Humboldt-Universität zu Berlin - DFG research unit SiAg 2010-2013


  Firm level    
SP 1
Agricultural entrepreneurs' decision making and structural change: An experimental approach
SP 2
Structural change and dynamic efficiency of German dairy farms
Odening / Hüttel
SP 3
Modeling farm level structural change in a dynamic competitive environment
SP 4
Market structure and organization in agri-food value chains: An application to the German dairy sector
Hockmann / von Schlippenbach
  Intermediate level    
  SP 5
Between path dependence and path creation: The impact of farmers’ behavior and policies on structural change in agriculture
Balmann / Kataria
SP 6
How should model linkages be designed to analyze the effects of global agricultural trade liberalization at the farm level?
Brockmeier / Offermann
  Policy- & sector level    
  SP 7
Energy from biomass: Linkages between the agricultural and the energy sector in the EU
Grethe / Blesl
SP 9
Econometric evaluation of
CAP impacts in Germany



Subproject 1:

Agricultural entrepreneurs’ decision making and structural change: An experimental approach

Short title: Entrepreneurial decision making

Principal investigator: Schade

The rational calculus assumed in most economic models is quite unrealistic and not predictive of economic agents’ decision-making. This statement holds for the existing literature on structural change in agriculture: based on a quite abstract characterization of agricultural entrepreneurs and not fully reflective of their actual decision-making. Hence, the understanding of structural change in agriculture can be improved via a realistic modeling of the decision-making of agricultural entrepreneurs. Applying and extending psychological and decision-oriented approaches from entrepreneurship research will help to better predict farmers’ behavior and choices. Economic experiments are important to evaluate the predictive power of such approaches and to test them against economic benchmark models.
The underlying paradigm of our research is that of bounded rationality. With the overall aim of elaborating implications for regulatory agencies, this project seeks at modeling decision-making in scenarios relevant for farming. It thus aims at overcoming the quite abstract characterization of the agricultural entrepreneur in the agricultural economics literature, at applying entrepreneurship research to farming and farmers in general, as well as generating experimental evidence on farmers’ decision-making, which is missing.
Building upon our experimental work on entrepreneurial behavior in general, on entrepreneurial disinvestment choices, and on coordination problems in the efficient allocation of land with economies of scale, we plan to focus on three research domains in the prolongation period:

  1. understanding the economic and psychological drivers of farm exit;
  2. deepening our understanding of evolving and persisting market structures as well as structural transition in agriculture; and, based on the first two,
  3. better characterizing the agricultural entrepreneur.

The first research domain is concerned with an issue crucial for understanding structural change in agriculture: inertia, i.e., “holding on for too long”, in farm exit and disinvestment choices. There are rational (real options-based) and psychological explanations for such inertia. Laboratory experiments have been successfully developed and carried out in the current research period to disentangle these effects, but further research is necessary to more deeply understand the psychological drivers. Methodologically, economic experiments will be enriched via psychological questionnaires and framing manipulations.
Based on experimental evidence generated in the current research period, the second domain focuses on better understanding barriers which hamper the transition to efficient land units and also focuses on developing and testing new auction types for allocating land plots that take into account the bounded rationality of the interacting farmers. Recommendations for the regulation of land markets and for the consulting of farmers are expected.
The third research domain is partially based on the first two, but requires carrying-out part of our experimental studies with different participants to better characterize the profile of agricultural entrepreneurs via a comparison with farmers that are non-farming entrepreneurs and non-entrepreneurs.

Subproject 2:

Structural change and dynamic efficiency of German dairy farms

Short title: Structural change and dynamic efficiency

Principal investigators: Odening / Hüttel

The German dairy sector is the biggest in the EU and reflects both the complexity of agricultural production in highly regulated markets and the complexity of the livestock production system. Since 1984 EU dairy policy has been characterized by a milk quota system and its associated intervention prices. Under the milk quota, classical economic principles such as growth and scale do not necessarily hold. Within the 2003 CAP reform, the decoupling of premia from any production level and a further reduction of the intervention prices induced sizeable adjustment processes at the farm level such as growth or abandonment of milk production. The 2008 health check and falling prices induce further pressure on farms to adjust, which may further alter the farm structure. It is widely acknowledged in the literature that structural change is closely related to the efficiency of firms within a sector. Firms showing superior performance and higher efficiency increase their market share at the expense of less efficient firms, thereby increasing concentration.
The overall objective of this subproject is to improve the understanding of structural change in the German dairy sector. The relevance of this topic emanates from the dynamics, which has characterized dairy production in the past, in conjunction with the value that the dairy sector adds to farm income in Germany and the EU. We want to identify the main drivers of the adjustment process at the farm level. Thereby we aim to pave the way for predictions of developments in this sector, in particular after the milk quota period. The first specific aim focuses on the impact of the milk quota scheme on structural change; of particular interest are the determinants of abandoning milk production. The second specific objective of the research is to investigate the relation between efficiency and structural change in the German dairy sector. It will be interesting to investigate if high efficiency actually translates into higher profitability and competiveness or if the “poor but efficient” hypothesis holds, particularly for small dairy farms. Moreover, we want to understand why certain farms turn out to be efficient and competitive and if efficiency is a crucial determinant of farm closure.
The aforementioned objectives require sophisticated methods that are not yet fully available. An important contribution of this subproject is the refinement of existing models that allow for a quantitative analysis of adjustment processes in dairy farming. The first focus lies on econometric models that explain exit decisions in dairy production. Basically, this refers to a real disinvestment option with a decreasing value the closer the (uncertain) abolishment of the milk quota is. Existing econometric models need to be broadened to account for uncertain future revenues, sunk costs and the role of the milk quota scheme. The second focus is on models for dynamic efficiency analyses. Based on a dual model of intertemporal decision-making, a shadow cost approach is used that allows for an econometric estimation of dynamic efficiency under uncertainty. Finally, it is aimed to test the impact of farm efficiency on exit decisions in the German dairy sector. Both approaches are based on German farm level panel data from the farm accountancy data network that requires the use of panel data methods.

Subproject 3:

Modeling farm level structural change in a dynamic competitive environment

Short title: Real options

Principal investigator: Mußhoff

Agrarian structures are considered to be an important factor for the competitiveness of farming. Therefore, agricultural economists have been concerned for a long time with describing, analyzing and modeling structural change in agriculture on the one hand, while on the other hand, high political attention is devoted to the competitiveness of agricultural production and the income situation of farms. To adequately forecast structural change in agriculture, in general, and to analyze the effects of different political schemes, in particular, a profound understanding of farm level structural change, i.e., investment and disinvestment decisions of farmers, is necessary.
When attempting to model, explain and forecast structural change in agriculture in the past, only sub-aspects of the complexity of real decision-making have been considered. Presently, no model exists that accounts for uncertainty, dynamics and irreversibility, although theses phenomena coexist in many decision problems. Another important issue, which is frequently neglected in farm level models, is competition, i.e., interactions between different agents. The overall objective of subproject 3 is to develop a normative model, which enables realistic prognosis of farms’ investment behavior and its implication for structural change in agriculture. The theoretical concept of the model is based on the real options approach. Real options theory allows for the quantification of the value of entrepreneurial flexibility under conditions of uncertainty and irreversibility based on the well-founded results of financial option pricing theory. Nevertheless, difficulties arise when analyzing (dis)investment decisions in a competitive environment because real options in contrast to financial options are non-exclusive. Thus, a direct determination of equilibria in competitive markets is necessary, which is commonly assessed as not practicable. We will do this by using an agent-based real options market model. Furthermore, this model will incorporate bounded rationality of decision-makers. For this purpose, results from the first project phase can be utilized. With these extensions, the real options model builds a starting point for an improved policy impact analysis.
The investment and disinvestment model will be applied to the dairy sector which shows a high (and further increasing) intensity of competition. Investments in the dairy sector are generally highly specific and thus cause sunk costs. Consequently, resale values of buildings are low. At the same time, demand shocks, fluctuations of weather conditions, climate change and epidemics such as BSE, in combination with a temporal delay in adjusting production, lead to high price fluctuations and long-run uncertainty in the milk market. Moreover, investments in the dairy sector are not “now or never” decisions. Such decisions can instead be temporally postponed. For decisions to invest in the dairy sector, all three conditions of “uncertainty”, “irreversibility” and “flexibility” of the real options approach are fulfilled. With respect to the dairy sector, the consequences of different policy schemes such as a rapid abolishment of the quota system, soft landing and / or the introduction of a minimum price for milk will be analyzed, i.e., changes in trigger prices, markets supply, farm profits and overall economic efficiency.

Subproject 4:

Market structure and organization in agri-food value chains: An application to the German dairy sector

Short title: Market structure

Principal investigators: Hockmann / von Schlippenbach

The German dairy sector is facing a period of significant changes. The number of both farmers as well as dairy processors has drastically declined over the last few decades. The wide range of literature tackling adjustment processes in agricultural sectors often ignores retail activities. As long as the retail sector was highly fragmented and constituted a “transparent” window between suppliers and consumers, this approach was appropriate; however, this has changed dramatically since the retail sector has been subject to a profound consolidation process. Due to the emergence of increasingly dominant agents at all stages of the dairy sector, strategic interactions along the value chain have become more important. In particular, the increasing buyer power of retailers has raised concerns in recent years.
Against this background, we aim at analyzing competitive forces along the value chain. So far, there exists no workhorse model that allows for the analysis of interdependencies in a three-layer structure where imperfect competition is assumed at all three stages. We close this gap by developing a general model that allows us to analyze negotiation outcomes between retailers, food processors and farmers. Based on this framework, our research project is divided into two work packages which tackle two different issues of structural change in agricultural value chains:
In the first work package, we focus on merger incentives at the processor level. If products are both conventional and strategic complements, a merger between processors may soften their bargaining position vis-à-vis retailers, while a merged processor has a better bargaining position vis-à-vis farmers. To capture this trade-off, we construct a simple theoretical model that considers the negotiations of processors with both farmers and retailers. When analyzing potential drivers of structural change, we take into account how changing marginal costs at the processors’ level, the processing capacities, as well as a variation of the number of upstream firms change the outcome in terms of endogenous market structure and delivery tariffs negotiated between processors and farmers as well as between processors and retailers. The empirical part of this work package focuses on econometric modeling and estimation of market structure development at the processor level. The empirical model will be developed within an Olley-Pakes framework which, in addition to the determinants of market structure, provides consistent information about productivity differences between processors.
In the second work package, we examine competition between different organizational forms such as cooperatives and for-profit firms at the processors’ level. We endogenize the size of cooperatives in a mixed duopoly equilibrium. We aim to understand how differences in ownership structures and in distribution of market power along the chain affect negotiation outcomes in vertical relations. We further intend to highlight how different ownership structures affect the endogenous quality choice of farmers and processors. The empirical investigation in the second work package focuses on the explanation of market performance and the identification of market power. Within this context, a structural market model will be developed that combines the theoretical findings of both work packages.

Subproject 5:

Between path dependence and path creation: the impact of farmers' behavior and policies on structural change in agriculture

Short title: Path dependence

Principal investigator: Balmann / Larsén

Compared to the regional heterogeneity of agricultural structures in Germany, the speed of structural change is slow and it cannot be expected that regionally different farm structures converge. This is puzzling considering regional heterogeneity is accompanied by the dominance of suboptimal farm sizes in many regions and – despite huge subsidies – by significant functional income disparities. One explanation for these observations is a path dependency of structural change. The main objective of this study is to have a better understanding of whether structural change is indeed path dependent, what the determinants are and how path dependence eventually can be overcome.
This study will especially focus on the German dairy sector. The first reason is that structural change in this sector seems to be particularly slow as a vast majority of dairy farms either operate with inferior techniques or apply them in a less economical way. Also, the regional heterogeneity of farm structures is huge with, e.g. many small farms in Bavaria and a relatively low number of large dairy farms in Saxony-Anhalt. The second reason is that the dairy sector is particularly affected by the ongoing liberalization of the CAP. Accordingly, dairy farmers, their representatives and politicians are highly concerned about the future of this sector. A third reason is that many German farms are engaged in this relatively labor intensive business which is often located in disfavored grassland regions.
The underlying hypothesis of this project is that structural change in the dairy sector is path dependent. In particular, it is hypothesized that existing structures are not the result of a long-term optimal development but instead are the result of historical events. Although historical events may have had a certain economic rationale, either dynamic inefficiencies accumulated overtime (or at least persisted) or the benefits of the incidents were exploited earlier. To analyze this, the concept of path dependence will be applied to identify and evaluate its potential reasons. Factors that could help overcoming path dependence in the direction of path creation or path breaking will also be explored.
The methodological starting point will be the agent-based model AgriPoliS which will be used to analyze structural change in agriculture by simulation experiments. In addition to previous studies which applied AgriPoliS or its predecessor to analyze the impacts of sunk costs, market frictions and agricultural policies, this study will particularly focus on the role of behavioral issues for path dependence. On the one hand, participatory laboratory experiments involving human players (farmers, students) will be used to improve the agents' behavioral foundation in AgriPoliS. On the other hand, simulation experiments with AgriPoliS will be used in combination with stakeholder workshops to learn how farmers, officials, and other stakeholders from two selected case study regions (Stendal in Saxony Anhalt and Allgäu in Bavaria) perceive issues and trends of structural change and agricultural policies. This will facilitate a better understanding of whether mental models of individuals and society affect the behavior of farmers and policy formation. At the same time, the workshops aim at validating AgriPoliS and its regional adaptation, as well as identifying realistic scenarios which will be analyzed jointly with the stakeholders.

Subproject 6:

How should model linkages be designed to analyze the effects of global agricultural trade liberalization at the farm level?

Short title: Assessment of model linkages

Principal investigator: Brockmeier / Offermann

The current round of WTO negotiations, also known as the "Doha Development Round", is an ambitious effort to make globalization more inclusive and help the world's poor, especially by slashing barriers and subsidies in farming. Nevertheless, the ultimate impacts of a WTO agreement are far from obvious. In particular, the often extremely complex policies governing agricultural sectors in many countries pose great challenges for a scientific policy impact assessment. The implications of agricultural trade liberalization need to be analyzed considering specific policies and effects at the single country, region or farm level, while at the same time accounting for the influence of trade liberalization at the global level. Currently, none of the existing analytical tools are able to answer the relevant policy questions at the global, sectoral and farm level. Thus, a model linkage is needed that captures impacts at both the micro and macro level. Model linkages of this kind are of great current interest, but approaches and techniques are mostly still under development or in their infancy. Additionally, model linkages are increasingly developed for and applied in policy advice, so that results need to be delivered with a very tight schedule. Under these conditions, it is nearly impossible to thoroughly consider the implications of different options to link models.
The overall objective of the project is therefore not only to link models that deliver results disaggregated to various levels, but also to analyze in detail the potentials and pitfalls of model linkages, and determine which links are most appropriate and feasible for the analysis of global agricultural trade liberalization. Therefore, the project aims to answer the following questions: How does structural change at the farm level influence aggregate supply and technical progress? Under which conditions is it possible to derive macro-relationships from micro-relationships? How does the aggregation level influence the model results and how can possible problems be overcome? What roles do bottom-up and top-down approaches play in this regard?
These research questions will be explored using a linkage between GTAP (Global Trade Analysis Project) and FARMIS (Farm Modeling Information System) to analyze the impacts of global agricultural trade liberalization in the context of structural change in the German agricultural sector. Foremost, the definition and specification of variables common to both models, the model behavior as well as behavioral parameters will be aligned and harmonized for a common baseline scenario. To increase consistency and the scope of results, the project will then investigate if bottom-up- and top-down linkages are more appropriate to answer the research question raised. This will entail an analysis of the conditions needed for an exact or suitable approximated aggregation of farm level structural change to macro modeling or for disaggregating macro level structural change to micro models. Sensitivity analyses will be extensively employed throughout the project to systematically highlight the implications of different approaches to link the models.

Subproject 7:

Energy from biomass: Linkages between the agricultural and the energy sector in the EU

Short title: Energy from biomass

Principal investigators: Grethe / Blesl

Over the last three decades real energy prices have increased relatively to real prices for agricultural products. Consequently, bioenergy as a share in total energy demand has increased worldwide and is expected to increase further. Partly, this tendency is due to the market mechanism: relative market prices are such that biomass is turned into energy, e.g., sugar cane into ethanol in Brazil or solid biomass for heating private homes in many industrialized countries as well. Partly, however, this tendency is policy driven. Due to a broad variety of drivers (climate policy, the security of energy supplies, agricultural and rural policy motives), many governments in industrialized as well as in developing countries are supporting the conversion of biomass into energy by means of tax exemptions, mandatory blending, subsidies and other policies. This also holds for the EU, where the production of fluid biofuels is policy driven and would not happen at market prices.
As a result, the potential supply of biomass for energy production (which depends on prices) has an impact on the future energy balance, and demand for energy from biomass has an impact on agricultural markets. This interrelationship has often been analyzed either based on energy system models, assuming a given biomass supply, or based on agricultural sector models assuming a given biomass demand for energy. Alternatively, some studies address this market interdependence based on general equilibrium models with a very stylized representation of the energy sector.
The objective of this subproject is to ex-ante analyze the interdependence between the energy and the agricultural sectors in the EU under energy as well as agricultural policy scenarios. To address the weaknesses of existing simulation models, the analysis will be based on the combined use of two well-established partial models: the Integrated Markal Efom System (TIMES) PanEU Model, which is a bottom-up dynamic energy system model with a rich technology representation of the EU energy system, and the European Simulation Model (ESIM), which is a partial equilibrium comparative static agricultural sector model with a rich representation of EU agricultural policies and market interdependencies. Both models are programmed in GAMS and both models already include, although in a rudimentary form, linkages between agricultural and energy markets: ESIM depicts the impact of an exogenously given demand for biofuels on agricultural markets and the TIMES PanEU model depicts the effect of an exogenously given supply of biomass on the EU energy system.
The work program includes the identification and creation of relevant interfaces and exchange variables for both models, the conceptualization of the regional dimension of bioenergy markets, the further development of both models (e.g. the integration of agricultural by-products relevant for energy production such as straw and manure in ESIM), as well as scenario development and analysis. Close interrelations exist with subproject 6: the interface with FARMIS in the first project period allows regional and farm specific effects of energy policy scenarios to be addressed; and subproject 7: the inclusion of agriculture in EU climate policy will have effects on the potential of the agricultural sector to supply biomass for energy, which will be taken into account in the ESIM / TIMES PanEU framework.

Subproject 9:

Econometric evaluation of CAP impacts in Germany

Short title: Impact evaluation

Principal investigator: Petrick

The aim of this subproject is to develop and apply regression models that analyze the effects of agricultural and rural development policies in German regions. Following the overall goal of structural econometric modeling of policy impacts, there are three stages to this process. First, microeconomic models of farm behavior will be developed that can accommodate the range of measures currently embodied in the Common Agricultural Policy (CAP). Second, econometric models are devised which address the identification problems inherent to a CAP-related empirical impact analysis. Third, the econometric models will be applied and their performance will be assessed. The research agenda of this subproject directly builds upon work in the ongoing subproject, “Econometric impact analysis of rural development policies” within the DFG Research Unit 986.
Currently, the EU spends approximately 5 billion Euros annually on decoupled direct payments to German farmers. Furthermore, agricultural policymakers have been increasingly relying on differentiated measures for ‘rural development’ which, besides supporting agricultural enterprises, should also assist environmental aims and bolster the economic strength of rural areas in general. The most important instruments in Germany – by relative budget allocation – for the recently expired aid period from 2000 to 2006, included agro-environment measures, measures for village regeneration, farm investment support and payments for farmers in less favored regions. In this period, approximately 8.7 billion Euro of EU funds were distributed in Germany for rural development; for the current aid period, 2007–2013, 8.1 billion Euros are budgeted. Thus, the question must be asked as to whether these measures will influence structural change in the agricultural sector in a socially desirable way.
Developing appropriate econometric methods for such an analysis and their application at the administrative district level (Landkreise) in selected German Länder is at the core of this subproject. The approach followed here aims at quantifying policy effects at the regionally aggregated level. Basing the analysis on territorial observation units also allows for investigating instruments which are not directly aimed at agricultural enterprises, e.g., the measures for village regeneration which are of particular importance in eastern Germany. The subproject draws upon recent literature dealing with multiple, continuous treatment effects in a panel data setting. Building upon ongoing work from the first phase of this subproject, the primary focus will be on three extensions to existing models:

  1. strengthening the microeconomic underpinnings related to CAP effects on farmers’ behavior;
  2. explicitly considering dynamics of farm structures in the econometric models;
  3. relaxing linearity assumptions that are typically central to these models.

The subproject addresses a central methodological aim of the research unit, namely the development of a methodological framework for analyzing and shaping structural change in the agricultural sector. Parameter estimates provide an important foundation for further analysis in other subprojects.