How to regulate raises questions about what interventions to select and how they should be designed to achieve the intended impact. Choice of intervention is sometimes called instrument choice or policy design.
Regulation narrowly defined includes the use of primary secondary and tertiary legislation. However traditional regulatory tools can be complimented by other broader instruments such as licencing and contracts.
According to Wiener and Richman, a distinction can be drawn between normative mechanism choice and positive mechanism choice. Normative mechanism choice is about what instruments we should choose, and is defined as follows:
“Mechanism choice is a kind of social ‘engineering’, the task of designing optimal instruments to achieve social objectives” (Wiener and Richman, at 363).
By contrast, positive mechanism choice is about what instruments actually get chosen and often relates to political and policy based considerations:
“If positive politics do affect the process of mechanism choice, then political forces predictably constrain policy options or raise the costs of certain policy instruments. Accordingly, understanding these political constraints (which may be impossible or too costly to change) should inform the normative instrument choice analysis.” (Wiener and Richman, at 371).
They describe many instruments in the regulatory tool box – rules about conduct, quantity/property, price/liability, information disclosure, and government ownership and private ordering (Wiener and Richman, 365-369).
At 365: “Conduct instruments involve government commands to firms and individuals, prescribing specific technologies, practises, methods, or behaviours that must be employed or must be avoided. These instruments mandate conduct, rather than outcomes.”
At 366: “Property- or quantity-based instruments limit the use of a resource or entry into a market by setting a quantity limit on such use or giving a property rightsholder the right to exclude”. Also; “Such limits might involve spatial parcelling of a resource (‘private property’), limited permits to use a resource, or limits on entry into a market. They include property rights… ambient standards… emissions or effluent standards… and use or extraction licences”
At 368: “If a market failure involves asymmetric information, or obstacles to bargaining because of incomplete information, then an information disclosure instrument can help remedy the problem and facilitate more efficient market transactions.”
At 368: “If regulation of market transactions is not sufficient, government can acquire ownership of a resource or enterprise.”; “Government ownership may enable wider public access to resources with bureaucratic procedures that let resource values dissipate or become captured by private users.”
At 368: “Industry custom, trade association codes of conduct, professional codes, and group property management systems are all examples of mechanisms that private actors may adopt to regulate conduct.”
The authors consider where the theory meets evidence, that is “the empirical reality that much legislation has been enacted that yields diffuse general benefits while imposing costs on industry” (Wiener and Richman, 374). Other more complex models are also considered:
“More complex models of positive politics look at a wider set of actors on the demand side, such as advocacy grounds and political entrepreneurs, and also give attention to the supply side of legislation, such as legislators and regulators.” (Wiener and Richman, 376).
Different authors categorise the range of interventions in different ways. Freiberg noted:
“There is no generally accepted taxonomy of regulatory tools. While many classifications have been proffered, none is overwhelmingly persuasive.” (Freiberg, at 6).
For a different taxonomy see the discussion at 225 – 234 of DL Weimer and AR Vining Policy Analysis: Concepts and Practice (4th ed, Prentice Hall, 2005) at 225-234.
For practical guidance and how to go about the policy design process see Chapter 2 at page 8 of The Treasury RIA Handbook.
 Jonathan B Wiener and Barak D Richman “Mechanism Choice” in Daniel A. Farber and Anne Joseph O’Connell (eds) Research Handbook on Public Choice and Public Law (Edward Elgar 2010).
 Arie Freiberg “Restocking the Regulatory Tool kit” (Jerusalem Papers in Regulation & Governance WP No 15 2010).
A key part of intervention design is the choice of institution. There are a variety of public organisational forms to choose from including departments, various types of Crown entities or local government. For official guidance on the choice of organisational forms see http://www.ssc.govt.nz/node/4006.
In addition to public organisations, there is also the option of using market and civil society actors. Levi Faur discusses using market and civil society actors to undertake second and third party regulation as an alternative to first party regulator by public organisations. 
Levi Faur defines second party regulation as:
“In second party regulation, there is a social, political, economic and administrative division of labour between the actors and the regulator is independent and distinct from the regulatee.” (Faur, at 11)
One of the examples Levi Faur uses is where bigger businesses set the standards for other businesses.
Levi Faur defines third party regulation as:
“In third party regulation, the relations between the regulator and the regulatee are mediated by a third party that acts as independent or semi-independent regulatory-auditor. Processes and procedures of accreditation by third parties is a central enforcement strategy and "contractual relationship between firm and the party auditing the facility in place of relying solely on the regulatory agency as enforcer." (Faur, at 11)
 Levi Faur “Regulation & Regulatory Governance” (Jerusalem Papers in Regulation & Governance WP No 1 2010).