (Espinosa e al., 2016). Indica i ely, ecu si e
p og amming a m models ha e been u ilized o he
de elopmen o a m i m g ow h models (e.g., Chien
and B ad o d, 1976; Ci adini e al., 2008; Dowson e
al., 2019) o in es iga e he economic consequences due
o a me s’ adap abili y o di e en wa e a ailabili y
scena ios (e.g., Iglesias e al., 2003; Rowan e al., 2011;
Robe e al., 2018; Dowson e al., 2019), o assess he
impac s o a ious policy e o m and p ice scena ios on
a m income and in es men beha io (e.g., Viaggi e
al., 2010; Viaggi e al., 2011; Da is e al., 2013; B i z e
356
Bio-based and Applied Economics 13(4): 353-386, 2024 | e-ISSN 2280-6172 | DOI: 10.36253/bae-14790
S ama is Man zia is e al.
al., 2016) and o analyze he impac o policies on a m
s uc u al change (e.g., Wibo g, 1998; Guinde e al.,
2005; Henningsen e al., 2005; O e mann and Ma ga-
ian, 2014; Djanibeko and Finge , 2018; Mi enzwei and
B i z, 2018).
The main s uc u al elemen s o a ecu si e p o-
g amming model co espond o a cons ained op imi-
za ion model and a da a gene a o , whe e he da a gen-
e a o , gi en he op imal alue o solu ion in pe iod ,
eini ializes he pa ame e s o pe iod +1, including a
se o cons ain s ha ela es he easible alues o cu -
en a iables o pas alues o a iables and exogenous
e en s (McCa l and Sp een, 1997). Following Chien and
B ad o d (1976) and McCa l and Sp een (1997), he gen-
e al o mula ion o he ecu si e p og amming a m
model is as ollows:
Max E{Π } = ∑
j E{Cj, }T Xj, (1)
Subjec o:
∑
j Ai,j, Xj, ≤ bi, ∀i (2)
Xj, ≥ 0 ∀j (3)
whe e E{ } deno es he expec a ion ope a o ; E{Π } is
a m’s expec ed g oss p o i in EUR which is maxi-
mized in yea ; E{Cj, } is he ec o o expec ed g oss
p o i in EUR/hec a e (ha) o he j c opping ac i i y in
pe iod ; Xj, is he ec o o he decisions a iables ha
deno es he le el o he j c opping ac i i y (hec a es o
c ops) in pe iod ; Ai,j, a e he esou ce I usages by he
j c opping ac i i y pe ha in pe iod ; bi, is he ec o o
a ailable esou ces i in pe iod , unc ionally dependen
upon lagged phenomena (Kay, 1971; McCa l and Sp een,
1997).
The eini ializa ion o he ec o o a ailable
esou ces (bi, ) is conduc ed h ough a m i m g ow h
ules such as he Endogenous Feedback Mechanism
(EFM) (e.g., Kay, 1971; Chien and B ad o d, 1976; McCa-
l and Sp een, 1997; Ci adini e al., 2008; Da is e al.,
2013; Robe e al., 2016). Al hough EFM has been
applied wi h some a ia ions, he gene al ma hema ical
o mula ion is as ollows:
bi, = (bi, -1, Xi, -1*, Vi, )) (4)
whe e he ec o o a ailable esou ces (bi, ) in pe iod
is de e mined by he ec o o a ailable esou ces in
he p e ious pe iod (bi, -1), he op imal decisions in he
p e ious pe iod (Xi, -1) and by he ec o Vi, ha allows
o ex e nal changes in he esou ce es ic ions due o
exogenous e en s ha will occu in he pe iod which
a e a he de e mined by ex e nal economic and en i-
onmen al ac o s (Kay, 1971; McCa l and Sp een, 1997;
Da is e al., 2013; Robe e al., 2016).
Since he p oposed model is used o s uc u al
change analysis, h ee mo e basic s uc u al elemen s
a e included o de e mine (i) a m iabili y, (ii) a m
g ow h/expansion, and (iii) capi al s ock e olu ion a he
a m le el. A de ailed desc ip ion o hese s uc u al ele-
men s o he model is ca ied ou in subsequen sec ions.
2.2. ARIMA modeling o economic o ecas ing in ag icul-
u e
The use ulness o such a simula ion model, which
is op imized sequen ially wi hin a dynamic amewo k,
lies in he abili y o p o ide esul s ou side he e e -
ence pe iod (ou -o -sample o ecas s). The e o e, o con-
duc ou -o -sample simula ions, he o ecas ed alues o
he exogenously de e mined pa ame e s o he a m a e
equi ed.
Va ious modele s ha e used ARIMA models o o e-
cas exogenously de e mined pa ame e s such as ag i-
cul u al commodi y p ices (e.g., Mao e al., 2022), c op
yields (e.g., Pe sakos e al., 2016), cos o p oduc ion ac-
o s (e.g., Hlouško á e al., 2018) and supply o a ious
esou ces (e.g., he o al amoun o ag icul u al land,
o al amoun o pes icides) (Cos ache e al., 2021).
ARIMA models a e i ed u ilizing he in o ma ion
in he se ies i sel o p edic u u e poin s in he se ies
(Ch is odoulos e al., 2010; Ga nie , n.d.), and he e-
o e he independen a iables a e lagged alues o he
se ies. Mo e speci ically, he u u e alues o he depend-
en a iable can only be desc ibed h ough hei p ob-
abili y dis ibu ion ende ing he se ies a s ochas ic p o-
cess2 (Pa doe, n.d.). In his ein, se e al modele s con-
side ha he use o ARIMA models is app op ia e o
economic o ecas ing in ag icul u e, especially in cases
o lack o well-de eloped heo y o limi ed in o ma-
ion (Pe sakos e al., 2016); as a esul , he o ecas ing o
exogenous a iables o en p esen p oblems o econo-
me ic model use s (Oli ei a e al., 1979).
Wi hin his con ex , he ARIMA s ochas ic p o-
cess is u ilized o es ima ing he alues o exogenously
de e mined pa ame e s o in e es (in ou case, ag icul-
u al commodi y p ices, c op yields, cos s, in e es a e,
o al a able land, and o al ci cula ing capi al) o pe -
o m ou -o -sample o ecas s in he medium e m. In
2 De ails on ARIMA modeling amewo k a e p o ided in Pa A:
Concep ual amewo k o ARIMA modeling in he supplemen a y
ma e ial.

357
Simula ing a m s uc u al change dynamics in Thessaly (G eece) using a ecu si e p og amming model
Bio-based and Applied Economics 13(4): 353-386, 2024 | e-ISSN 2280-6172 | DOI: 10.36253/bae-14790
addi ion, ARIMA models a e u ilized o es ima e he
alues o andom/s ochas ic pa ame e s, such as ag i-
cul u al commodi y p ices and c op yields, o exp ess
agen s’ quasi- a ional expec a ions mechanism (Ne lo e
and Bessle , 2001; Siegle e al., 2024).
2.3. Simula ion model speci ica ion and assump ions
2.3.1. Model’s basic s uc u e
The ini ial endowmen s wi h p oduc ion ac o s
a e speci ied be o e he sequen ial simula ion s a s (in
ou case, a able land, i iga ed land, ci cula ing capi al,
capi al s ock, and bo owed capi al) (Happe e al., 2008)
(see Figu e 1). To simula e a ms’ p oduc i e decisions
h ough he p oposed a m-le el ecu si e linea p o-
g amming model, we assume ha a ms op imize he
expec ed g oss p o i (e.g., Rowan e al., 2011) o each
yea gi en he a m’s esou ce, policy, and lexibili y
cons ain s. To elabo a e mo e, esou ce cons ain s con-
ain: (i) A able land cons ain ; (ii) I iga ed land con-
s ain ; and (iii) Ci cula ing capi al cons ain .
Policy cons ain s con ain: (i) 2013 CAP e o m
cons ain s (g eening obliga ions); (ii) CAP Pos -2020
e o m scena io cons ain s; (iii) Ni a e pollu ion educ-
ion p og am cons ain s; and (i ) O ganic a ming p o-
g am cons ain . Flexibili y cons ain co esponds o
he cons ain o mul iannual con ac a ming3.
Each sub-model (based on ep esen a i e indi idual
eal-wo ld a m) op imized ecu si ely4 o a sequence o
15 yea s ( om 2012 o 2026). Time p og esses in disc e e
ime in e als, symbolizing he commencemen o a
g owing season a ime (see Figu e 1). To pe o m ou -
o -sample simula ions (i.e., ou side he e e ence pe iod,
speci ically a e 2019), mainly ARIMA models a e used
o o ecas he alues o he exogenously de e mined
pa ame e s o in e es (see Figu e 1).
2.3.2. Fa m agen s’ expec a ions speci ica ion and model
alida ion
Va ious au ho s (e.g., Femenia e al., 2017) conside
naï e and quasi- a ional expec a ions (ARIMA mod-
eling), bo h based on pas obse a ions, o be he mos
equen expec a ion mechanisms5 in some ypes o
3 A de ailed desc ip ion o he objec i e unc ion and cons ain s is
p o ided in Pa B: S uc u e o he model’s objec i e unc ion and
cons ain s in he supplemen a y ma e ial.
4 The model is w i en in GAMS language.
5 A de ailed desc ip ion o a m agen s’ expec a ions mechanisms is
p o ided in Ne lo e and Bessle (2001), Haile e al. (2016), Femenia e
al. (2017), and Siegle e al. (2024).
a ming. In luenced by his inding, we emphasize hese
wo mechanisms o expec a ions ega ding ag icul u al
commodi y p ices and c op yields in he p esen s udy,
conside ing ha hey will be ep esen a i e o sample
a ms and he in o ma ion a ailable o hem (mainly
based on pas obse a ions).
Mo e speci ically, we ha e o mula ed wo al e -
na i e models; one e e ed o as he Quasi-Ra ional
expec a ions (QR) model and he o he as he Naï e and
Quasi-Ra ional expec a ions (NV&QR) model. In mo e
de ail, in he QR model case, he agen ’’ expec a ions a e
exp essed h ough quasi- a ional expec a ions (ARIMA
modeling) o ag icul u al commodi y p ices and c op
yields (e.g., Na ayana and Pa ikh, 1981; Ne lo e and
Bessle , 2001; Siegle e al., 2024). In he NV&QR mod-
el case, he agen ’’ expec a ions a e exp essed h ough
naï e p ice expec a ions o ag icul u al commodi y
p ices (e.g., Ne lo e and Bessle , 2001; Robe e al., 2018;
Siegle e al., 2024) and h ough quasi- a ional expec a-
ions o c op yields.
Then he wo p oposed models a e alida ed o
hei capabili y o ep oduce ac i i ies alloca ion
(Gómez-Limón e al., 2016), he numbe o su i ing
a ms (Becke s e al., 2018), and he a m size dis ibu-
ion (F eeman e al., 2009; Becke s e al., 2018).
2.3.3. De e mining a m iabili y
Usual app oaches o de ining a m iabili y a e
based on he oppo uni y cos o a ming (e.g., Lough ey
e al., 2022) and he po e y line (e.g., Mille e al.,
1981; Lough ey e al., 2022). O he app oaches o de in-
ing a m iabili y ocus on mone a y e u ns, whe e he
a m income should ensu e long- e m a m g ow h in
equi y, o a leas he equi y should emain s able in o
he u u e (e.g., B igh e al., 2007; Ba nes e al., 2015).
Ano he in e es ing app oach o de ining a m ia-
bili y om a socio-economic pe spec i e is based on
he “Keeping up wi h he Joneses” (KUJ) p e e ences
(Mille e al., 1981; Pa oissien e al., 2021). Fa me s may
compa e hei p o i s o he o e all s anda d o li ing
(a e age li ing expendi u es/a e age consump ion le el)
o socially close e e ence g oup (neighbo ing a ms),
which is conside ed he socie al consump ion bench-
ma k o social e e ence poin o consump ion le el
(Pa oissien e al., 2021).
F om his pe spec i e, agen s ha s and below
hei socie al e e ence poin (in he sense o no being
able o inance his le el o consump ion) a e o ced o
s ay ou o he “ a ace o keeping up wi h he Joneses”
(Ba ne e al., 2010), may expe ience lowe li e sa is ac-
ion and p o essional well-being, a si ua ion which may
358
Bio-based and Applied Economics 13(4): 353-386, 2024 | e-ISSN 2280-6172 | DOI: 10.36253/bae-14790
S ama is Man zia is e al.
c ea e incen i es o exi he sys em (Pa oissien e al.,
2021; Nguyen and He on, 2021). The e o e, a a m mus
achie e a minimum le el o p o i abili y, allowing en y
in o he “ a ace” (Lomba do, 2021) acco ding o KUJ
p e e ences (i.e., keeping up wi h a benchma k p opo -
ional o he a e age le el o consump ion o he socially
close e e ence g oup (Ba ne e al., 2010) such as neigh-
bo ing a ms).
The in luences o his hypo hesis come om e olu-
iona y and social psychology, whe e a ious esea ch-
e s assume ha he ques o s a us – equen ly e e ed
o in his con ex as “Keeping-up-wi h- he Joneses”–
depends on he social no ms ela ed o a benchma k
consump ion le el such as he a e age consump ion le el
o he socially close e e ence g oup (Fishe and Hei-
jd a, 2009; Lomba do, 2021; Mageli e al., 2022). Based
on he abo e easoning, a ious esea che s assume ha
he ques o social s a us can be linked o he s i ing
o su i e (Mageli e al., 2022). No ably, since social
g oups can dis ibu e esou ces among hei membe s,
an agen ’s chances o su i e and ep oduce a e g ea -
ly enhanced i she/he belongs o a g oup and i she/he
holds a ela i ely high social ank wi hin he g oup, in
he sense ha an agen ’s ela i e posi ion may gi e he /
him a su i al ad an age h ough access o ma e ial and
ep oduc i e esou ces (Mageli e al., 2022).
Al e na i ely, a m iabili y can be de ined acco d-
ing o a combina ion o mone a y alue and socio-eco-
Ini ial condi ions : Numbe o neighbo ing su i ing
a ms, a m size dis ibu ion,a ailable a able land a
a m le el, a ailable ci cula ing capi al a a m le el,
capi al s ock a a m le el, bo owed capi al a a m
le el
Fa m-le el da a
(Field su ey)
Fa m-le el op imiza ion model :
Expec ed Fa m G oss P o i maximiza ion
unde esou ce, policy and lexibili y
cons ain s
Fa m agen s' expeca ions
Quasi- a ional agen s’ expec a ions
o p ices and c op yields
OR
Nai e agen s’expeca ions o p ices
and quasi- a ional agen s’
expec a ions o c op yields
Ou -o -sample simula ions
ARIMA models:
Fo ecas ing alues o
cos s, p ices, c op yields,in e es
a e, o al a able land, and o al
ci cula ing capi al
Linea end model:Fo ecas ing
alue o li ing expendi u es index
Fa m iabili y algo i hm:
i) Op imal Fa m Ne P o i A e Tax ≥
Simula ed Socie al Consump ion
Benchma k o neighbo ing a ms AND
ii) Op imal Fa m G ow h in Equi y ≥ 0
Numbe o
neighbo ing
su i ing a ms
Fa m size
dis ibu ion
Sha e o
a mland by a m
size classes
Regional land use
change dynamics
Economic
pe o mance by
a m size classes
En i onmen al
impac
assessemen
Re-ini ializa ion o esou ces & a m
i m g ow h ules o su i ing a ms:
-Endogenous Feedback Mechanism
(EFM) o esou ces (land, ci cula ing
capi al)
Coupled wi h: Rela i e op imal Fa m
G ow h in Equi y o ealloca e/alloca e
esou ces (abandoned/ elesased land
a ailable o en , egional a ailabili y
o ci cula ingcapi al)be ween
neighbo ing su i ing a ms
Nex g owing
season ( = +1)
Pos -solu ion module o Economic
indica o s :
i) Op imal Fa m Ne P o i A e Tax ii)
Op imal Fa m G ow h in Equi y
Pos -solu ion module o Means-based
en i onmen al indica o s: i) Fe ilize
use ii) Pes icide use iii) Wa e use
Capi al s ock e olu ion (In es emen module):
-Pe pe ual In en o y Me hod (PIM) coupled wi h:
-Leon ie p oduc ion ela ionship be ween capi al s ock and land
Ac ual p ices; Ac ual c op yields
De e mining equi ed bo owing
capi al:
i) Requi ed bo owing ci cula ing
capi al
ii) Requi ed bo owing in es men
capi al
Upda ed
in o ma ion on
ini ial condi ions
Op imal c opping ac i i ies alloca ion
Regional c op supply
Numbe o
neighbo ing
su i ing
a ms
Fa m size
dis ibu ion
Times se ies da a
(S a is ical
au ho i y, Ru al
ins i u ions, FADN)
Policy
scena ios
Figu e 1. Concep ual diag am o he p oposed modeling amewo k. No es: A pos -solu ion module o means-based en i onmen al indica-
o s enables he model o es ima e he en i onmen al pe o mance o a ms. Howe e , o limi he size o his pape , he en i onmen al
impac assessmen will no be p esen ed he e. Sou ce: Au ho s
359
Simula ing a m s uc u al change dynamics in Thessaly (G eece) using a ecu si e p og amming model
Bio-based and Applied Economics 13(4): 353-386, 2024 | e-ISSN 2280-6172 | DOI: 10.36253/bae-14790
nomic c i e ia (Be e al., 2011; Mi enzwei and B i z,
2018; Seidel and B i z, 2019).
In he p esen modeling app oach, a sample a m is
conside ed iable/su i ing by sa is ying wo iabili y
c i e ia: (i) he c i e ion o socie al consump ion bench-
ma k o neighbo ing a ms (NBF)6 acco ding o he KUJ
p e e ences and, (ii) he c i e ion o non-nega i e op i-
mal a m g ow h in equi y. A his poin , we would like
o men ion ha , ollowing simila simula ion models
(Be e al., 2011; O e mann and Ma ga ian, 2014; Mi -
enzwei and B i z, 2018; Seidel and B i z, 2019) we sim-
ula e only a m exi acco ding o he a m exi module
conside ing economic and socioeconomic c i e ia. Con-
sequen ly, we do no model he li e cycle o agen s who
en e a ming, ge old, and e i e (Be e al., 2011).
The e o e, ollowing each disc e e op imiza ion
ime-s ep (annual), e e y neighbo ing a m nb decides
whe he o emain in he sys em o exi (see also Figu e
1). Speci ically, a neighbo ing a m is conside ed iable
and emains in he p oduc ion sys em when a he end
o he yea mee s bo h iabili y c i e ia, i.e., (i) he
op imal Fa m Ne P o i a e Tax (FNPAT*nb , ) should
be a leas equal o he simula ed a e age li ing expendi-
u es o neighbo ing a ms in yea (LENBF, sim), and (ii)
op imal a m g ow h in equi y (FGE*nb , ) should be a
leas equal o ze o.
6 The li e a u e on whom agen s compe e wi h o social s a us,
i.e., who he Joneses a e, is ela i ely limi ed (Mageli e al., 2022).
Ne e heless, i is concei able ha agen s compa e mo e in ensely
wi h agen s who a e socially p oxima e o hem (Mageli e al., 2022).
Fo example, socie y se es as a socially dis an e e ence g oup,
whe eas colleagues a e socially close e e ence g oups (Mageli e al.,
2022). In his amewo k, we could conside a socially close e e ence
g oup o each agen (indi idual eal-wo ld a m), a ms wi h he same
p oduc i e specializa ion loca ed in he same egion, i.e., neighbo ing
a ms (NBF) co espond o a able c op a ms o he egional uni o
Ka di sa (NUTS-3 le el). In pa icula , a me s o his e e ence g oup
could be conside ed colleagues due o hei simila p o essional goals
and in ense p o essional in e ac ions, which a e exp essed h ough hei
p o essional collec i e bodies, such as ade union bodies, g oups o
p oduce s, and coope a i es, which a e mainly made up o a me s o
common p oduc i e specializa ion. F om his pe spec i e, he in ense
p o essional and, consequen ly, social in e ac ions may p o ide each
agen o he e e ence g oup (neighbo ing a m) wi h a compa a i ely
be e le el o in o ma ion abou he economic pe o mance o i s
neighbo s and he li elihood le el (consump ion le el, pa icula ly o
isual commodi ies ha a e connec ed o income o weal h, e.g., ca s
and houses) (Mageli e al., 2022) han o socially dis an e e ence
g oups (i.e., a ms wi h di e en p oduc i e specializa ions compa ed
o he agen ). Consequen ly, his comp ehensi e in o ma ion signals he
p ocess o o ming social no ms based on which a social g oup’s social
s a us o posi ion is de e mined. In ou case, he ques o social s a us
is e lec ed in KUJ p e e ences (Fishe and Heijd a, 2009; Lomba do,
2021; Mageli e al., 2022). Finally, we also elied on a s ic de ini ion
o neighbo ing a ms o his selec ion based on he ele an li e a u e
(Pa oissien e al., 2021), whe e only a ms wi h he same specializa ion
loca ed in he same egion a e included in he socially close e e ence
g oup (neighbo ing a ms).
As ega ds he ma hema ical o mula ions o he
speci ic p o i abili y measu es a e as ollows conside ing
he ele an li e a u e (GRDC, 2015):
FNPAT* , = Π* , – (DEP , + LRC , + SFNC , +
LFNC , + SIC , + FPTX , ) (5)
FGE* , = FNPAT* , – LE , (6)
whe e FNPAT* , is he op imal Fa m Ne P o i a e Tax
in yea ; Π* , is he op imal g oss p o i o a m in
yea ; DEP , is he dep ecia ion o machine y o a m
in yea ; LRC , a e he land en al cos s7 o a m in
yea ; SFNC , a e he sho - e m inance cos s which
co espond o he in e es paid o sho - e m loans o
a m in yea ; LFNC , a e he long- e m inance cos s
which co espond o he in e es paid o long- e m
loans o a m in yea ; SIC , a e he social insu ance
con ibu ions paid by a m in yea ; FPTX , is he a m
p o i ax paid by a m in yea ; FGE* , is he op imal
Fa m G ow h in Equi y o a m in yea ; LE , a e he
li ing expendi u es8 o a m in yea .
2.3.4. Re-ini ializa ion o esou ces and a m i m g ow h
ules
The annual e-ini ializa ion o esou ces equi ed
o he a ms’ ope a ion and g ow h/expansion p ocess
is conduc ed h ough he Εndogenous Feedback Mecha-
nism (EFM) (whose gene al s uc u e has been p esen -
ed in he 2.1 sec ion). An essen ial pa o he li e a u e
indica es ha g ow h in equi y de e mines he p ospec s
o g ow h/expansion o he a m (e.g., Pain e , 2005;
Ci adini e al., 2008; Be e al., 2011; GRDC, 2015), ha
is, ha he acquisi ion o esou ces will be de e mined
h ough his p o i abili y measu e. Hence, we conside
ha op imal a m g ow h in equi y could be used as an
al e na i e c i e ion o a m expansion/g ow h o a-
7 In case ha a m en s ou pa o owned a mland, hen ecei es land
en al income LRINC , . Consequen ly he equa ion (5) is adap ed as
ollows: FNPAT* , = (Π* , + LRINC , ) – (DEP , + SFNC , + LFNC , +
SIC , + FPTX , ), indica ing ha a a m canno simul aneously en in
and en ou a mland, a condi ion we also ind in simila simula ion
models (e.g., Dona i e al., 2024).
8 The es ima ion o li ing expendi u es ollowing he base yea (2012)
is ca ied ou by u ilizing he li ing expendi u es index (LEI) o
households in u al a eas (ELSTAT, 2021). Tha is, he e ogenei y
be ween a ms in he li ing expendi u es in he base yea (2012) is
cap u ed, bu i s e olu ion o e ime is based on he exogenously
de e mined li ing expendi u es index (LEI). Since he a ailable
ime se ies o he li ing expendi u es index (LEI) does no mee he
minimum equi ed ime ho izon o 16 da a poin s o he ARIMA model
(Ch is odoulos e al., 2010), we use a linea end model ins ead o he
ARIMA model o make pos -sample o ecas s.
360
Bio-based and Applied Economics 13(4): 353-386, 2024 | e-ISSN 2280-6172 | DOI: 10.36253/bae-14790
S ama is Man zia is e al.
di ional c i e ia such as he shadow alues o esou ces
(e.g., land, ci cula ing capi al) (Guinde e al., 2005; Hen-
nessy, 2007; Espinosa e al., 2016).
Howe e , gi en esou ce cons ain s, especially
land, a m expansion is possible when neighbo ing
a ms decide o downsize o abandon ag icul u al p o-
duc ion (Plogmann e al., 2022). Essen ially, he p o-
cess o s uc u al change d i es he ealloca ion o he
esou ces equi ed o expansion, whe e he esou ces o
non- iable neighbo ing a ms (e.g., land) a e ealloca ed
o iable ones (see also Figu e 1). Va ious modele s (e.g.,
Be e al., 2011; Sheng e al., 2015; He e a e al., 2022;
Sun e al., 2022) highligh he ole o ela i e p o i abil-
i y as a c i e ion/mechanism o he ealloca ion/alloca-
ion o esou ces be ween su i ing a ms. Wi hin his
con ex , ou conce n was how op imal a m g ow h in
equi y could be exp essed as a c i e ion/mechanism o
esou ce ealloca ion among iable a ms and in eg a ed
in o he EFM. To model his mechanism, we adap ed
he concep o e icien alloca ion (Aye s e al., 2020;
Chen e al., 2022). Acco ding o he p oposed adap a-
ion, we eplace ela i e a ming p oduc i i y wi h ela-
i e a m g ow h in equi y. We conside his adjus men
o be easonable since Fos e e al. (2008) ound ha
“ i ms’ sel -selec ion beha io (in choosing an ope -
a ing scale, o o en e o exi ) is made based on i m
p o i abili y a he han i m p oduc i i y and conse-
quen ly esou ce ealloca ion may no always align wi h
i m p oduc i i y g ow h, pa icula ly in he sho un”
(Sheng e al., 2015: 75).
By inco po a ing he p oposed esou ce ealloca-
ion/alloca ion mechanism in o he EFM, each a m’s
annual le el o esou ce is de e mined by he a ailable
le el o he esou ce a he beginning o he p e ious
g owing season, he ela i e op imal g ow h in equi y a
he end o he p e ious g owing season (indica ing he
op imal decisions), and by exogenous e en s9 ha will
occu in he cu en g owing season.
Since we ha e ensu ed ( om he iabili y de e mi-
na ion assump ions) ha a iable a m will no e eal
nega i e op imal g ow h in equi y, he ma hema ical
o mula ion o he sha e o any esou ce ∈ {AL,CRC}
alloca ed o ealloca ed is as ollows:
Ωsim ,nb , = , o = 1… T,
0 ≤ Ωsim ,nb , (7)
9 We assume ha exogenous e en s a e exp essed h ough successi e
di e ences in he agg ega e le el o esou ces whe e he ela i e op imal
g ow h in equi y o he p e ious g owing season alloca es hese posi i e
o nega i e di e ences ac oss a ms.
whe e Ωsim ,nb , is he simula ed sha e o esou ce allo-
ca ed/ ealloca ed o iable neighbo ing a m in yea ;
FGE* ,nb , is he op imal Fa m G ow h in Equi y o ia-
ble neighbo ing a m in yea ; FGE* ,nb ,
is he agg ega e op imal Fa m G ow h in Equi y o ia-
ble neighbo ing a ms in yea .
Essen ially he simula ed sha e o esou ce allo-
ca ed o iable neighbo ing a m in pe iod (Ωsim ,nb , )
exp esses he pa o EFM which co esponds o op imal
decisions (Xj *) while conside ing he in e dependence o
op imal decisions o iable neighbo ing a ms, indica -
ing compe i i eness o esou ces. I is also wo h no ing
ha he simula ed sha e (Ωsim ,nb , ) emains he same o
each esou ce alloca ed/ ealloca ed.
(i) A able land
The e o e, conside ing he abo e, he EFM mecha-
nism o he esou ce o a able land will be o mula ed
as ollows:
AL ,nb , = AL ,nb , -1 + ΩsimAL , -1 [
ALn ,nb , -1sim + (TALNBF, – TALNBF, -1)], o =2…T (8)
whe e AL ,nb , is he a ailable a able land o iable
neighbo ing a m in yea ; ALn ,nb , -1 is he a ailable
a able land o iable neighbo ing a m a he beginning
o yea -1; ΩsimAL ,nb , -1 is he simula ed sha e o a able
land ealloca ed o iable neighbo ing a m a he end o
he yea -1, ha is, ollowing he annual op imiza ion;
ALn ,nb , -1sim is he simula ed agg ega e
a able land o non- iable neighbo ing a ms a he end
o he yea -1, ha is, ollowing he annual op imiza-
ion; TALNBF, is he ac ual o al a able land o neighbo -
ing a ms in yea ; TALNBF, -1 is he ac ual o al a able
land o neighbo ing a ms in yea -1.
Essen ially, he p oduc ΩsimAL ,nb , -1(TALNBF, –
TALNBF, -1) co esponds o he ec o Vi o EFM ha allows
o ex e nal changes in he esou ce es ic ions due o
exogenous e en s, and p obably e lec s he compe i ion
o esou ces wi h o he ypes o a ms o non-ag icul u al
sec o s which ope a e wi hin he same egion.
Howe e , compe i i e p essu es a e likely o lead o
an un a o able si ua ion, i.e., TALNBF, – TALNBF, -1 < 0
and consequen ly o a dec ease o a ailable a able land
o he iable neighbo ing a ms, which will be eal-
loca ed among hem u ilizing he in e se o m o he
simula ed sha e o a able land (ΩsimAL ,nb , -1
-1), ha is, less
p o i able albei iable a ms will abandon p opo ion-
a ely mo e o hei a able land.
As can be easily unde s ood by he eade , he abo e
p ocedu e is also applied o he a ailable i iga ed land
367
Simula ing a m s uc u al change dynamics in Thessaly (G eece) using a ecu si e p og amming model
Bio-based and Applied Economics 13(4): 353-386, 2024 | e-ISSN 2280-6172 | DOI: 10.36253/bae-14790
scena ios is p ojec ed o enhance he p o i abili y o
hese a ms u he . I is also wo h no ing ha be ween
hese wo scena ios, no subs an ial di e ences can be
ound in he e olu ion o p o i abili y.
3.4. Simula ed land use change
As ega ds he simula ed land use change dynam-
ics illus a ed in Figu e 5, he main change can be seen
in he p og essi e expansion o he p ocessing ege able
a ea and especially o p ocessing peppe . This inding
ho oughly e i ies a me s’ expec a ions o he u he
expansion o hese c ops. In pa icula , he p ocess-
ing peppe a me s o he sample s a e ha hei expo
ac i i y will inc ease signi ican ly in he coming yea s
since hey ecei e mo e han double commodi y p ices
compa ed o domes ic p ices. P ocessing oma o a m-
e s aspi e o a signi ican expansion o hei p oduc-
i e ac i i y due o he posi i e g ow h p ospec s o he
local oma o p ocessing indus y, as hey also conside
he ole o he local g oup o p ocessing oma o a me s
o be pa icula ly bene icial. An inc easing end in he
p ocessing ege able a ea is simula ed o bo h scena ios.
S ill, a mo e signi ican upwa d end is simula ed o
he CAP Pos -2020 e o m scena io, possibly due o he
inc eased a e o s uc u al change leading o mo e e i-
cien use o esou ces, in he sense ha su i ing a ms
end o alloca e a mland a ea o compa a i ely mo e
p o i able ac i i ies20.
Con e sely, we simula ed a signi ican g adual
dec ease in he co on and obacco a eas. In ac , o
he CAP Pos -2020 e o m scena io, we obse e a u -
20 De ails a e p o ided in Table A1 in he Appendix.
Figu e 4. E olu ion o simula ed a e age Fa m Ne P o i a e Tax (FNPAT) by scena io. No e: The p o isions o he CAP Pos -2020 sce-
na io apply om he yea 2023. Sou ce: Au ho s, based on sample da a.
Table 4. Simula ed mean Fa m Ne P o i a e Tax (FNPAT) in EUR by a m size classes (2012-2026).
Fa m size class in ha
(Cha ac e iza ion) 2012 2019 2026 (BAU scena io) 2026 (CAP Pos -2020
scena io)
2026 (CAP Pos -2020
& LWA scena io)
<10 (Ve y Small) 18,156 13,706 12,196 - -
10-<30 (Small) 41,931 26,707 25,989 24,987 25,010
30-<50 (Medium) 59,829 68,250 98,608 104,309 107,964
50-<100 (La ge) 110,370 69,918 115,803 122,013 126,539
≥100 (Ve y La ge) - 695,181 1.738,668 1.855,216 1.865,563
Agg ega e 39,526 84,644 271,183 323,692 327,622
No e: The de e mina ion and cha ac e iza ion o a m size classes is based on Happe e al. (2008), and Hue el & Ma ga ian (2009).
Sou ce: Au ho s, based on sample da a.

368
Bio-based and Applied Economics 13(4): 353-386, 2024 | e-ISSN 2280-6172 | DOI: 10.36253/bae-14790
S ama is Man zia is e al.
he educ ion o he co on and obacco a eas. Du um
whea a ea inc eases signi ican ly o e ime o he BAU
scena io, while o he CAP Pos -2020 e o m scena io,
a dec ease a e 2022 is o eseen due o he se -aside
applied by he as majo i y o sample a ms (mo e han
90%) in he con ex o eco-scheme paymen s. Based on
his inding, we conclude ha a ms ha e a s ong incen-
i e o adop eco-schemes since he majo i y exceed 10
hec a es and, he e o e, would be equi ed o implemen
se -aside on 4% o a able land wi hou ex a paymen .
In he CAP Pos -2020 & LWA scena io, an expec ed
inc ease is simula ed o he a ea o he g ain (du um
whea , maize), especially maize, due o he possible
inc ease and main enance o a m ga e p ices a high
le els due o he Uk ainian c isis. Acco dingly, a u he
educ ion in co on and obacco a eas is simula ed.
4. DISCUSSION AND CONCLUSIONS
Dynamic modeling me hodologies a e deemed
c ucial o comp ehending he e olu ion o economic
agen s’ beha io s in esponse o shi s in he economic
en i onmen o policies (Ga deb oek and Oude Lansink,
2008). Conside ing he ola ile economic en i onmen
in which a ms ope a e due o ecen in e na ional de el-
opmen s, such assessmen s gain signi ican weigh when
using simula ion models like he one we p opose he ein
since hey can suppo policy analys s in o mula ing
and speci ying he app op ia e policy measu es.
In his con ex , his s udy desc ibed he concep ual
amewo k o a newly de eloped a m-le el ecu si e
linea p og amming model p ima ily aiming a simu-
la ing he impac o policy e o m on s uc u al change
in he a able p oduc ion sys em o he egion o Ka -
di sa (NUTS-3 le el), one o he cen al g owing egions
o a able c ops in G eece. While managing o cap u e
mainly endogenously he dynamics o s uc u al change
adap a ion, he p oposed simula ion model can simul-
aneously be cha ac e ized by a compa a i ely low le el
o modeling complexi y compa ed o o he simula ion
models, such as agen -based models.
F om a gene al pe spec i e, his pape seeks o con-
ibu e o he deba e on dynamic assessmen s o he
mul idimensional e ec s in he con ex o he CAP Pos -
2020 e o m while conside ing ecen geopoli ical de el-
opmen s in he con ex o he Uk ainian c isis.
Valida ion esul s demons a e sa is ac o y pe -
o mance o he simula ion model in ep oducing pas
changes. The e o e, we can use he model o assess he
e ec s o a ious scena ios on he ag icul u al p oduc-
ion sys em. By ca ying ou policy expe imen s o
wo di e en policy scena ios and a combined scena -
io (policy and geopoli ical) we es ima ed an inc eased
a e o s uc u al change compa ed o he e e ence
pe iod (2012-19), and especially o he CAP Pos -2020
and CAP Pos -2020 & Long Wa o A i ion (LWA)
scena ios. The p oposed model simula ed an e iden
g adual concen a ion o a mland in ela i ely la ge
a ms ( a m size ≥50 ha), accompanied by a dec ease
in he numbe o ela i ely small a ms ( a m size < 30
ha), making hese indings consis en wi h he esul s
ob ained om simula ion models (e.g., Happe e al.,
2008; Be e al., 2011; Dona i e al., 2024) and o he
dynamic modeling app oaches (He e a e al., 2022;
Schuh e al., 2022).
(a) BAU scena io
(b) CAP Pos -2020 scena io
(c) CAP Pos -2020 & LWA scena io
Figu e 5. Simula ed a able land alloca ion by scena io. No e: The
p o isions o he CAP Pos -2020 e o m scena io apply om he
yea 2023. Sou ce: Au ho s, based on sample da a.
369
Simula ing a m s uc u al change dynamics in Thessaly (G eece) using a ecu si e p og amming model
Bio-based and Applied Economics 13(4): 353-386, 2024 | e-ISSN 2280-6172 | DOI: 10.36253/bae-14790
Rega dless o he examined scena io, he simula -
ed a e age a m p o i abili y shows a g adual inc ease,
which is pa ly explained by he ac ha ela i ely
mo e p o i able a ms emain in he p oduc ion sys em
con i ming p e ious indings ob ained om simula-
ion models (Happe e al., 2008; Be e al., 2011) and
o he dynamic modeling app oaches (He e a e al.,
2022; Schuh e al., 2022). Ob iously, he su i ing a ms
which achie e g ow h in equi y end o alloca e hei
g owing esou ces (such as a mland, ci cula ing capi al
and ixed asse s) mo e e icien ly, i.e., o ela i ely mo e
p o i able p oduc i e ac i i ies (in ou case, p ocess-
ing ege ables), u he enhancing a e age a m p o i -
abili y (Be e al., 2011). Howe e , a downwa d end is
simula ed o he a e age p o i abili y o ela i ely small
a ms ( a m size < 30 ha).
In e ms o land use change dynamics, ega dless o
he scena io, ou model simula ed an inc easing end o
he land alloca ed o ood c ops such as p ocessing ege a-
bles and a simul aneous dec easing end o he a mland
alloca ed o indus ial c ops such as co on and obacco.
The a ionale explains his esul discussed ea lie , namely
ha su i ing a ms end o expand p oduc i e ac i i ies
wi h compa a i ely highe p o i abili y, a inding ha is
also consis en wi h indings ob ained om a simula ion
model applied o he ag icul u al sys em o he A gen ine
Pampas (Be e al., 2011). Addi ionally, Be e al. (2011)
conside ha his beha io o he a ms is in e p e ed by
hei su i al s a egy. Conside ing he abo e, i could be
said ha a co ela ion o land use change wi h s uc u al
change eme ges, in he sense ha he iabili y o a ms
is s ongly dependen on he land use chosen (Be e al.,
2011) and is exp essed h ough hei su i al s a egy o
alloca e hei a mland a ea and capi al o he mos p o i -
able c opping ac i i y g adually.
Focusing on he pape ’s main inding – namely, he
ag icul u al p oduc ion concen a ion in ela i ely la ge
a ms ( a m size ≥ 50 ha) – i is ound ha his has some
signi ican policy implica ions. In pa icula , an in ensi-
ying con inua ion o p essu es owa ds ewe bu la ge
a ms (i.e., an inc easing a e o s uc u al change) could
lead o a b eakdown o social cohesion, a p e equisi e
o add essing u al communi ies’ challenges (Knu son
e al., 1986). F om his pe spec i e, app op ia e policy
measu es could ocus, o example, on he enhancemen
o a me s’ ma ke access since small and medium-sized
a ms ha e issues accessing ma ke s, achie ing a p ope
sha e in he EU ood chain, including alue-added p o-
cessing, and main aining ba gaining powe (Schuh e al.,
2022). In his ein, coope a i es a e one way o imp o e
a me s’ access o ma ke s and s eng hen ba gaining
powe , p ima ily h ough e ical in eg a ion, which can
o en play a signi ican ole in inc easing he economic
bene i s o a me s (Schuh e al., 2022). The e o e, i is
essen ial o p io i ize examining exempla y coope a i e
p ac ices and suppo ing he adop ion o simila ope a-
ional models h ough policy ac ions (Schuh e al., 2022).
E en i essen ial insigh s we e gained, his modeling
exe cise is cha ac e ized by se e al ca ea s, whe e we will
ocus on he main ones. Fi s , al hough he p oposed
ecu si e linea p og amming model u ilizes inpu da a
o ep esen a i e indi idual eal-wo ld a ms, e ec i ely
cap u ing he he e ogenei y in a m s uc u e and epli-
ca ing a ied a m beha io , i does no explici ly cap u e
he in e ac ion be ween indi idual a ms in he sense o
no inco po a ing an endogenous p ice o ma ion mech-
anism o he ma ke o locally a ailable esou ce like
land (Be ge , 2001; T oos and Be ge , 2015; K emmy-
das, 2019). Addi ionally, i does no ully conside spa ial
ela ionships, o e looking he impe ec land alloca ion
among a ms by dis ega ding in e nal anspo cos s
and he physical immobili y o land (Be ge , 2001; T oos
and Be ge , 2015; K emmydas, 2019). In his con ex , he
de e mina ion o he egional le el a which a ms can be
ega ded as compe i o s o he a mland o e ed is le o
he subjec i i y o he modele . Al hough adminis a i e
uni s a e o en used as a ealis ic app oach (in ou case,
he egional uni o Ka di sa (NUTS-3 le el)), ideally, he
egional le el could be de ined by he iewpoin o ac i e
a me s who ope a e he land (Plogmann e al., 2022).
Consequen ly, hese weaknesses o he p oposed mod-
el limi i s abili y o ully cap u e in e ac ions be ween
a ms and spa ial dynamics, limi ing i s explana o y
powe in policy analysis. Especially, he model canno
p o ide de ailed insigh s in o he impac s o policy sce-
na ios/op ions on a m s uc u e due o hei e ec s on
local esou ce ma ke s (K emmydas, 2019). Fu he mo e,
he incomple e inco po a ion o spa ial dynamics cu ails
he model’s explana o y capaci y ega ding policy e ec s
on he en i onmen , whe e spa ial aspec s hold conside -
able impo ance (K emmydas, 2019).
Second, al hough he p oposed simula ion model con-
side s he di e ences in p o i s among neighbo ing a ms
cul i a ing di e en a mland a eas in he base yea , p o-
iding a easonable ep esen a ion o he a m g ow h p o-
cess, i does no conside economies o scale in an in e -
empo al con ex . The cap u e o economies o scale a a
longi udinal le el by he p oposed model was no ca ied
ou o main ain i s compu a ional complexi y. Howe e , a
mo e de ailed model ha conside s his dimension could
enhance he ep esen a ion o a m he e ogenei y and,
consequen ly, policy ep esen a ion owa ds a mo e ealis-
ic amewo k. The e o e, u u e de elopmen s o he p o-
posed simula ion model could inco po a e cos educ ions
370
Bio-based and Applied Economics 13(4): 353-386, 2024 | e-ISSN 2280-6172 | DOI: 10.36253/bae-14790
S ama is Man zia is e al.
as a unc ion o a m expansion and/o echnological p o-
g ess (Happe e al., 2008; Be e al., 2011).
Thi d, due o he lack o a m-le el da a o he
in e im yea s o he e e ence pe iod, we we e o ced o
use he a ailable na ional-le el ime se ies o pa am-
e e s o in e es o b idge he ime se ies da a gap a he
a m le el. Howe e , a ious au ho s ha e highligh ed
and documen ed he s a is ical di e ences be ween
egional/na ional and a m-le el ime se ies da a asso-
cia ed wi h unde es ima ion o a iabili y (e.g., Deb ah
and Hall, 1989). In pa icula , agg ega ed da a ends o
unde es ima e he a iabili y o pa ame e s such as p ic-
es and yields a he a m le el (Deb ah and Hall, 1989),
which may lead o a less adequa e ep esen a ion o eal-
i y ega ding a ms’ beha io and adap a ion.
This modeling exe cise has iden i ied many a enues
o u he esea ch, highligh ing only a ew. Fi s , he
geog aphical and sec o al co e age should be expanded.
Second, i is o pa icula impo ance o un simula ions
using al e na i e alloca ion/ ealloca ion mechanisms o
esou ces, such as ela i e shadow alues o esou ces.
Thi d, an in e es ing a enue o u he esea ch is o
conduc an en i onmen al impac assessmen by u iliz-
ing mean- and e ec -based indica o s (Lebacq e al.,
2013; Dona i e al., 2024) bu also o inco po a e social
indica o s, allowing us o assess sus ainabili y pe o -
mance a he a m le el (e.g., Lai ez e al., 2023). Finally,
u he esea ch could be conduc ed on he in es iga ion
o a m iabili y using al e na i e mone a y and socio-
economic iabili y c i e ia.
To conclude, al hough ou modeling esul s may
no ep esen all G eek egions, hey may be pa icula ly
in o ma i e o ends ha may eme ge due o s uc u al
and land-use changes in u al a eas wi h simila a able
p oduc ion sys ems, no only in he coun y bu also in
he wide Medi e anean a ea.
ACKNOWLEDGEMENTS
This esea ch is co- inanced by G eece and he Eu o-
pean Union (Eu opean Social Fund-ESF) h ough he
Ope a ional P og amme “Human Resou ces De elop-
men , Educa ion and Li elong Lea ning” in he con ex
o he p ojec “S eng hening Human Resou ces Resea ch
Po en ial ia Doc o a e Resea ch” (MIS-5000432), imple-
men ed by he S a e Schola ships Founda ion (ΙΚΥ).
REFERENCES
Aye s , S., B and , L., and Res uccia, D. (2020). Ma -
ke cons ain s, misalloca ion, and p oduc i i y in
Vie nam ag icul u e, Food Policy, 94. h ps://doi.
o g/10.1016/j. oodpol.2020.101840
Baldi, L., Calzolai, S., A ini, F., & Dona i, M. (2024). P e-
dic ing he e ec o he Common Ag icul u al Poli-
cy pos -2020 using an agen -based model based on
PMP me hodology. Bio-Based and Applied Econom-
ics. h ps://doi.o g/10.36253/bae-14592
Balmann, A. (1997). Fa m-based modelling o
egional s uc u al change: A cellula au oma a
app oach. Eu opean Re iew o Ag icul u al Eco-
nomics, 24(1): 85–108. h ps://doi.o g/10.1093/
e ae/24.1.85
Balmann, Al., and Valen ino , V. (2016). Towa ds a he-
o y o s uc u al change in ag icul u e: Jus econom-
ics? 149 Th EAAE Semina : S uc u al Change in
Ag i-Food Chain, New Rela ions be ween Fa m Sec-
o , Food Indus y and Re ail Sec o , 3. h ps://doi.
o g/10.22004/ag.econ.246420
Ba nes, A. P., Hansson, H., Mane ska-Tase ska, G.,
Sh es ha, S. S., and Thomson, S. G. (2015). The in lu-
ence o di e si ica ion on long- e m iabili y o he
ag icul u al sec o . Land Use Policy, 49, 404–412. h -
ps://doi.o g/10.1016/j.landusepol.2015.08.023
Ba ne , R. C., Bha acha ya, J., and Bunzel, H. (2010).
Choosing o keep up wi h he Joneses and income
inequali y. Economic Theo y, 45(3): 469–496. h ps://
doi.o g/10.1007/s00199-009-0494-5
Becke s, V., Becke s, J., Vanmae cke, M., an Hecke, E., an
Rompaey, A., and Dendoncke , N. (2018). Modelling
a m g ow h and i s impac on ag icul u al land use:
A coun y scale applica ion o an agen -based model.
Land, 7(3). h ps://doi.o g/10.3390/land7030109
Be ge , T. (2001). Agen -based spa ial models applied o
ag icul u e: a simula ion ool o echnology di u-
sion, esou ce use changes and policy analysis. Ag i-
cul u al Economics, 25(2-3): 245-260. h ps://doi.
o g/10.1111/j.1574-0862.2001. b00205.x
Be , F. E., Podes á, G. P., Ro e e, S. L., Menéndez, Á. N.,
No h, M., Ta a a, E., Laciana, C. E., Webe , E., and
To anzo, F. R. (2011). An agen -based model o sim-
ula e s uc u al and land use changes in ag icul u al
sys ems o he A gen ine pampas. Ecological Model-
ling, 222(19): 3486–3499. h ps://doi.o g/10.1016/j.
ecolmodel.2011.08.007
B igh , G., Flo ey, B., and Adams, J. (2007). Indica o s
and de e minan s o a m inancial success. Jou -
nal o Fa m Managemen 13(2): 75-93. h ps://www.
iag m.com/ iles.php? ile=la ge/jou nals/jo m/ ol-
ume_13/jo m- ol13-no2-pages-75-93.pd
B i z, W., Lenge s, B., Kuhn, T., and Schä e , D. (2016). A
highly de ailed empla e model o dynamic op imi-
za ion o a ms-FARMDYN. Model documen a ion.
371
Simula ing a m s uc u al change dynamics in Thessaly (G eece) using a ecu si e p og amming model
Bio-based and Applied Economics 13(4): 353-386, 2024 | e-ISSN 2280-6172 | DOI: 10.36253/bae-14790
Ins i u e o Food and Resou ce Economics, Uni e -
si y Bonn. h ps://www.il 1.uni-bonn.de/en/ esea ch/
esea ch-g oups/economic-modeling-o -ag icul u al-
sys ems/ a mdyn
Chen, C., Res uccia, D., and San aeulàlia-Llopis, R.
(2022). The e ec s o land ma ke s on esou ce
alloca ion and ag icul u al p oduc i i y. Re iew
o Economic Dynamics, 45, 41–54. h ps://doi.
o g/10.1016/j. ed.2021.04.006
Chien, Y. I., and B ad o d, G. L. (1976). A Sequen ial
Model o he Fa m Fi m G ow h P ocess. Ame ican
Jou nal o Ag icul u al Economics, 58(3): 456-465.
h ps://doi.o g/10.2307/2290724
Ch is odoulos, C., Michalakelis, C., and Va ou as, D.
(2010). Fo ecas ing wi h limi ed da a: Combining
ARIMA and di usion models. Technological Fo e-
cas ing and Social Change, 77(4), 558–565. h ps://
doi.o g/10.1016/j. ech o e.2010.01.009
Ciaian, P., Espinosa, M., Paloma, G.Y., Heckelei, T.,
Sckokai, P., Lang ell, S., Louhichi, K., Thomas, A.,
and Va d, T. (2013). Fa m le el modelling o he
CAP: a me hodological o e iew. EUR 25873. Lux-
embou g (Luxembou g): Publica ions O ice o he
Eu opean Union; 2013. JRC79969. h ps://pub-
lica ions.j c.ec.eu opa.eu/ eposi o y/bi s eam/
JRC79969/j c79969online inal.pd
Ci adini, E. D., Lubbe s, M. T. M. H., de Ridde , N., an
Keulen, H., and Claassen, G. D. H. (2008). Explo ing
op ions o a m-le el s a egic and ac ical decision-
making in ui p oduc ion sys ems o Sou h Pa ago-
nia, A gen ina. Ag icul u al Sys ems, 98(3): 189–198.
h ps://doi.o g/10.1016/j.agsy.2008.07.001
Coch ane, W. (1958). Fa m P ices: My h and Reali y.
Minneapolis: Uni e si y o Minneso a P ess, 1958.
h ps://sea chwo ks.s an o d.edu/ iew/9972413
Cos ache, M., Sebas ian C is ea, D., Pe ea, S. M., Necu-
li a, M., Raho eanu, M. M. T., Simiono , I. A., Mog-
odan, A., Sa pe, D., and Raho eanu, A. T. (2021).
In eg a ing aquaponics p oduc ion sys ems in o
he Romanian g een p ocu emen ne wo k. Land
Use Policy, 108. h ps://doi.o g/10.1016/j.landuse-
pol.2021.105531
Da is, J., Caskie, P., and Wallace, M. (2013). P omo ing
s uc u al adjus men in ag icul u e: The economics
o New En an Schemes o a me s. Food Policy, 40,
90–96. h ps://doi.o g/10.1016/j. oodpol.2013.02.006
Day, R. H., and Cigno, A. (1978). Modelling economic
change: The ecu si e p og amming app oach. Day,
R H, Cigno, A (eds). Modelling Economic Change.
Else ie No h-Holland, Inc.
Deb ah, S., and Hall, H. H. (1989). Da a Agg ega ion and
Fa m Risk Analysis. In Ag icul u al Sys ems (Vol. 31).
Djanibeko , U., and Finge , R. (2018). Ag icul u al isks
and a m land consolida ion p ocess in ansi ion
coun ies: The case o co on p oduc ion in Uzbeki-
s an. Ag icul u al Sys ems, 164, 223–235. h ps://doi.
o g/10.1016/j.agsy.2018.03.009
Dowson, O., Philpo , A., Mason, A. and Downwa d, A.
(2019). A mul i-s age s ochas ic op imiza ion model
o a pas o al dai y a m. Eu opean Jou nal o Ope a-
ional Resea ch 274: 1077–1089.
ELSTAT (2021). Household Budge Su ey (a e 2008),
2020. Hellenic S a is ical Au ho i y. h ps://www.s a-
is ics.g /en/s a is ics/-/publica ion/SFA05/2020
Espinosa, M., Goch , A., Heckelei, T., and Paloma, S. G.
y. (2016). Inco po a ing a m s uc u al change in
models assessing he Common Ag icul u al Policy:
An applica ion in he CAPRI a m ype model. Jou -
nal o Policy Modeling, 38(6): 1040–1059. h ps://doi.
o g/10.1016/j.jpolmod.2016.03.005
Eu opean Council (2022). In og aphic - Uk ainian g ain
expo s explained. h ps://www.consilium.eu opa.eu/
en/in og aphics/uk ainian-g ain-expo s-explained/
FADN Public Da abase. Fa m Accoun ancy Da a Ne -
wo k Public Da abase. Di ec o a e-Gene al o Ag i-
cul u e and Ru al De elopmen . h ps://ag ida a.
ec.eu opa.eu/ex ensions/FADNPublicDa abase/FAD-
NPublicDa abase.h ml
Femenia, F., La u e, L., and Cha as, J. P. (2017). Respon-
si eness o a m in es men o p ice changes: An
empi ical s udy o he F ench c op sec o . Jou nées
de Mic oéconomie Appliquées (JMA), G oupe
d’Analyse des I iné ai es e Ni eaux Sala iaux
(GAINS). FRA. Uni e si é du Maine, FRA., Jun 2017,
Le Mans, F ance. h ps://hal.science/hal-01571567
Finge , R., and el Benni, N. (2021). Fa m income in Eu o-
pean ag icul u e: New pe spec i es on measu emen
and implica ions o policy e alua ion. Eu opean
Re iew o Ag icul u al Economics, 48(2): 253-265.
h ps://doi.o g/10.1093/e ae/jbab011
Finge , J. M., and K einin, M. E. (1979). A Meas-
u e o ‘Expo Simila i y’ and I s Possible Uses.
The Economic Jou nal, 89(356). h ps://doi.
o g/10.2307/2231506
Fishe , W. H., and Heijd a, B. J. (2009). Keeping up wi h
he ageing Joneses. Jou nal o Economic Dynamics
and Con ol, 33(1): 53–64. h ps://doi.o g/10.1016/j.
jedc.2008.04.010
Fos e , L., Hal iwange , J., and Sy e son, C. (2008). Reallo-
ca ion, i m u no e , and e iciency: Selec ion on p o-
duc i i y o p o i abili y? Ame ican Economic Re iew,
98(1): 394-425. h ps://doi.o g/10.1257/ae .98.1.394
F eeman, T., Nolan, J., and Schoney, R. (2009). An agen -
based simula ion model o s uc u al change in
372
Bio-based and Applied Economics 13(4): 353-386, 2024 | e-ISSN 2280-6172 | DOI: 10.36253/bae-14790
S ama is Man zia is e al.
Canadian p ai ie ag icul u e, 1960-2000. Canadian
Jou nal o Ag icul u al Economics, 57(4): 537–554.
h ps://doi.o g/10.1111/j.1744-7976.2009.01169.x
Ga deb oek, C., and Oude Lansink, A., 2008. Dynamic
mic oeconome ic app oaches o analysing ag icul u al
policy. Pape p esen ed a he 107 h EAAE Semina
“Modelling o Ag icul u al and Ru al De elopmen
Policies”, Janua y 29–Feb ua y 1, 2008, Se ille, Spain.
Ga nie , H. (n.d.). In oduc ion o ime se ies analysis and
o ecas ing. Uni e si y o Lo aine. h p://w3.c an.
uni -lo aine. /pe so/hugues.ga nie /Enseignemen /
TSAF/C-TSAF-Box-Jenkins_me hod.pd
Godda d, E., Wee sink, A., Chen, K., and Tu ey, C. G.
(1993). Economics o S uc u al Change in Ag icul-
u e. Canadian Jou nal o Ag icul u al Economics/
Re ue Canadienne d’ag oeconomie, 41(4): 475-489.
h ps://doi.o g/10.1111/j.1744-7976.1993. b03772.x
Gómez-Limón, J. A., Gu ié ez-Ma ín, C., and Riesgo, L.
(2016). Modeling a a m le el: Posi i e Mul i-A ib-
u e U ili y P og amming. Omega (Uni ed Kingdom),
65: 17-27. h ps://doi.o g/10.1016/j.omega.2015.12.004
G eek Minis y o Ru al De elopmen and Food (2014).
The con ex o Common Ag icul u al Policy a e
2015-Fi s Pilla : Di ec Paymen s (in G eek). h p://
minag ic.g /images/s o ies/ag opol/G eek/Ag o_pol/
KAP/Pa ousiasiFINAL2015_290715.pd
G eek Minis y o Ru al De elopmen and Food
(2022). G eek S a egic Plan p oposal o he CAP
2023-2027 (in G eek). h ps://ead.g /wp-con en /
uploads/2022/01/cap_sp_p oposal_30_12_2021.pd
GRDC (2015). How Do I Measu e The Financial Pe o -
mance O My Fa m Business? G ains Resea ch and
De elopmen Co po a ion (GRDC).
Guinde, L., Mille , G., Rozakis, S., Sou ie, J.-C., and
T égue , D. (2005). The CAP Mid-Te m Re o m
Impac s o F ench Ce eal-O ien ed Fa ms. P o-
ceedings o he 89 h EAAE Semina .10.22004/
ag.econ.232607
Haile, M. G., Kalkuhl, M., and on B aun, J. (2016). Wo ld-
wide ac eage and yield esponse o in e na ional p ice
change and ola ili y: A dynamic panel da a analysis
o whea , ice, co n, and soybeans. In Food P ice Vola-
ili y and I s Implica ions o Food Secu i y and Policy.
h ps://doi.o g/10.1007/978-3-319-28201-5_7
Happe, K., Balmann, A., Kelle mann, K., and Sah bach-
e , C. (2008). Does s uc u e ma e ? The impac o
swi ching he ag icul u al policy egime on a m
s uc u es. Jou nal o Economic Beha io and O gan-
iza ion, 67(2): 431–444. h ps://doi.o g/10.1016/j.
jebo.2006.10.009
Hennessy, T. C. (2007). Modelling he E ec o Policy
Re o m on S uc u al Change in I ish Fa ming.
I ish Ag icul u e and Food De elopmen Au ho i y,
Ru al Economy Resea ch Cen e, Ionad Taighde Eac-
namaíoch a Tua ha.
Henningsen, A., Henning, C., S u e, C., and Mülle -
Scheessel, J. (2005). Economic impac o he Mid-Te m
Re iew on ag icul u al p oduc ion, a m income and
a m su i al: A quan i a i e analysis o local sub-
egions o Schleswig-Hols ein in Ge many, 11 h EAAE
Cong ess, Augus 24-27, 2005. Copenhagen, Denma k.
He e a, H., Schü z, L., Paas, W., Reidsma, P., and
Kopainsky, B. (2022). Unde s anding esilience o
a ming sys ems: Insigh s om sys em dynamics
modelling o an a able a ming sys em in he Ne h-
e lands. Ecological Modelling, 464: 109848. h ps://
doi.o g/10.1016/j.ecolmodel.2021.109848
Hlouško á, Z., Ženíško á, P., and P ášilo á, M. (2018).
Compa ison o ag icul u al cos s p edic ion app oach-
es. Ag is On-Line Pape s in Economics and In o ma -
ics, 10(1). h ps://doi.o g/10.7160/aol.2018.100101
Huang, W.-Y., Weisz, R. N., & Heady, E. O. (1980). An
econome ic-p og amming model o ag icul u al pol-
icy analysis. CARD Repo s. 89, Cen e o Ag icul-
u al and Ru al De elopmen , Iowa S a e Uni e si y.
h p://lib.d .ias a e.edu/ca d_ epo s/89
Hue el, S., and Ma ga ian, A. (2009). S uc u al change
in he Wes Ge man ag icul u al sec o . Ag icul u al
Economics, 40: 759–772. h ps://doi.o g/10.1111/
j.1574-0862.2009.00413.x
Iglesias, E., Ga ido, A., and Gómez-Ramos, A. (2003).
E alua ion o d ough managemen in i iga ed a eas.
Ag icul u al Economics, 29(2): 211–229. h ps://doi.
o g/10.1016/S0169-5150(03)00084-7
Kay, R. D. (1971). A dynamic linea p og amming model
o a m i m g ow h in No h Cen al Iowa. Re o-
spec i e Theses and Disse a ions. 4470, Iowa S a e
Uni e si y. h ps://lib.d .ias a e.edu/ d
Khanal, A. R., and Omobi an, O. (2020). Ru al Finance,
Capi al Cons ained Small Fa ms, and Financial Pe -
o mance: Findings om a P ima y Su ey. Jou nal
o Ag icul u al and Applied Economics, 52(2). h -
ps://doi.o g/10.1017/aae.2019.45
Knu son, R., Richa dson, J., and Smi h, E. (1986). Impac o
Fa m Policies on Ag icul u e, Fa m S uc u e, and Ru al
Communi ies. Jou nal o Ru al Social Sciences, 4(1).
K emmydas, D. (2019). In eg a ed ma hema ical p o-
g amming and agen based model o policy analysis
o G eek ag icul u e. PhD Thesis, Ag icul u al Uni-
e si y o A hens.
K emmydas, D., Ciaian, P., and Baldoni, E. (2023). Mod-
eling con e sion o o ganic ag icul u e wi h an EU-
wide a m model. Bio-Based and Applied Economics,
12(4), 261–304. h ps://doi.o g/10.36253/bae-13925

373
Simula ing a m s uc u al change dynamics in Thessaly (G eece) using a ecu si e p og amming model
Bio-based and Applied Economics 13(4): 353-386, 2024 | e-ISSN 2280-6172 | DOI: 10.36253/bae-14790
Lai ez, J., Jou dain, D., Lopez-Ridau a, S., Sy ongxay,
C., and A holde , F. (2023). Mul ic i e ia assess-
men o al e na i e c opping sys ems a a m le el. A
case wi h maize on amily a ms o Sou h Eas Asia.
Ag icul u al Sys ems, 212. h ps://doi.o g/10.1016/j.
agsy.2023.103777
Lebacq, T., Ba e , P. ., and S ilman , D. (2013). Sus ain-
abili y indica o s o li es ock a ming. A e iew. In
Ag onomy o Sus ainable De elopmen (Vol. 33,
Issue 2, pp. 311–327). h ps://doi.o g/10.1007/s13593-
012-0121-x
Lomba do, V. (2021). Social inclusion h ough social s a-
us and he eme gence o de elopmen aps. Me o-
economica, 72(4), 798–825. h ps://doi.o g/10.1111/
meca.12348
Lough ey, J., O’Donoghue, C., and Conneely, R. (2022).
Al e na i e measu es o Family Fa m Viabil-
i y – Inco po a ing gap measu es. Jou nal o Ru al
S udies, 89: 257–274. h ps://doi.o g/10.1016/j.
j u s ud.2021.11.022
Mageli, I., Mannbe g, A., and Heen, E. E. (2022). Wi h
whom, and abou wha , do we compe e o social
s a us? E ec s o social closeness and ele ance o
e e ence g oups o posi ional conce ns. Jou nal o
Beha io al and Expe imen al Economics, 98. h ps://
doi.o g/10.1016/j.socec.2022.101867
Mao, L., Huang, Y., Zhang, X., Li, S., and Huang, X.
(2022). ARIMA model o ecas ing analysis o he
p ices o mul iple ege ables unde he impac o he
COVID-19. PLoS ONE 17(7): e0271594. h ps:// doi.
o g/10.1371/jou nal.pone.0271594
McCa l, B.A., and Sp een, T.H. (1997). Applied Ma h-
ema ical P og amming Using Algeb aic Sys ems.
Texas A&M Uni e si y, College S a ion, Texas. h -
ps://agecon2. amu.edu/people/ acul y/mcca l-b uce/
mccsp / hebook.pd
Mille , T. A., Rodewald, G. E., and Mcel oy, R. G. (1981).
Economies o Size in U.S. Field C op Fa ming.
Ag icul u al Economic Repo No. 472, Na ional
Economics Di ision, Economics and S a is ics Se -
ice, U.S. Depa men o Ag icul u e. h ps://doi.
o g/10.22004/ag.econ.307907
Mi enzwei, K., and B i z, W. (2018). Analysing Fa m-
speci ic Paymen s o No way using he Ag ispace
Model. Jou nal o Ag icul u al Economics, 69(3):
777-793. h ps://doi.o g/10.1111/1477-9552.12268
Mode n Wa Ins i u e (2022). Wha Comes Nex in
Uk aine: Th ee Scena ios. Mode n Wa Ins i u e –
Wes Poin . h ps://mwi.usma.edu/wha -comes-nex -
in-uk aine- h ee-scena ios/
Na ayana, N. S. S., and Pa ikh, K. S. (1981). Es ima ion o
a m supply esponse and ac eage alloca ion: a case
s udy o Indian ag icul u e. Resea ch Repo , In e -
na ional Ins i u e o Applied Sys ems Analysis, 81–1.
Ne lo e, M. and D.A. Bessle (2001), “Expec a ions,
In o ma ion and Dynamics” in B. Ga dne and G.
Rausse (Eds.) (2001), Handbook o Ag icul u al
Economics, Vol.1 (Else ie Science BV), pp. 155-206.
Neuen eld , S., Goch , A., Heckelei, T. and Ciaian, P.
(2019). Explaining a m s uc u al change in he Eu o-
pean ag icul u e: A no el analy ical amewo k. Eu o-
pean Re iew o Ag icul u al Economics 46: 713–768.
Nguyen, D. J., and He on, A. (2021). Keeping Up Wi h he
Joneses o Feeling P iced Ou ?: Explo ing How Low-
Income S uden s’ Financial Posi ion Shapes Sense o
Belonging. Jou nal o Di e si y in Highe Educa ion,
14(3): 429-440. h ps://doi.o g/10.1037/dhe0000191
O e mann, F., and Ma ga ian, A. (2014). Modelling S uc-
u al Change in Ex-An e Policy Impac Analysis. In:
Zopounidis, C., Kaloge as, N., Ma as, K., an Dijk, G.,
Baou akis, G. (eds) Ag icul u al Coope a i e Manage-
men and Policy. Coope a i e Managemen . Sp inge ,
Cham. h ps://doi.o g/10.1007/978-3-319-06635-6_8
Oli ei a, R. A., O ’conno , C. W., and Smi h, G. W.
(1979). Sho -Run Fo ecas ing Models o Bee P ices.
Wes e n Jou nal o Ag icul u al Economics, 4(1).
Pain e , M. J. (2005). Re u ns o Fa mland and Fa m
Labou and Managemen in Wes e n Canada. Jou -
nal o Fa m Managemen , 12(3): 123-141. h ps://
ageconsea ch.umn.edu/nanna/ eco d/122234/ iles/
Pain e %201.pd ?wi hWa e ma k=0&wi hMe ada a=
0& e sion=1®is e Download=1
Pa doe, I. (n.d.). Cou se no es o  STAT 501: Reg es-
sion Me hods: T.2.5.1 - ARIMA Models. Depa men
o S a is ics, Pennsyl ania S a e Uni e si y. h ps://
online.s a .psu.edu/s a 501/lesson/ / .2/ .2.5/ .2.5.1-
a ima-models
Pa oissien, E., La u e, L., and Pie , L. (2021). Ea ly exi
om business, pe o mance and neighbou s’ in lu-
ence: a s udy o a me s in F ance. Eu opean Re iew
o Ag icul u al Economics, 48(5): 1132–1161. h ps://
doi.o g/10.1093/e ae/jbab008
Pe sakos, A., Ha eau, G., P adel, W., and Sua ez, V.
(2016). Fo ecas ing Po a o and Swee po a o Yields o
2050. In e na ional Po a o Cen e (CIP) Lima, Pe u.
Wo king Pape 2016-1. 26 p.
Plogmann, J., Mußho , O., Odening, M., and Ri e , M.
(2022). Fa m g ow h and land concen a ion. Land
Use Policy, 115: 106036. h ps://doi.o g/10.1016/j.lan-
dusepol.2022.106036
Qua ey-Papa io, T. K., Ja ed, S. A., and Liu, S. (2021).
Fo ecas ing cocoa p oduc ion o six majo p oduce s
h ough ARIMA and g ey models. G ey Sys ems, 11(3):
434-462. h ps://doi.o g/10.1108/GS-04-2020-0050
374
Bio-based and Applied Economics 13(4): 353-386, 2024 | e-ISSN 2280-6172 | DOI: 10.36253/bae-14790
S ama is Man zia is e al.
RIRDC (2007). D i e s o S uc u al Change in Aus-
alian Ag icul u e. Ru al Indus ies Resea ch and
De elopmen Co po a ion (RIRDC). Publica ion No.
07/057.
Robe , M., Be gez, J. E., and Thomas, A. (2018). A s o-
chas ic dynamic p og amming app oach o ana-
lyze adap a ion o clima e change – Applica ion o
g oundwa e i iga ion in India. Eu opean Jou nal o
Ope a ional Resea ch, 265(3): 1033–1045. h ps://doi.
o g/10.1016/j.ejo .2017.08.029
Robe , M., Thomas, A., and Be gez, J. E. (2016). P o-
cesses o adap a ion in a m decision-making models.
A e iew. Ag onomy o Sus ainable De elopmen ,
36(4): 64. h ps://doi.o g/10.1007/s13593-016-0402-x
Rowan, T. S. C., Maie , H. R., Conno , J., and Dandy, G.
C. (2011). An in eg a ed dynamic modeling ame-
wo k o in es iga ing he impac o clima e change
and a iabili y on i iga ed ag icul u e. Wa e
Resou ces Resea ch, 47(7): W0752. h ps://doi.
o g/10.1029/2010WR010195
Sain -Cy , L. D. F., S o m, H., Heckelei, T., and Pie , L.
(2019). He e ogeneous impac s o neighbou ing a m
size on he decision o exi : E idence om B i any.
Eu opean Re iew o Ag icul u al Economics, 46(2):
237–266. h ps://doi.o g/10.1093/e ae/jby029
Schuh, B. e al. (2022) Resea ch o AGRI Commi ee –
The Fu u e o he Eu opean Fa ming Model: Socio-
economic and e i o ial implica ions o he decline
in he numbe o a ms and a me s in he EU, Eu o-
pean Pa liamen , Policy Depa men o S uc u al
and Cohesion Policies, B ussels. h ps://www.eu o-
pa l.eu opa.eu/RegDa a/e udes/STUD/2022/699620/
IPOL_STU(2022)699620(SUM01)_EN.pd
Seidel, C., and B i z, W. (2019). Es ima ing a dual alue
unc ion as a me a-model o a de ailed dynam-
ic ma hema ical p og amming model. Bio-Based
and Applied Economics, 8(1): 75-99. h ps://doi.
o g/10.13128/bae-8147
Sheng, Y., Jackson, T., and Gooday, P. (2015). Resou ce
ealloca ion and i s con ibu ion o p oduc i i y
g ow h in Aus alian b oadac e ag icul u e. Aus al-
ian Jou nal o Ag icul u al and Resou ce Economics,
61(1): 56-75. h ps://doi.o g/10.1111/1467-8489.12137
Siegle, J., As ill, G., Plakias, Z., and T egeagle, D. (2024).
Es ima ing pe ennial c op supply esponse: A me h-
odology li e a u e e iew. Ag icul u al Econom-
ics (Uni ed Kingdom). h ps://doi.o g/10.1111/
agec.12812
Smale M., Saupe W.E., Salan P. (1986). Fa m amily cha -
ac e is ics and he iabili y o a m households in
Wisconsin, Mississippi, and Tennessee. Ag icul u al
Economics Resea ch, 38: 11–27.
Sun, R., Nolan, J., and Kulsh esh ha, S. (2022). Agen -
based modeling o policy induced ag i-en i onmen-
al echnology adop ion. SN Business & Economics,
2(8). h ps://doi.o g/10.1007/s43546-022-00275-6
T oos , C., and Be ge , T. (2015). Dealing wi h unce ain-
y in agen -based simula ion: Fa m-le el modeling o
adap a ion o clima e change in sou hwes Ge many.
Ame ican Jou nal o Ag icul u al Economics, 97(3),
833–854. h ps://doi.o g/10.1093/ajae/aau076
T oos , C., and Be ge , T. (2016). Simula ing s uc u al
change in ag icul u e: Modelling a ming households
and a m succession. In: Sau age, S., Sánchez-Pé ez,
J.M., Rizzoli, A.E. (Eds.), 2016. P oceedings o he
8 h In e na ional Cong ess on En i onmen al Mod-
elling and So wa e, July 10-14, Toulouse, F ance.
ISBN: 978-88-9035-745-9.
T oos , C., and Be ge , T. (2020). Fo malising alida ion?
Towa ds c i e ia o alid conclusions om agen -
based simula ion. 10 h In e na ional Cong ess on
En i onmen al Modelling and So wa e B ussels, Bel-
gium. h ps://schola sa chi e.byu.edu/iemsscon e -
ence/2020/
Viaggi, D., Raggi, M., and Gomez Y Paloma, S. (2010).
An in ege p og amming dynamic a m-household
model o e alua e he impac o ag icul u al policy
e o ms on a m in es men beha iou . Eu opean
Jou nal o Ope a ional Resea ch, 207(2): 1130–1139.
h ps://doi.o g/10.1016/j.ejo .2010.05.012
Viaggi, D., Ba olini, F., Raggi, M., Sa donini, L., Sam-
me h, F., Gomez, Y., Paloma, S. (2011). Fa m in es -
men beha iou unde he CAP e o m P ocess. JRC
Eu opean Comission: Luxembou g.
Weye s ass, K. (2016). A ool o suppo ing economic
policy-making in he o me Yugosla ia. Documen a-
ion and applica ions o a mac oeconomic mul i-coun-
y model. Os eu opa: Geschich e, Wi scha , Poli ik,
49. Wien: LIT-Ve lag. 224 p.
Wibo g, T. (1998). KRAM - A Sec o Model o Danish
Ag icul u e: Backg ound and F amewo k De elopmen .
h p://lib.d .ias a e.edu/ca d_wo kingpape s/221
Zimme mann, A., and Heckelei, T. (2012). S uc u -
al Change o Eu opean Dai y Fa ms - A C oss-
Regional Analysis. Jou nal o Ag icul u al Econom-
ics, 63(3): 576-603. h ps://doi.o g/10.1111/j.1477-
9552.2012.00355.x
Zimme mann, A., Heckelei, T., and Domínguez, I. P.
(2009). Modelling a m s uc u al change o in e-
g a ed ex-an e assessmen : e iew o me hods and
de e minan s. En i onmen al Science and Poli-
cy, 12(5): 601-618. h ps://doi.o g/10.1016/j.en s-
ci.2009.01.014
375
Simula ing a m s uc u al change dynamics in Thessaly (G eece) using a ecu si e p og amming model
Bio-based and Applied Economics 13(4): 353-386, 2024 | e-ISSN 2280-6172 | DOI: 10.36253/bae-14790
APPENDIX
Table A1. Simula ed a e age g oss ma gin o each c opping ac i -
i y (EUR/ha).
2012 2019
Co on 1,176 1,549
Tobacco (Vi ginia) 4,750 4,757
Maize 2,300 1,409
P ocessing Toma o 6,370 4,863
P ocessing Peppe 17,331 27,800
Al al a (hay) 807.3 817.6
Al al a (seed p oduc ion) - 509.5
Du um Whea 258.2 207.3
Sou ce: Au ho s, based on sample da a.
SUPPLEMENTARY MATERIAL TO
“SIMULATING FARM STRUCTURAL CHANGE
DYNAMICS IN THESSALY (GREECE) USING
A RECURSIVE PROGRAMMING MODEL”
Pa A: Concep ual amewo k o ARIMA modeling
The Box-Jenkins me hod o Au o eg essi e In eg a ed
Mo ing A e age (ARIMA) models is conside ed one o
he mos e icien ime se ies o ecas ing me hods u iliz-
ing almos any se o da a (Ch is odoulos e al., 2010). In
his amewo k, o he au ho s conside ha ARIMA mod-
els ha e been ema kably success ul wi h an excellen pe -
o mance on small da a se s (Ga nie , n.d.). Acco ding o
a ious modele s, ARIMA models can p o ide accep able
esul s when a leas 16- ime se ies da a poin s a e a ail-
able (Go a di & Sca so, 1994; Ch is odoulos e al., 2010).
An impo an class o s ochas ic models o desc ib-
ing ime se ies a e called s a iona y models o Au o e-
g essi e-Mo ing A e age (ARMA) models a ying abou
a ixed cons an mean le el and wi h cons an a iance
(Box e al., 2016).
An ARMA (p,q) model is o mula ed as ollows:
Y = φiY -i + ε – θjε -j, (A1)
whe e φ1 .…, φp a e he au o eg essi e (AR) pa am-
e e s o be es ima ed, θ1 ,…,θq a e he mo ing a e age
(MA) pa ame e s o be es ima ed, and ε1…ε a e a se ies
o unknown andom “shocks” (o esiduals) ha a e
assumed o ollow a no mal dis ibu ion (Pa doe, n.d.).
The model can be simpli ied by in oducing he
Box-Jenkins backwa d shi ope a o 21 whe e BiY = Y -i
21 The Backwa d shi ope a o is a use ul no a ional de ice exp essing
and Bjε = ε -j; Y1,…,Y is any ime se ies ; p< and q<
(Pa doe, n.d.).
Subs i u ing backwa d shi ope a o s in equa ion
(A1), we ob ain he ollowing o m:
(1 – φiBi)Y = (1 – θjBj)ε (A2)
Which is o en educed u he o (Pa doe, n.d.):
φp(B)Y = θq(B)ε (A3)
Many se ies encoun e ed in indus y o business
e eal nons a iona y beha io 22and do no a y abou a
ixed mean, showing a s ochas ic end (Box e al., 2016).
We should he e o e con e a non-s a iona y ime se ies
o a s a iona y one by di e encing he ARMA (p,q) model.
Then he ARMA (p,q) model can be ex ended and
w i en using di e ences ΔY = (1 – Β)dY = ∇dY as ol-
lows:
φp(B)(1 – Β)dY = θq(B)ε (A4)
whe e d is he o de o di e encing. Replacing in he
ARMA model wi h he di e ences abo e, we ob ain he
o mal ARIMA p,d,q) model (Pa doe, n.d.).
To de ec non-s a iona i ies, we u ilize one o he
mos well-known es s, which co esponds o he aug-
men ed Dickey-Fulle (ADF) es (As e iou & Hall, 2007;
Mahan e al., 2015; Box e al., 2016). The iden i ica ion
o possible model o de s (p,q) is app oached h ough he
u iliza ion o Au oco ela ion unc ion (ACF) and Pa -
ial Au oco ela ion unc ion (PACF) plo s (Mahan e
al., 2015; Box e al., 2016; Ga nie , n.d.) while ying o
keep he model o de s a low le els (≤ 2) o mos o he
es ima ed models (Go a di & Sca so, 1994). A e es i-
ma ing se e al models, we es whe he he condi ion o
in e ibili y (As e iou & Hall, 2007; Ga nie , n.d.) and
s a is ical signi icance o he AR and MA pa s o he
model a e sa is ied (Mossad & Alazba, 2015). The es i-
ma ed models a e hen compa ed acco ding o he Akai-
ke in o ma ion c i e ion (AIC) by selec ing he model
wi h he lowes alue (Mahan e al., 2015; Box e al.,
2016; Ga nie , n.d.).
The diagnos ic check o he model is hen pe o med,
which is applied o esiduals o de ec whe he hey exhibi
he leng h o p e ious da a he model uses o p o ide o ecas s
(Ch is odoulos e al., 2010).
22 ARIMA modeling equi es ha he ime se ies be s a iona y (Scha e
e al., 2021). A s a iona y se ies is cha ac e ized by h ee p ope ies: a
cons an mean, cons an a iance, and cons an co a iance ha depends
only on he ime in e als (Scha e e al., 2021). Time se ies wi h ends
o changing a iance is non-s a iona y.
376
Bio-based and Applied Economics 13(4): 353-386, 2024 | e-ISSN 2280-6172 | DOI: 10.36253/bae-14790
S ama is Man zia is e al.
au oco ela ion, u ilizing he B eusch-God ey Lag ange
Mul iplie (LM) es (Mahan e al., 2015; Weye s ass,
2016; Ayele e al., 2017). The null hypo hesis o he LM es
is ha he e is no au oco ela ion in he esiduals se ies
up o he p e-de e mined lag o de (p=2 in ou analysis)
a he 5% le el o signi icance (Weye s ass, 2016; Ayele
e al., 2017). Rega ding he measu emen o he o ecas -
ing accu acy o ARIMA models, he e is no uni e sally
p e e ed measu e; howe e , acco ding o a ious model-
e s (Go a di and Sca so, 1994; Ch is odoulos e al., 2010),
pa icula emphasis is gi en o he measu e o Mean Abso-
lu e Pe cen age E o (MAPE). A his poin , we would like
o poin ou ha he e is no commonly accep ed h eshold
o MAPE in he in e na ional li e a u e; howe e , some
au ho s conside ha a o ecas ing model is cha ac e -
ized by good o ecas ing accu acy (o goodness-o - i )
when MAPE does no exceed 20%, whe eas when i does
no exceed 10%, he o ecas ing accu acy is cha ac e ized
as high o pe ec (e.g., Qua ey-Papa io e al., 2021). Es i-
ma es and s a is ical es s o ARIMA models we e pe -
o med using EViews s a is ical package.
Pa A: Re e ences
As e iou, D., & Hall, S. (2007). Applied Econome ics: A
Mode n App oach. Palg a e Macmillan, New Yo k.
Ayele, A. W., Gab eyohannes, E., & Tes ay, Y. Y.
(2017). Mac oeconomic De e minan s o Vola il-
i y o he Gold P ice in E hiopia: The Applica-
ion o GARCH and EWMA Vola ili y Models.
Global Business Re iew, 18(2), 308–326. h ps://doi.
o g/10.1177/0972150916668601
Box, G., Jenkins, G., Reinsel, G., & Ljung G. (2016). Time
Se ies Analysis: Fo ecas ing and Con ol, 5 h Edi ion,
by Geo ge E. P. Box, Gwilym M. Jenkins, G ego y
C.Reinsel and G e a M. Ljung, 2015. Published by
John Wiley and Sons Inc.
Ch is odoulos, C., Michalakelis, C., & Va ou as, D.
(2010). Fo ecas ing wi h limi ed da a: Combining
ARIMA and di usion models. Technological Fo ecas -
ing and Social Change, 77(4), 558–565. h ps://doi.
o g/10.1016/j. ech o e.2010.01.009
Ga nie , H. (n.d.). In oduc ion o ime se ies analysis and
o ecas ing. Uni e si y o Lo aine. h ps://w3.c an.
uni -lo aine. /pe so/hugues.ga nie /Enseignemen /
TSAF/C-TSAF-Box-Jenkins_me hod.pd
Go a di, G., & Sca so, E. (1994). Di usion models in
o ecas ing: A compa ison wi h he Box-Jenkins
app oach. Eu opean Jou nal o Ope a ional Resea ch,
75(3). h ps://doi.o g/10.1016/0377-2217(94)90300-X
Mahan, M., Cho n, C., & Geo gopoulos, A. (2015). Whi e
Noise Tes : de ec ing au oco ela ion and nons a ion-
a i ies in long ime se ies a e ARIMA modeling. P o-
ceedings o he 14 h Py hon in Science Con e ence,
97–104. h ps://doi.o g/10.25080/majo a-7b98e3ed-00
Mossad, A., & Alazba, A. A. (2015). D ough o ecas -
ing using s ochas ic models in a hype -a id clima e.
A mosphe e, 6(4), 410–430. h ps://doi.o g/10.3390/
a mos6040410
Pa doe, I. (n.d.). Cou se no es o  STAT 501: Reg es-
sion Me hods: T.2.5.1 - ARIMA Models. Depa men
o S a is ics, Pennsyl ania S a e Uni e si y. h ps://
online.s a .psu.edu/s a 501/lesson/ / .2/ .2.5/ .2.5.1-
a ima-models
Qua ey-Papa io, T. K., Ja ed, S. A., & Liu, S. (2021).
Fo ecas ing cocoa p oduc ion o six majo p oduc-
e s h ough ARIMA and g ey models. G ey Sys ems,
11(3), 434-462. h ps://doi.o g/10.1108/GS-04-2020-
0050
Scha e , A.L., Dobbins, T.A. & Pea son, SA. (2021).
In e up ed ime se ies analysis using au o eg essi e
in eg a ed mo ing a e age (ARIMA) models: a guide
o e alua ing la ge-scale heal h in e en ions. BMC
Medical Resea ch Me hodology, 58(21), h ps://doi.
o g/10.1186/s12874-021-01235-8
Weye s ass, K. (2016). A ool o suppo ing economic
policy-making in he o me Yugosla ia. Documen a-
ion and applica ions o a mac oeconomic mul i-coun-
y model. Os eu opa: Geschich e, Wi scha , Poli ik,
49. Wien: LIT-Ve lag. 224 p.
Pa B: S uc u e o he model’s objec i e unc ion and con-
s ain s
The ollowing desc ibes he objec i e unc ion’s
s uc u e and he cons ain s ypical o each sub-model.
The objec i e unc ion o he expec ed g oss p o i o he
a m in yea is de ined as ollows:
Max E{Π , } = XT ,j, [E{p ,j, } E{y ,j, } – c ,j,
+ lsj, + e , ecop1 ,j, + ε , ecop2 ,j, ] + DP , DL , +
b , NP , NL , + β , OP , OL , + , RP , AL ,
(B1)
Subjec o:
A able land cons ain
X ,j, = AL , , o = 1,…,T, j ∈ J (B2)
I iga ed land cons ain
X ,wj, ≤ IL , , o = 1,…,T, wj ∈ WJ, WJ ⊆ J (B3)
Ci cula ing capi al cons ain
X ,j, c ,j, ≤ CRC , , o = 1,…,T, j ∈ J (B4)
X ,j, ≥ 0 , o = 1,…,T (B5)
383
Simula ing a m s uc u al change dynamics in Thessaly (G eece) using a ecu si e p og amming model
Bio-based and Applied Economics 13(4): 353-386, 2024 | e-ISSN 2280-6172 | DOI: 10.36253/bae-14790
Pa F: ARIMA and linea end models es ima ions
Table F1. ARIMA models o exogenously de e mined pa ame e s o in e es
Εxogenously de e mined
pa ame e o a m model (Y )
Time se ies da a
poin s
[pe iod]
ARIMA
Model
(p,d,q)
Φ1Φ2Φ3Θ1Θ2Θ3μMAPE
(%) AIC
Augmen ed
Dickey-Fulle
-S a is ic
B eusch-God ey Se ial
Co ela ion LM Tes
[P ob. X2 (p)]
Hi ed labo p ice index 19
[2001-2018 ] (2,0,1) 1.64***
(0.05)
-0.82***
(0.04) --0.99***
(0.09) - - 92.79***
(0.59) 0.95 3.84 -3.55*P ob. X2 (2)=0.059
Inpu p ice index 20
[2000-2019 ] (1,1,1) 0.87***
(0.06) - - -0.99***
(0.12) - - - 3.78 6.19 -4.29** P ob. X2 (2)=0.44
Machine y en al p ice index 20
[2000-2019 ] (0,2,1) - - - -0.50**
(0.21) - - - 1.29 4.09 -7.00*** P ob. X2 (2)=0.92
Land en al p ice index 19
[2000-2018 ] (1,0,1) 0.53**
(0.20) - - 0.99***
(0.06) - - 99.05***
(1.57) 1.05 3.76 -4.01** P ob. X2 (2)=0.40
In e es a e index 21
[2000-2020 ] (0,2,1) - - - -0.91***
(0.07) - - -5.51 6.49 -6.92*** P ob. X2 (2)=0.99
Co on yield
(Kg/Ha)
57
[1961-2017 ] (1,0,1) 0.92***
(0.01) - - -0.97***
(0.03) - - 299.23***
(5.36) 7.55 9.21 -4.21*** P ob.X2 (2)=0.54
D. whea yield
(Kg/ 0.1 Ha)
57
[1961-2017 ] (2,0,1) 0.50***
(0.15)
0.43***
(0.14) --0.62***
(0.14) - - 278.04***
(51.51) 11.30 9.83 -3.47** P ob.X2 (2)=0.98
Tobacco yield (Vi ginia)
(Kg/0.1 Ha)
39
[1979-2017 ] (1,0,3) 0.87***
(0.02) - - -1.00***
(0.15)
0.48**
(0.21)
-0.46***
(0.15)
336.52***
(8.17) 6.73 9.40 -4.21*** P ob.X2 (2)=0.35
Peppe yield
(Kg/0.1 Ha)
47
[1961-2007 ] (1,0,2) 0.98***
(0.09) - - -0.66***
(0.14)
-0.29**
(0.14) -4396.72***
(1079.63) 6.28 13.21 -4.10*** P ob.X2 (2)=0.81
Toma o yield
(Kg/0.1 Ha)
47
[1961-2007 ] (1,1,0) -0.44***
(0.14) -----81.27**
(31.97) 5.5 14.34 -7.21*** P ob.X2 (2)=0.36
Legumes c ops yield
[Al al a (hay & seed)]
(Kg/0.1 Ha)
18
[2000-2017 ] (0,0,2) - - - 0.31***
(0.08)
0.93***
(0.02) -740.24***
(26.53) 4.69 10.83 -4.82*** P ob.X2 (2)=0.10
Maize yield
(Kg/0.1 Ha)
37
[1981-2017 ] (2,0,0) 0.54***
(0.16)
0.33***
(0.16) ----1086.85***
(122.19) 3.77 10.67 -4.18** P ob.X2 (2)=0.36
Co on p ice
(EUR/kg)
20
[2000-2019 ] (1,0,1) 0.85***
(0.06) - - -0.96***
(0.04) - -
0.48***
(0.048) 15.17 -2.19 -3.39*P ob.X2 (2)=0.16
D. whea p ice
(EUR/kg)
20
[2000-2019 ] (3,0,0) 0.84***
(0.23)
-0.62**
(0.29)
0.44*
(0.22) ---0.19***
(0.02) 10.81 -4.08 -3.19*P ob.X2 (2)=0.32
Legume c ops p ice
(Al al a-hay)
(EUR/kg)
20
[2000-2019 ] (1,1,0) -0.43*
(0.22) ---- - - 6.00 -6.36 -4.36** P ob.X2 (2)=0.80
Maize p ice
(EUR/kg)
20
[2000-2019 ] (1,0,1) 0.83***
(0.06) - - -0.99***
(0.10) - -
0.18***
(0.00) 8.48 -4.70 -3.40* P ob.X2 (2)=0.08
(Con inued)

384
Bio-based and Applied Economics 13(4): 353-386, 2024 | e-ISSN 2280-6172 | DOI: 10.36253/bae-14790
S ama is Man zia is e al.
Εxogenously de e mined
pa ame e o a m model (Y )
Time se ies da a
poin s
[pe iod]
ARIMA
Model
(p,d,q)
Φ1Φ2Φ3Θ1Θ2Θ3μMAPE
(%) AIC
Augmen ed
Dickey-Fulle
-S a is ic
B eusch-God ey Se ial
Co ela ion LM Tes
[P ob. X2 (p)]
To al a able land index 16
[2004-2019 ] (0,1,3) - - - -1.26***
(0.28)
1.17***
(0.22)
-0.82***
(0.14)
3.19***
(0.89) 3.67 6.98 -7.75*** P ob.X2 (2)=0.64
To al ci cula ing capi al index 16
[2004-2019 ] (0,1,1) - - - -0.93***
(0.06) - - 29. 7***
(2.31) 10.91 10.24 -3.75** P ob.X2 (2)=0.19
No es: ∇dY = μ + ϕ1∇dY -1 + ⋯ ϕp∇dY -p + ε − θ1ε -1 − ⋯ − θqε -q;
Φ1, . . . , Φp: au o eg essi e (AR) model pa ame e s o o de p; Θ1, … , Θq: mo ing a e age (MA) model pa ame e s o o de q (Ma ínez-Acos a e al., 2020) ; ε is whi e noise; μ= a
cons an equal o he mean o he se ies i d = 0 (Na ayana & Pa ikh, 1981); * indica es signi icance a 0.1 le el, ** indica es signi icance a 0.05 le el, *** indica es signi icance a 0.01
le el; The null hypo hesis H0 o he B eusch-God ey Se ial Co ela ion LM Tes is ha he e is no au oco ela ion in he esiduals se ies up o p e-de e mined lag o de (p=2 in ou
analysis) a he 0.05 le el o signi icance (Weye s ass, 2016).
Sou ce: Au ho s, based on ELSTAT (2019b), ELSTAT (2019c), FADN Public Da abase, G eek Minis y o Ru al De elopmen and Food, G eek Minis y o Ru al De elopmen and
Food (2019).
Table F2. Linea end model eg ession s a is ics o u al households’ li ing expendi u e index (LEI)
Va iable Coe icien S d. E o -S a is ic P ob.
C 0.954777 0.019644 48.60363 0.0000
@TREND -0.023925 0.002778 -8.612057 0.0000
R-squa ed 0.870843 Mean dependen a 0.811226
Adjus ed R-squa ed 0.859101 S.D. dependen a 0.099846
S.E. o eg ession 0.037479 Akaike in o c i e ion -3.589453
Sum squa ed esid 0.015451 Schwa z c i e ion -3.502538
Log likelihood 25.33145 Hannan-Quinn c i e . -3.607318
F-s a is ic 74.16753 Du bin-Wa son s a 0.642814
P ob(F-s a is ic) 0.000003
Sou ce: Au ho s, based on ELSTAT (2021) da a.
Table F1. (Con inued).
385
P edic ing he e ec o he Common Ag icul u al Policy pos -2020 using an agen -based model based on PMP me hodology
Bio-based and Applied Economics 13(4): 353-386, 2024 | e-ISSN 2280-6172 | DOI: 10.36253/bae-14790
(a)
(b)
(c)
(d)
(e)
( )
(g)
(h)
(i)
(j)
(k)
(l)
(m)
(n)
(o)
(p)
(q)
( )
(s)
Figu e F1. ARIMA and linea end models o he exogenously de e mined pa ame e s o in e es . No es: he ho izon al axis indica es he
yea ; 0.1 Ha (hec a e) =1 s emma is he G eek uni o land a ea. (a) Hi ed labo p ice index; (b) Inpu p ice index; (c) Machine y en al
p ice index; (d) Land en al p ice index; (e) In e es a e index; ( ) Co on yield (kg/0.1 Ha); (g) Du um whea yield (kg/0.1 Ha); (h) Tobac-
co yield (kg/0.1 Ha); (i) Peppe yield (kg/0.1 Ha); (j) Toma o yield (kg/0.1 Ha); (k) Legume c ops yield (kg/0.1 Ha) including Al al a (hay
& seed); (l) Maize yield (kg/0.1 Ha); (m) Co on p ice (EUR/kg); (n) Du um whea p ice (EUR/kg); (o) Al al a (hay) p ice (EUR/kg); (p)
Maize p ice (EUR/kg); (q) To al a able land index; ( ) To al ci cula ing capi al index; (s) Li ing expendi u es index. Sou ce: Au ho s, based
on ELSTAT (2019b), ELSTAT (2019c), ELSTAT (2021), FADN Public Da abase, G eek Minis y o Ru al De elopmen and Food, G eek
Minis y o Ru al De elopmen and Food (2019).
386
Bio-based and Applied Economics 13(4): 353-386, 2024 | e-ISSN 2280-6172 | DOI: 10.36253/bae-14790
S ama is Man zia is e al.
Pa F: Re e ences
ELSTAT (2019b). Cos s o he ac o s o Ag icul u al and
Li es ock P oduc ion (Cos s indices), 2019. Hellenic
S a is ical Au ho i y. h ps://www.s a is ics.g /en/s a-
is ics/-/publica ion/DKT33/2019
ELSTAT (2019c). Inpu and Ou pu P ice Indices in Ag i-
cul u al and Li es ock P oduc ion, 2019. Hellenic S a-
is ical Au ho i y. h ps://www.s a is ics.g /en/s a is-
ics/-/publica ion/DKT30/2019-M12
ELSTAT (2021). Household Budge Su ey (a e 2008),
2020. Hellenic S a is ical Au ho i y.h ps://www.s a-
is ics.g /en/s a is ics/-/publica ion/SFA05/2020
Ma ínez-Acos a, L., Med ano-Ba boza, J. P., López-
Ramos, Á., López, J. F. R., & López-Lamb año, Á.
A. (2020). SARIMA app oach o gene a ing syn-
he ic mon hly ain all in he Sinú i e wa e -
shed in Colombia. A mosphe e, 11(6). h ps://doi.
o g/10.3390/a mos11060602
FADN Public Da abase. Fa m Accoun ancy Da a Ne -
wo k Public Da abase. Di ec o a e-Gene al o Ag i-
cul u e and Ru al De elopmen . h ps://ag ida a.
ec.eu opa.eu/ex ensions/FADNPublicDa abase/FAD-
NPublicDa abase.h ml
G eek Minis y o Ru al De elopmen and Food. S a is i-
cal Da a- ime se ies (in G eek). h p://wwww.minag-
ic.g /g eek/ag o_pol/3.h m
G eek Minis y o Ru al De elopmen and Food (2019).
S a is ics Da a o a eas and p oduc ion o plan p od-
uc s (in G eek). h p://www.minag ic.g /index.php/el/
he-minis y-2/s a is ikes- ekmh ioshs/8510-s a is i-
ka-ek -pa ag- y ikonp oion on
Na ayana, N. S. S., & Pa ikh, K. S. (1981). Es ima ion o
a m supply esponse and ac eage alloca ion: a case
s udy o Indian ag icul u e. Resea ch Repo , In e na-
ional Ins i u e o Applied Sys ems Analysis, 81–1.
Weye s ass, K. (2016). A ool o suppo ing economic
policy-making in he o me Yugosla ia. Documen a-
ion and applica ions o a mac oeconomic mul i-coun-
y model. Os eu opa: Geschich e, Wi scha , Poli ik,
49. Wien: LIT-Ve lag. 224 p.