Keywords

Lobbying is the backbone of representative democracy. While elections are the main vehicle for selecting politicians and holding them accountable, lobbying is the principal means to influence public policies between elections. In fact, lobbying is so deeply intertwined with politics that understanding policy outcomes is rarely possible without taking lobbying into account. While lobbying is often seen in a very negative light in public discourses, it is often seen more benevolently in scientific discourses.

One of the main aspects of lobbying is the provision of information to policymakers. Interest groups are better informed than politicians on specific effects and aspects of public policies. It seems natural that politicians would rely on this expertise to draft better legislation (Esterling 2004; Truman 1971). However, one problem arises: By definition, interest groups are self-interested political actors, pushing their own agenda. This leads to the important question when will interest groups share their information truthfully with policymakers and when will they try to misrepresent the information? The answer to this question depends crucially on the communication strategy of interest groups.

This book set out to improve our understanding of interest groups’ choice of lobbying strategies, which are understood as the way in which an interest group communicates with politicians. It is crucial to take the political context of lobbying into account. Any attempt to try and understand lobbying strategies without reference to the political process in which lobbying is embedded will at best provide a partial understanding of interest group strategy choice.

While my approach has a strong micro foundation, it moves beyond the numerous attempts which try to explain interest group strategy choice by individual group characteristics. These often fail to take the context of actions into account. But interest groups do not always use the same strategies. I have shown that some of the strategies are structurally equivalent once one conceives their intended purpose as the reference.

Many of the studies who engage with these questions are focused on the US Congress (e.g. Kollman 1997; Baumgartner et al. 2009). While the research on Congress helped us in many ways to enhance our knowledge of lobbying processes, it is hard to generalize to other political systems due to the many peculiarities of the political process. Similarly, theoretical work is largely institution free in the sense that generic communication processes are modeled (e.g. Crawford and Sobel 1982; Potters and van Winden 1992).

The political process is a prime example of a situation of decision-making under uncertainty. Legislators face uncertainty when deciding whether to adopt or reject a bill. This fundamental uncertainty results from the difference between the policy that is adopted and the outcome of the policy. The demand for lobbying is created by politicians’ dependence on information which is essential to make a better decision. Interest groups, on the other hand, face process uncertainty generated by the bargaining environment. The logic of political decision-making in a parliamentary system with proportional representation is very different from the US context.

Adapting the existing models to the setting of a parliamentary system was the first challenge to be tackled in the book. Toward this end, I developed a game theoretic model of lobbying which takes into account the institutional setup in a parliamentary system such as Germany. The model illuminates the differences in the bargaining environment and how they structure incentives for lobbying. The activities of interest groups were modeled by combining a signaling model of lobbying with a veto bargaining model. The model allowed me to open the black box of the single decision-maker in standard models of informational lobbying (e.g. Potters and van Winden 1992).

Similar to standard models, the assumption is that interest groups are better informed about the fundamental causal mechanisms. Interest groups possess valuable knowledge about how policies translate into policy outcomes, they face no fundamental uncertainty. However, as political decision-making is a bargaining situation in which the outcome is the result of a legislative process, the outcome of this process is usually not certain. The interest group, therefore, faces a serious amount of uncertainty about the potential outcome of its lobbying strategy. I have called this uncertainty process uncertainty. A couple of factors influences the degree of process uncertainty. One of the most important is the level of fundamental uncertainty. As such, the two types of uncertainty are closely connected. This is where my model deviates significantly from standard models of lobbying.

Due to process uncertainty, interest groups lack full knowledge of the exact outcome of their lobbying strategy ex-ante. This is often overlooked in analyses of lobbying but has strong implications. The exchange relations inherent to the lobbying situation are characterized by a double uncertainty. The higher the fundamental uncertainty, the higher the demand for information, i.e. lobbying. At the same time, interest groups become less able to predict the policy outcome. In order to understand interest group strategy choices, it is vital to look at what happens after the influence attempt from the interest group’s point of view.

The choice of lobbying strategy encompasses two interrelated decisions: The first is to become active or not, i.e. when do interest groups mobilize? The second is—conditional on mobilization—which type of communication does an interest group choose? Choosing a strategy implies to opt for a target and a mode of communication. The mode concerns private vs. public communication activities. Potential targets of the communication acts are the relevant decision-makers, which can be either the chambers in a bicameral legislature or the coalition partners in government in the unicameral case.

For understanding process uncertainty and how it affects interest groups’ incentives to provide information, it is essential to see the world from an interest groups perspective. It is central to identify the likely outcomes which result from different lobbying strategies in order to understand what drives interest groups’ activities regarding a specific bill. Working toward a prediction of interest group behavior, I have turned the problem around and asked the question what is the expected policy outcome conditional on the interest groups strategy. Interest groups base their choice of communication strategy on the expected outcome and the expected variation in the outcome. The latter part can be considerable due to the risk aversion of interest groups.

The bargaining stage considerably changes the strategic incentives of the interest group compared to a situation where the interest group signals to one decision-maker. A second dimension is added to the problem as by choosing a communication strategy the interest group can alter the range and variation of policy outcomes but not necessarily the expected policy outcome. Thus, the interest group faces two—possibly opposing—incentives: (a) While a communication strategy may shift the expected outcome in a more favorable direction, and (b) this may come at the expense of higher variability of outcomes. This trade-off is the result of the interdependence of the decision of the political actors.

My model implies that interest group is less powerful than is widely believed. Interest groups which are against a change of the Status Quo will lobby not to prevent a change, which is impossible, but to reduce the variation of outcomes. They are more likely to use a public communication strategy compared to groups who are in favor of a change of policy. The latter are more likely to use a strategy which involves private messages.

Interest groups mainly affect the variation of outcomes by changing the information structure in the political system. However, they can only modestly affect expected policy outcomes. This potentially opens a new direction for empirical research as it shows that many scholars may have been looking for effects in the wrong places.

The interest group’s ability to affect political bargaining varies considerably with the degree of fundamental uncertainty. The influence of interest groups increases with the level of fundamental uncertainty as for higher degrees of uncertainty a change of the Status Quo without interference by the interest group is less likely.

The constraining effect of the compromising actor is also stronger. Ceteris paribus, a higher level of process uncertainty implies lower fundamental uncertainty as the relative size of the two types of uncertainty matters.

Another factor is the risk aversion of interest groups which creates trade-offs between expectation and variance of policies. In case of low uncertainty, the utility loss from the variation of policy outcomes is indeed greater than the gains from changes in expected policies.

Decision-making institutions thus play a major role in the determination of policy outcomes. They feed back on interest group strategies as the groups need to anticipate the likely results of their activities. By highlighting preference constellations, I move beyond approaches who only explain lobbying by the context of a bill.

I have derived several hypotheses from the model. First, interest groups which prefer a move away from the Status Quo are more likely to use private messages, while interest groups opposing a change of the Status Quo are using more public strategies. I have operationalized this by the distance of the interest group to the center of the bargaining range. There is strong empirical support for this hypothesis, although the effect is reversed for opposition bills. This suggests that differences in interest group behavior between opposition and majority bills deserve more attention.

Likewise, as predicted by the theoretical model, pro-change groups are much more constrained in their lobbying activities and on average less active than anti-change groups. Other, predictions are that higher costs imply less activity and that higher fundamental uncertainty increases lobbying activity. The data strongly support both claims.

I have used both standard logistic regressions and cross-classified hierarchical models to estimate the effects. While the latter account for likely correlations in the error structure of the data, the results are identical. I have shown in several additional analyses with alternative specifications and alternative operationalizations of the dependent variable that the results are robust.

My endeavor has demonstrated the potential to derive better predictions of the lobbying strategies of interest groups if one take political processes and preferences into account. The perils of oversimplifying by clustering interest groups into types or by ignoring the political process cannot be overstated. More work should be dedicated to disentangle different aspects of interest group’s utility calculus to better understand their actions. One aspect that should receive more attention in the future is how interest group competition affects the choice of strategies.

If uncertainty is not too high, interest groups affect mainly the variance rather than the mean of the policy outcome. This is an important feature of the model and may explain many of the non-findings in empirical research. If the expected outcome is (almost) the same with or without lobbying, a mean centered idea of causality will never lead to strong empirical insights.

Once a causal effect is defined in terms of the difference in variances as Braumoeller (2006) does, one may be able to find empirical evidence on the effectiveness of interest groups. My findings call for a new approach to empirical analyses of the determinants and effects of interest group activities. Current approaches center mainly on policy outcomes. A more nuanced approach that analyzes the variance of political outcomes as a function of political uncertainty may be the more fruitful approach. A major advantage is that political uncertainty varies with the institutional setup of a polity and thus my model opens new possibilities for comparative research designs. The model strongly suggests to direct the effort more into this direction.

The formal model has several normative implications, which are in line with the empirical findings. First, similar to other models of information transmission, very extreme pro-change interest groups are effectively taken out of the equation. This can be seen in the distribution of ideal points. There are no ideal points more extreme than the political decision-makers on the pro-change side. Extreme pro-change groups are unable to truthfully communicate with decision-makers. The presence of a moderate second chamber (or coalition partner in a unicameral setting) creates this constraint for both public and private communication.

At the same time, compared to a moderate chamber or coalition party, the presence of a relatively progressive second chamber increases the possibility of information transmission by public messages. The moderate chamber is the constraining factor in the political process. Inducing change requires loosening this constraint. More groups will be able to truthfully communicate with this chamber using public compared to private communication. Public communication is therefore not necessarily evidence for Status Quo groups trying to preserve public policies.

Second, in the cases where truthful communication is possible, the presence of lobbying enhances the welfare of both the interest group and the decision-makers. To the degree that decision-makers’ preferences reflect preferences of their constituencies, this also enhances the welfare of society as a whole. Biased information can be beneficial.

Lastly, the model implies that there may be a new rationale for public communication by Status Quo groups. It is often claimed that public communication has three goals: Signaling to constituencies, mobilization and conflict expansion (Kollman 1998).

In parliamentary systems, these goals are important as well. However, given that interest groups are unable to prevent change once political decision-makers agree to change public policies, risk aversion may add another dimension: The reduction of variation in public policies. Ex-ante, interest groups face uncertainty about the exact outcome of the political process. Successful public communication will not affect the (ex-ante) expected outcome, but will reduce variability of outcomes. Risk averse interest groups do value this, as it makes the process more predictable.

The predictions of the theoretical model are less accurate for opposition bills. Interest groups which are in favor of a change of the Status Quo are more likely to send public messages in case of opposition bills. Given that opposition bills have little chances of being accepted in the German parliamentary democracy, this seems to imply a difference in the strategic calculus of interest groups. This case is not directly covered by the model and would require further analysis. A possible explanation is that the messages are a signal to the group’s constituency helping to ensure group maintenance.

Future work should also take the claim serious that some of what lobbyists do affects the variance rather than the (expected) outcome of policies. A main challenge will be the operationalization and measurement of these concepts and to develop a clear empirical strategy to test for the effect.