| Abstract | Republika Hrvatska je od 1. srpnja 2013. godine punopravna članica Europske unije (EU).
Poljoprivreda je strateška grana hrvatskog gospodarstva u kojoj su se pristupanjem dogodile
znatne promjene i prilagodbe. Gospodarske koristi otvorenog tržišta prisutne su već i prije
same integracije države u EU. Kao potpora donošenju odluka i strateškom planiranju u
praksi se koriste rezultati evaluacijskih i analitičkih alata. Jedan od alata za procjenu učinaka
koji na temelju kvalitativnih i kvalitativnih pokazatelja analizira promjene na tržištu
poljoprivrednih proizvoda je model parcijalne ravnoteže. Prema dostupnoj literaturi, slične
analize - konkretno za domicilno tržište poljoprivrednih proizvoda, nisu zastupljene u
dovoljnoj mjeri. Integracijom domaćeg poljoprivrednog tržišta zajedničkom tržištu EU na
razini domaće proizvodnje i potrošnje događaju se promjene. Neki od čimbenika koji oblikuju
domaću ponudu i potražnju su cijene na tržištu EU, razina samodostatnosti u EU, agrarno
političke mjere, razvoj tehnologije i sl.
Cilj rada je razviti i primijeniti ekonometrijski model parcijalne ravnoteže u ratarskom sektoru
hrvatske poljoprivrede u kontekstu Zajedničke poljoprivredne politike (ZPP) tj. procijeniti
učinak agrarno političkih mjera u sektoru. Rezultati modela projiciraju razvoj tržišnih
pokazatelja izabranih ratarskih proizvoda u uvjetima postojeće ZPP-a, uz uvjet ceteris
paribus. Budući je primijenjeni hrvatski model integrirani podmodel EU modela, u radu su
prikazani i rezultati projekcija za tržišta ratarskih proizvoda u EU i u državama članicama koje
„diktiraju“ tokove na tržištima analiziranih proizvoda u EU.
Pristup modeliranju preuzet je iz postojeće metodologije ustanovljene AGMEMOD
(Agriculture Member State Modelling) partnerstvom. Sustav ekonometrijski procijenjenih
bihevioralnih jednadžbi, potom kalibriranih parametara jednadžbi, predstavlja tržišta pojedinih
ratarskih kultura (pšenica, ječam, kukuruz, zob, raž, uljana repica, suncokret, soja). Tim
jednadžbama stvara se međuveza dva sustava (EU i Hrvatska) kojom se izračunavaju
trgovinski pokazatelji (neto trgovina). Domaće cijene proizvoda su vezane na EU cijene
jednadžbama koje opisuju utjecaj EU na domaće cijene. Cijene, uz uključene elemente
politike u model, određuju razine proizvodnje i potrošnje svih analiziranih proizvoda. Tzv.
politički harmoniziran pristup primijenjen u modelu omogućuje strukturiranu prezentaciju
mjera ZPP-a te se njime prikazuju učinci mjera na proizvodnju. Međudjelovanje pojedinih
tržišta uključenih u model i određena biološka ograničenja uzeti su u obzir i neposredno su
uključeni u model. Nakon ekonometrijske procjene parametara bihevioralnih jednadžbi i
rješavanja modela, multiplikatori učinaka pokazali su da se struktura modela i parametri
ponašaju kako je predviđeno. Cjelokupni model osjetljiv je na egzogene šokove koji se a
priori očekuju, vodeći računa o biološkim i ekonomskim ograničenjima.
Analiza rezultata započinje generiranjem tzv. Polaznog scenarija (Baseline-a) za Hrvatsku.
Polazni scenarij pretpostavlja agrarnu politiku iz 2015. godine (koja u odnosu na politiku prije
pristupa nudi veći proračun poljoprivredi) koja vrijedi za domaće tržište poljoprivrednih
proizvoda. Posljednja promatrana godina tržišnih ulaznih podataka u model je 2012., a
budući je Hrvatska već tada preuzela mjere ZPP-a s primjenom u ratarstvu (nevezana
plaćanja), u polaznom scenariju uključena je agrarna politika EU za razdoblje 2012./2014. i
2014./2020. godine.
Rezultati projekcije su pokazali da će cijene domaćih ratarskih proizvoda pratiti trend cijena u
EU i biti na nešto višoj razini u odnosu na prosjek 2010/2012 zbog troška domaće
proizvodnje. Pretpostavljeni tehnološki napredak će biti glavni pokretač povećanja prinosa, a
posljedično povećanju cijena očekuje se i povećanje proizvodnih površina jer u Hrvatskoj
postoji dovoljno „neograničenih poljoprivrednih površina“. Scenarij pokazuje pozitivan učinak
ulaska Hrvatske u EU na ratarsku proizvodnju. Drugi set rezultata u radu prikazuje trendove
ukupne ratarske proizvodnje u EU i kod pojedinih država članica, tzv. ključnih tržišta. Ta
tržišta (npr. francusko, njemačko, nizozemsko) su važna kako za EU tako i za hrvatski
model, jer su uključena u jednadžbe domaćih proizvoda i jer posredno prenose utjecaj
europskih ali i svjetskih cijena. Na strani ponude proizvoda, pretpostavke cijena i politike
djeluju podjednako važno u scenarijskoj projekciji. Potražnja svih poljoprivrednih proizvoda je
prvenstveno pod utjecajem promjena cijena, dok je učinak promjene politike indirektan. U
pozitivnom vrednovanju scenarija, gdje se testira model, uz povećanje EU cijena za 10% i
smanjenje učinka cjenovnog indeksa mineralnog gnojiva za 10% , proizvodnja svih proizvoda
raste dok je domaća potrošnja promjenjiva. Također, model reagira i u obrnutoj situaciji
(negativno vrednovanje: -10% EU cijene; +10% parametar indeksa troškova mineralnih
gnojiva). Model reagira na te egzogene šokove u skladu s ekonomskom teorijom.
Svrha ovoga rada nije samo razvoj modela koji se može koristi za analizu (politike) nego i
prikaz mogućnosti kreiranja modela u uvjetima ograničenih ili deficitarnih podataka.
Identificiranje varijabli i odnosa koji se analiziraju je središnji problem kod razvoja modela. U
procesu specifikacije, procjene, vrednovanja ili simulacije modela, često se dogodi da model,
kao što je originalno postavljen ne prikazuje korisne rezultate. Promjena, ponovna procjena
(engl. re-estimation) i/ili kalibracija pojedinih jednadžbi u modelu događa se nekoliko puta u
procesu razvoja modela sa svrhom postizanja što boljeg prikaza realnog stanja. Ograničenje
podataka i nedostatak detaljnih studija o učincima pojedinih varijabli (parametri u
jednadžbama) mogu biti identificirani kao glavna ograničenja prilikom kreiranja modela. U
ovom trenutku, jedna od prednosti trenutne verzije modela, može biti činjenica da je model
postavljen/izgrađen i da je konzistentan s ekonomskom teorijom te procijenjen od hrvatskih
stručnjaka. Ipak, u ocjeni modela, što je možda najbitnije, važno je razumljivo prikazati
rezultate modela ljudima koji imaju puno znanja o pitanjima koja se rješavaju u modelu, a
istovremeno su kritični prema metodi modeliranja, općenito. |
| Abstract (english) | Republic of Croatia is a full member of the EU from 1st of July 2013. Agriculture is a strategic
sector of the Croatian economy where were occurred the biggest changes and adjustments
by joining. Even before the country's integration to the EU there were present the economic
benefits of open market. Results of analytical tools are used in practice as support for
evaluation and decision-making and strategic planning. Model of partial equilibrium is one of
the tools for assessing the impact which on the basis of quantitative values and qualitative
indicators analyzes changes in the market of agricultural products. According to available
literature similar analysis, specifically for the domestic market of agricultural products, are
insufficiently present.
The aim is to build and validate econometric partial equilibrium model for Croatian crop
sector for projection and policy simulations. In general, AGMEMOD (Agriculture Member
State Modelling) models consist of sub-models for grains (soft wheat, durum wheat, barley,
maize, rye, and oats), oilseeds (rapeseed, sunflower, and soya), livestock (cattle, pigs,
poultry, sheep and goats) along with its products and dairy products (milk, cheese, butter,
skimmed milk powder, whole milk powder). Results of the model estimate market
development of selected agricultural crop products in terms of the existing agri-policy
measures, provided ceteris paribus. Since the Croatian model is a part of integrated submodel
of the EU model, the paper presents the estimated results for the agricultural products
market in the EU and in Member States which market trends affected the most analyzed
products markets in the EU.
Country model and his commodity sub-models are built on a common format in order to link
up with other partners’ models and such provide integrated model for the EU as a whole. On
the other side, each national model has to reflect specific problems and characteristics
typical for agricultural sectors in that country. National model joints sub-models of individual
commodities that are linked together through cross-price effects in supply and demand
equations and price transmission equations which link domestic prices with international
prices. Commodity sub-models describe acreage, animal stocks, yield levels, production,
commodity stocks, food and feed demand, imports and exports. Individual crop sector
models are linked through the allocation of land, while crop and livestock sectors are linked
through the use of feeds. Sub-models consist of numerous components which reflect various
aspects of demand, supply and price determination and such represent particular commodity
market structure in the form of linear equations. Each commodity market is in equilibrium
when sum of production, beginning stocks and imports equals to sum of domestic use,
ending stocks and exports.
The Croatian market of crops (wheat, barley, corn, oats, rye, rapeseed, sunflower, and
soybean) is described using the system of econometric estimated behavioral (linear)
equations, which are then then calibrated. Behavioral equations reflect supply and demand
side. These equations produce also the interconnection of two systems (EU and Croatia)
which addresses trade indicators (net trade). Domestic prices of products are related to the
EU price equations that describe the impact of the EU on domestic prices. Prices with
included elements of policy model, determine the level of production and consumption of the
products analyzed. The so-called policy harmonized approach in the model provides a
structured presentation of measures of the CAP and shall show the effects of the measures
on production. The interaction between the markets included in the model and certain
biological constraints are taken into account and are directly involved in the model.
After econometric estimation of parameters of behavioral equations and solving models,
multipliers effects showed that the model structure and parameters behaved as anticipated.
The entire system is vulnerable to exogenous shocks that are a priori expected, taking into
account the biological and economic constraints. Production levels answer of crop production
is more limited by the sown area size than a sudden change in domestic consumption. Net
trade is the difference between domestic production and consumption and therefore
simulated value of net trade show all the errors committed in the rest of the simulation model.
Analysis of the results begins by generating Baseline scenario for Croatia. Baseline scenario
assumes the agricultural policy from 2015 (which, in relation to the policy before access
offers a bigger budget to agriculture), which is applied to the domestic market of agricultural
products. Last observed years of market inputs to the model in 2012, and since that time
Croatia has already taken measures of the CAP with the application in crop production
(decoupled payments) The Baseline includes national and EU agricultural policy for the
period 2012-2014 and 2014-2020. Results projections showed that the prices of domestic
agricultural products follow the trend of prices in the EU and to be somewhat higher due to
the cost of domestic production. The assumed technological advances will be the main driver
of yield growth, and consequently an increase in price is expected to increase production
areas in Croatia because there is enough "unlimited agricultural land." The scenario shows
the positive effect of Croatian accession to the EU according to agricultural production. The
second set of results in work shows the trends of the total agricultural production in the EU
and in individual Member States, so-called key markets. These markets (e.g. France,
Germany, and the Netherlands) are evenly important - for the EU and the Croatian model,
because they are included in the observation equations of domestic products and as
indirectly convey the impact of European and world prices.
On the supply side, assumptions like prices and policy work act evenly important in the
scenario projections. The demand of agricultural products is primarily influenced by changes
in prices, while the policy change effect is indirect. In the positve validation option of
Baseline, where we tested the model, with an increase in EU prices by 10% and reduce the
impact of price index of mineral fertilizers by 10%, production of all products to grow while
domestic consumption variable. Also, model reacted in opposite situation (negative validation
option: -10% EU prices: +10% fertilizer cost index parameter). Model responds to these
shocks in accordance with economic theory.
The purpose of this paper is not only the development of models that can be used to analyze
the (policy), but also because of the possibilities for creating a model in conditions of limited
or scarce data. Identifying the variables and relationships to be analyzed is the central
problem in the development of the model. In the process of specification, evaluation,
validation or model simulation, it often happens that the model, as it was originally set up -
does not work or does not show useful results. Change, re-estimation) and/or calibration of
certain equations in the model occurs several times in the process of developing models in
order to achieve the best image of the real situation. The limitation of data and the lack of
detailed studies on the effects of individual variables (parameters in the equation) can be
identified as the main constraints when creating the model.
At this moment, one of the advantages of the current model version is the fact that the model
is set up/constructed and that is consistent with economic theory and evaluated by Croatian
experts. However, in the model assessment, which is perhaps the most important, it is
important to show the results of the model understandable to people who have a lot of
knowledge of the issues covered in the model, as well as being critical to the method of
modeling, in general. |
