Presentation on theme: "A Comparative Theory of Legislation, Discretion, and Policy making Process (Huber&Shipan) Two crucial elements in the politicians- bureaucrats interaction."— Presentation transcript:
A Comparative Theory of Legislation, Discretion, and Policy making Process (Huber&Shipan) Two crucial elements in the politicians- bureaucrats interaction 1)Policy Preference divergence less delegation, more details in statutes to micromanage policy implementation 2)Technical complexity (and policy uncertainty) of the policy issues more delegation to utilize the bureaucracy expertise and vaguer legislation.
A Comparative Theory of Legislation, Discretion, and Policy making Process (Huber&Shipan) Additional elements 3)Legislative capacity that affects the costs to politicians of drafting detailed legislation (informational costs and decision making costs) 4)Bargaining environment (the number of actors have the authority to propose or veto legislation) 5)Non statutory features of political system that oblige the bureaucrats to behave consistently with politicians’ preferences.
Model of Designing Legislative Statutes Assumptions (1) Politician= legislators who determine the level of policy discretion in a statute Bureacrat= who implements the policy once the statute is adopted. Both actors are motivated by policy outcomes Policy=what is spelled out in legislation and during implementation Outcomes=what happens in the real world once a policy is adopted and implemented.
The Bureaucrat is better informed than the Politician about the repercussions of particular policy choices Designing a statute is costly to the Politician. These costs depend on 1.The extent to which the Politician limits discretion 2.General legislative capacity The Bureaucrat have a choice about whether to implement the policy chosen by the Politician. If he/she is caught to implementing a policy that the Politician does not like, he/she will pay a price. If no policy change occurs, then the status quo is retained Model of Designing Legislative Statutes Assumptions (2)
As it is empirically impossible to obtain minimally accurate measures of the actual preferences of bureaucrats H.&S. assume that political actors at the apex of bureaucratic structures strongly influence the policy preferences of the bureaucrats. In the case of the Parliamentary democracies these political actors are the Cabinet ministers. Model of Designing Legislative Statutes Empirical implication
Parliamentary model: only one politician (median member of Parliament or leader of Majority party or pivotal member of a majority coalition) unilaterally establishing the contours of the legislation Veto model: two politicians, one proposing, the other accepting or vetoing. Useful to understand logics in presidential systems Bicameral system: two politicians with both proposal and veto power. Three Institutional environments, three Models
Stage 1: the Politician decides whether to adopt new statute and the degree to which the statute actually limits the type of policies that the Bureaucrat can implement. In one dimension the Politician specifies the left and right boundaries on policy Basic model’s stages L R1R2 Statute 1 Statute 2 Less delegationMore delegation
Stage 2:The Bureaucrat implements the policy (x). Bureaucrats knows which outcomes will result from which policy outcomes. The politicians don’t. In one dimension the outcome (y) will be a mapping that is a specific direction and distance away from the original policy. If the distance is = 0 and x=y then Outcome= Policy; What happens if on the contrary the distance>0 and y x? Basic model’s stages
Stage 2. How policy uncertainty can affect the game ? Two scenarios: a.The policy implemented by B. is also the policy outcome (x0=y0) b.The outcome is 1 unit to the left of the policy implemented by B (x1>y1) Basic model’s stages LR2R1B x0=y0 y1x1 Politician’s preferred outcome Bureaucrat’s preferred outcome 1 unit
Stage 2. 1.Politician believes with some probability that a) or b) are true. 2.Bureaucrat knows with certainty if a) or b) is true. Basic model’s stages LR2R1B x0=y0 y1x1 Politician’s preferred outcome Bureaucrat’s preferred outcome 1 unit
Stage 2. First type of politician’s mistake 1.Suppose that the Politician L believes that (b) is very likely (x1>y1) and he/she allows the Bureaucrat B to implement x1. This requires a policy with a right boundary R1 (large delegation). 2.Suppose, against L’s expectation that (a) is true (x0=y0) B will implement its preferred outcome, far from L. Too much delegation! Basic model’s stages LR2R1B x0=y0 y1x1 Politician’s preferred outcome Bureaucrat’s preferred outcome 1 unit !
Stage 2. Second type of politician’s mistake 1.Suppose that the Politician L believes that (a) is very likely (y0= x0). He/she wants to constrain B by adopting a right- side boundary very close to L, such as R2 (Small delegation) Suppose, against L’s expectation that (b) is true (y2< x2) 2.Suppose B complies with the statute by implementing x2 that will produce y2, a bad outcome for both actors. Too small delegation! However… Basic model’s stages LR2R1B x0=y0 y2x2 Politician’s preferred outcome Bureaucrat’s preferred outcome 1 unit !
Stage 2. Why The Bureaucrat should have to comply with the statute even when the size of delegation does not allow to achieve the Politician’s interest ? The Bureaucrat must decide whether to comply with the statute by choosing a policy that is within the boundaries set by the Politician. In the previous assumption the Bureaucrat who does comply with the statute cannot be punished. So the Bureaucrat can prefer to be “unefficient” but “compliant”. Basic model’s stages
Stage 3. Non statutory factors (independent central banks, judges etc) can influence policy outcomes. Assumption: After the Bureaucrats act one of two things occurs: 1)No non statutory factors intervene to influence the policy outcome 2)These factors intervene shifting the policy outcome to the Politician’s preferred policy. In this last circumstance if the Bureaucrat is “disloyal” he/she will be punished and must pay a cost Both the Politician and the Bureaucrat know the likelihood of these two events. Basic model’s stages
The Bureaucrat’s implementation Decision LR2R1B y1x1 Politician’s preferred outcome Bureaucrat’s preferred outcome 1 unit CB1 y2 x2 1 unit Statute 2 Statute 1 If the Bureaucrat knows that the policy outcome will be one unit to the left of the policy that is implemented, then he/she would like to be able to implement a policy x1, that is one unit to the right of his/her ideal point B and yields y1 as an outcome. This possibility is given by the statute 1 as statute 1 gives enough discretion CB2
The Bureaucrat’s implementation Decision LR2R1B y1x1 Politician’s preferred outcome Bureaucrat’s preferred outcome 1 unit CB1 y2 x2 1 unit Statute 2 Statute 1 What happens when the policy the Bureaucrat wants to implement is outside the set of policies allowed by the statute (for instance when the statute 2 is adopted) ? CB2
The Bureaucrat’s implementation Decision LR2R1B y1x1 Politician’s preferred outcome Bureaucrat’s preferred outcome 1 unit CB1 y2 x2 1 unit Statute 2 Statute 1 If he/she adopts his/her most preferred policy x1 he may be caught and sanctioned, since this policy is outside L-R2 (statute 2); Alternatively he/she can adopt the optimal safe policy x2
The Bureaucrat’s implementation Decision LR2R1B y1x1 Politician’s preferred outcome Bureaucrat’s preferred outcome 1 unit CB1 y2 x2 1 unit Statute 2 Statute 1 In order to decide which policy to implement, the Bureaucrat will compare (nc) the net benefits from non compliance to (c) the benefits he would obtain from simply (and safely) complying with the statute. If the (c)>(nc) the Bureaucrat will comply with the law. CB2
The Bureaucrat’s implementation Decision LR2B Politician’s preferred outcome Bureaucrat’s preferred outcome 1 unity2 x2 Statute 2 The compliance boundary (CB) is the point at which the combined benefits and costs of non compliance are equal to the benefits of compliance. B will be indifferent between implementing a law that is a the compliance boundary and implementing a policy that yields the Bureaucrat’s preferred policy but that also risks sanction when the right boundary of the law, R, will coincide with CB. CB2 Compliance boundary depends on : a)Influence of non statutory factors b)Magnitude of sanctions c)Relationship between policies and outcome. If policy and outcome coincide then CB will be more to left.
The Bureaucrat’s implementation Decision LR2R1B Politician’s preferred outcome Bureaucrat’s preferred outcome 1 unit CB1 y2 x2 Statute 2 Statute 1 If sanctions or the probability of getting caught are large, the compliance boundary CB can be far to the left of B, for instance in CB1. In these circumstances both statutes 1 and 2 are right boundaries (R1 and R2) to the right (above) of CB1. Bureaucrats prefer to comply with both statutes than implementing a policy that yields B but also risks sanctions.
The Bureaucrat’s implementation Decision LR2R1B y1x1 Politician’s preferred outcome Bureaucrat’s preferred outcome 1 unity2 x2 1 unit Statute 2 Statute 1 When the compliance boundary depicts a situation where sanctions or probability of getting caught (or both) are quite low as in CB2, then the bureaucrat B will comply with Statute 1 with right boundary R1 (he can obtain B) but he won’t comply with Statute 2, which has a right boundary R2 to the left of CB2. (too small delegation!) CB2
The Bureaucrat’s Implementation Decision If the statute allows sufficient discretion (right boundary above, to the right of CB) the Bureaucrat will implement the policy that he most prefers among the policies that comply with the statute If the statute limits discretion too much (right boundary below, to the left of CB) the Bureaucrat will take a chance and risk sanctions by implementing the policy that yields his preferred outcome Bureaucrat is more likely to follow the Politician’s instructions if 1) the political system is likely to catch non compliant behaviour 2) If the potential non compliant behaviour’s benefits are low
The Politician’s optimal statute in certain circumstances For the politician will never be optimal to adopt a statute that the Bureaucrat is certain to ignore (when the right boundary R is to the right the CB). Adopting a statute that limits discretion is costly. The politician never would like to pay uselessly such a cost. The Politician would prefer to adopt a vague low cost statute that allows the Bureaucrat to implement the policy he/she most prefers than adopt a low-discretion, high-cost statute that results in exactly the same policy outcome. In other terms if for the disloyal Bureaucrat the probability to be caught is zero or the price to pay is null, then limiting the discretion by a detailed law does not make any sense.
Legislation and Discretion in the Parliamentary Model Politician must pay a cost to adopt a low- discretion law; this cost is a function of legislative capacity. Policy conflict (between Politicians and Bureaucrats) and legislative capacity interact to influence legislative statutes (and their level of discretion) Simplifying assumptions 1)No policy uncertainty 2)No nonstatutory factors (therefore no “CB”) 3)Bureaucrats will comply with the statute
Legislation and Discretion in the Parliamentary Model L B1B2 B3 Threshold c1 c2c3 c4c5c6 1)B1, B2, B3 = potential policy preferences of the Bureaucrats 2)c1, c2, c3 etc= different levels of costs which correspond different levels of legislative capacity. Less cost, more legislative capacity. c1 corresponds to the highest capacity; c6 correspond to the lowest capacity
Legislation and Discretion in the Parliamentary Model L B1B2 B3 Threshold c1 c2c3 c4c5c6 1)Imagine B1 and c1; a)Policy benefit of low discretion Statute is -|L-L|-c1 (the maximum of utility is of course 0); b)Policy benefit of high discretion Statute = -|L-B1|; Therefore the condition to have a low discretion Statute is –c1>-|L-B1| namely |L-B1|>c1. Is it respected ? Yes 2) Imagine B1 and c2; a)Policy benefit of low discretion Statute is -|L-L|-c2 b)Policy benefit of high discretion Statute = -|L-B1|; Therefore the condition to have a low discretion Statute is –c2>-|L-B1| namely |L-B1|>c2. Is it respected ? No. in fact |L-B1|
Legislation and Discretion in the Parliamentary Model L B1B2 B3 Threshold c1 c2c3 c4c5c6 1)Imagine c2 and B1; The policy benefit from the low discretion statute is smaller than the policy benefit from an high discretion statute: –c2 |L-B1|; a small increase of the policy divergence does not matter. 2) Imagine c2 and B2; now –c2> -|L-B2| namely c2<|L-B2|; The policy benefit from the low discretion statute is greater than the policy benefit from an high discretion statute discretion statute Therefore holding legislative capacity fixed, an increase in policy conflict from B1 to B2 leads to adoption of a low-discretion statute
Legislation and Discretion in the Parliamentary Model L B1B2 B3 Threshold c1 c2c3 c4c5c6 1)Legislative capacity and policy conflict can have an interactive effect. May exist a threshold of policy conflict that must be surpassed before the legislative capacity can influence discretion in legislation or may exist a threshold of legislative capacity before the policy conflict can influence discretion in legislation. 2)two distribution of legislative capacities (two cross-country dataset). Distr.1 has a much wider range of legislative capacities than Distr.2 Distribution 1 of Legislative Capacity Distribution 2 of Legislative Capacity
Legislation and Discretion in the Parliamentary Model L B1B2 B3 Threshold c1 c2c3 c4c5c6 1)If you consider Distribution 1 it would be unlikely to uncover direct effects of policy conflict and legislative capacity on discretion because many of the political systems have insufficient capacity to ever adopt a low-discretion statute. 2)As the costs of adopting low discretion laws decrease from c6 to c5 or even to c4, nothing happens ; 3)As the costs are above the threshold, then changes in policy conflict from B1 to B2 or to B3 would have no effect on the adoption of low-discretion statute Distribution 1 of Legislative Capacity
Legislation and Discretion in the Veto model A politician (the Legislature) adopts a policy. Another Politician (the President) has a veto prerogative. Additional simplifying assumption: President and Bureaucrats have the same preferences
Legislation and Discretion in the Veto model In general the same results of the Parliamentary model. Relative value of low discretion statutes increases 1) As policy conflict between Politician and Bureaucrat increases 2) As Legislative capacity increases 3) As non-statutory factors become less reliable. However in addition Status quo position matters as now gridlock is possible The threat of veto can never lead to less discretion but sometimes to more discretion than we would find in the Parliamentary model.
L=R1 P=BQ 1)For simplicity assume that: a)All the players in the model know that the final policy outcome will be the same as the policy the Bureaucrat implement (x=y) b)The Politician (the Legislature) has legislative capacity to adopt any statute he/she wishes (c 0) c)The Bureaucrat will comply with any bill. 2) Without any veto a Politician could implement a low-discretion bill with a left and right boundary at L, his most preferred policy. (L=R1). 3) The President P would realize that he prefers the status quo Q to L and would veto such a bill. 4) The politician L must give the Bureaucrat leeway to implement a policy that the President will accept.For example a policy with an upper bound at R2. The Bureaucrat would implement R2 and P would not veto anymore.. R2
Legislation and Discretion in the Bicameral Model (with veto) Both Chambers (Upper and Lower) can make policy proposals and they do sequentially If the first proposal is met with a counterproposal by the other house, then the two chamber must pay a bargaining cost associated with the opportunity cost of delay.
L=R1 P=U=BQ For simplicity assume that: a)Bargaining cost of a counterproposal = c b)U=B=P c)Neither Chamber pays any other cost of making a proposal d)The Upper Chamber U has made a policy proposal and the Lower Chamber L now has to decide whether to pay the bargaining costs c of making a counterproposal. e)The best proposal the Lower Chamber L can make is R2 because of the presidential veto. However such a proposal entails also a cost c f)Upper chamber can take advantage of this situation and propose R3 as an upper bound on policy since the beginning. Upper chamber has extracted the bargaining cost from the Lower Chamber in the form of increased legislative discretion. When Bicameral bargaining is necessary, the discretion in the statutes cannot decline and often will be greater. R2R3 c