Anatomy of a Policy Area: The Case of Shippingqed.econ.queensu.ca/pub/cpp/Dec97/Pal.pdf · Anatomy of a Policy Area: The Case of Shipping 399 ... based on William Dunn’s methodology

Anatomy of a Policy Area: The Case of Shipping399

CANADIAN PUBLIC POLICY – ANALYSE DE POLITIQUES, VOL. XXIII , NO. 4 1997

Anatomy of a Policy Area:The Case of ShippingDANIEL HOSSEUS AND LESLIE A. PAL

School of Public AdministrationCarleton UniversityOttawa, Ontario

Les champs de la politique sont habituellement définis par un usage conventionnel plutôt que par une analysesystématique. Cet article est une expérience à l’aide de la méthodologie de l’analyse de frontière qui construitde façon inductive à travers l’aggrégation des “sujets” et des instruments de politique. À l’aide du cas de lapolitique en matière d’envois maritimes, nous développons une liste de 473 sujets tirés à partir de plusieursbibliographies et de sources politiques. Ceux-ci sont ensuite aggrégés dans 135 catégories avec les instru-ments de politique qui les accompagnent. Bien que la méthodologie ait des limites évidentes, l’expérienceet l’étude de cas démontrent qu’il est possible d’estimer systématiquement tous les contenus possible d’unchamp politique. Cette approche est suffisament prometteuse pour justifier des raffinements et des applica-tions plus larges à toute une gamme de problèmes d’estimation. Cet article est aussi une contribution modesteà l’étude des instruments politiques qui semble s’être arrêtée à l’étape de la “classification” sans beaucoupd’analyse empirique de l’application des catégories d’instruments.

Policy fields are usually defined by conventional usage rather than by systematic analysis. This paperexperiments with a methodology of boundary analysis that builds inductively through the aggregation of“topics” and policy instruments. Using the case of shipping policy, we develop a list of 473 topics drawnfrom several bibliographical and policy sources. These are then aggregated into 135 categories with accom-panying policy instruments. While the methodology has obvious limitations, the experiment and case studydemonstrate that it is possible to systematically estimate the complete possible contents of a policy field.The approach shows sufficient promise to warrant refinements and broader applications to a wide variety ofestimation problems. It also makes a modest contribution to the study of policy instruments, which hastended to get stuck at the “classification” stage without much detailed empirical analysis of how and wheninstrument categories actually apply.

Do policy analysts really know what they aretalking about? For example, how do we know

the difference between transportation policy and ag-ricultural policy, or between environmental policyand social policy? Pigeon-holing is probably one ofthe most fundamental aspects of any policy analy-sis, and yet also one of the least examined. Policyanalysis and policy development clearly cannot oc-

cur without some implicit boundaries and catego-ries that define a policy field. Wildavsky thought of“public policies as divided into sectors” and used aspatial metaphor to argue that these sectors couldbe “densely rather than lightly packed” (1979,p. 64). In practice, while there may be deductiveprinciples that help define a policy field (e.g., trans-portation policy self-evidently is about moving

400 Daniel Hosseus and Leslie A. Pal


people and things through space), most analysts relyon conventional understandings of core legislationand instruments (e.g., transportation policy is largelydefined by legislation with the word transportationin it).

When policy fields are stable or “lightly packed,”this conventional approach works reasonably well.As turbulence increases and as fields get morecrowded and overlapped, however, the need for pre-cision and more systematic definitions increases.Major changes in policy can have “second- andthird-order consequences that impair or change in-stitutions, patterns, and arrangements” in unan-ticipated ways (Dror 1971, p. 65). Policies are inti-mately linked: changing one will often affect an-other. To understand what we are doing, we need inthe first instance to know with what we are work-ing. Taking Wildavsky’s metaphor a step further, weneed to inventory the existing “policy space.” Forexample, the Canadian government has recentlybegun a review of the Canada Shipping Act. Thecentury-old act is archaic in structure and content,yet constitutes the core of the current policy frame-work for Canada’s shipping industry. Over the years,the Shipping Act has spawned a plethora of regula-tions and guidelines, and is referenced by other laws.Any overhaul of the Canada Shipping Act will haveramifications for all these laws, regulations, andguidelines. A systematic, comprehensive, and em-pirical inventory of the policy space would seem tobe a useful prerequisite for an overhaul of core leg-islation such as the Canada Shipping Act. Moreover,a generic inventory methodology could be appliedacross policy fields in any single jurisdiction, andacross the same policy field in different jurisdictions,to allow more precise comparisons of policy pro-files. A careful inventory of a policy space shouldmake for better policy analysis, policy making, andpolicy management, as well as a better generalunderstanding of a policy area.

This article explores one method of compiling acomprehensive, empirically based policy inventory

which can serve as a guide to a given policy space,based on William Dunn’s methodology of bound-ary analysis (Dunn 1994). The Canadian commer-cial marine industry and its governing policy regimewill serve as an illustration. We make several he-roic assumptions throughout, but despite the cru-dity of many aspects of the apparatus, we are con-vinced that the question to which it responds is bothvital and devilishly difficult: “If you had to com-pletely describe the contents of policy field X, howwould you do it?” This question — though not usu-ally put in as pointed a fashion — underlies a hugeamount of practical and academic policy analysis,from legislative histories to briefing books to schol-arly monographs. As we noted above, we also be-lieve that as policy fields become more crowded andpolicy making becomes more “horizontal” (Pal1997), some greater measure of precision in answer-ing this question becomes vital. We hope to showthat a policy inventory is a useful, albeit somewhatunwieldy, tool for both policy management andpolicy analysis. We emphasize that this paper is ex-ploratory, and as we are approaching the subjectfrom a methodological point of view, we hesitate tomake grand claims about its theoretical and practi-cal implications.

Finally, most readers will detect a whiff of ra-tionalism in this article; after all, we are urginggreater precision in analytical technique, and areoffering an application of one such technique. Therecent “argumentative turn” (Fischer and Forester1993; Stone 1988) in the policy sciences has em-phasized the social construction of policy problemsand hence, by extension, of policy fields. We haveno desire to reopen the tired and stale debate be-tween constructivists and rationalists. We admit toa cautious belief that better policy can arise (notmust arise) from better technique. Indeed, for ourmore constructivist-leaning readers, the methodol-ogy we propose can be viewed as a species of lin-guistic analysis that tries to make better sense ofthe words people use to describe what they are doing(Dryzek 1990).



POLICY INVENTORY

A policy inventory will be defined as a catalogue oftopics and their associated policy instruments.Schematically, an inventory looks like the table be-low. In theory, a complete inventory will provide alist of all possible topics related to a policy field,though if a government is not addressing that topic,the instrument category will be empty.

Policy Topic Policy Instrument

topic 1 policy instrument(s) for topic 1topic 2 policy instrument(s) for topic 2topic ... policy instrument(s) for topic ...topic n policy instrument(s) for topic n

Policy inventories represent a systematic gridwhich can be applied to policy regimes, i.e., thecollection of specific policies governing a policyarea,1 in other times or jurisdictions, and thus serveas a common basis for comparison. They ensure thatidentical topics are compared.

TopicsIdeally, a policy topic refers to a very narrow sub-category of a policy area. Topics are headers whichshould capture condition descriptions (which formthe basis of problem definitions2) as objectively aspossible. Every policy field consists of a myriad ofproblems or topics, and typically these are aggre-gated into broader (and more workable) categories.The point of an inventory, however, is to begin withthe smallest “policy datum” or topic, and aggregatein more systematic and transparent ways. Forexample, a policy topic in the shipping policy fieldmight be “air quality.” Air quality is a very narrowterm under which a set of condition statements canbe categorized (e.g., “the air in ships’ engine roomstends to cause headaches among crew members whowork there”). By keeping the term as neutral as pos-sible, a policy inventory’s list of topics can beapplied more readily as a grid or template for

comparisons across jurisdictions. Readers willimmediately object that the definition of topics willvary across jurisdictions. They do to some extent,but as we show below, there are strong similaritiesas well, in part perhaps because of policy diffusion,emulation, and the increasing salience of globalpolicy regimes.

Policy InstrumentsWhereas topics are visible terms embedded in thelinguistic field of a policy domain which can besearched for inductively, the policy instrument withwhich that topic is associated is a more abstract cat-egory, and we require some guidance from theexisting literature.

As a field of study, the analysis of policy instru-ments is itself relatively new. Salamon noted in 1981that “the systematic study of the techniques of gov-ernment action has hardly gotten off the ground”and that “most tools have hardly been scrutinized atall, and there has been a virtual absence of system-atic comparative work analyzing different tools orexamining the changing forms of action as a whole”(1981, pp. 262, 263). One exception was the workby Kirschen and colleagues in the early 1960s, build-ing on the 1953 remark by Dahl and Lindblom thata fundamental feature of contemporary governmentwas the proliferation of new techniques of socialaction (1953, p. 8). Kirschen identified five “fami-lies” of policy instruments: (1) public finance,(2) money and credit, (3) exchange-rate, (4) directcontrol, and (5) changes in institutional framework,which together accounted for 62 distinct instruments(Kirschen et al. 1964). Kirschen’s is easily the mostdetailed and finely grained schema of policy instru-ments, but there are five others that have attractedacademic attention.

Hood (1984) developed a typology of instrumentsbased on what he called the NATO scheme, stand-ing for the different resources that governments haveat their disposal to effect policy change. N standsfor nodality or information resources, A for



authority, T for treasure or money, and O for or-ganization or personnel. In Hood’s framework, gov-ernment is modeled as a cybernetic system, bothacting upon and reacting to its environment, muchlike a thermostat. Pal (1992, p. 143) combined theNATO schema with the Doern and Phidd (1992) ar-gument of a sliding scale of coercion, to present anextended inventory of policy instruments linked tospecific governmental resources and objectives.

McDonnell and Elmore (1987) proposed ascheme with four classes of instruments: (1) man-dates (rules governing action), (2) inducements(transfers of money), (3) capacity-building (trans-fer of money for the purpose of investment in mate-rial, intellectual, or human resources), and (4) sys-tem-changing (“transfer of official authority amongindividuals and agencies to alter the system by whichpublic goods and services are delivered”). As theysaw it,

policymakers face a discrete number of poten-tially powerful choices when they respond to apolicy problem. They can set rules, they can con-ditionally transfer money, they can invest in fu-ture capacity, or they can grant or withdraw au-thority to individuals and agencies. Each of theseoptions is expected to carry a particular effect —compliance, production, capacity, or authority.And each carries a package of benefits and costsfor different actors. (p. 12)

In an important article, Linder and Peters (1989)reviewed the various instrument schemes, and de-termined that several basic classes appeared repeat-edly: “1) direct provision 2) subsidy 3) tax 4) con-tract 5) authority 6) regulation (the only consensusclass), and 7) exhortation” (p. 44). Actual policymakers provided this list of instruments: cash grant,government-sponsored enterprise, in-kind transfer,loan guarantee, “tax break,” fee/charge, certification/screening, government provision, fine, administeredcontract, quota, prohibition, quality standard,“jawboning,” public promotion, information/demonstration, procedural “guideline,” insurance,

loan, licence/permit, price control, public invest-ment, franchise (ibid., p. 53).

Schneider and Ingram (1993) took a behaviouralapproach to the analysis of instruments, on the as-sumption that “public policy almost always attemptsto get people to do things that they might not other-wise do; or it enables people to do things that theymight not have done otherwise” (p. 513). They iden-tified five classes of instruments. Authority tools are“statements backed by the legitimate authority ofgovernment that grant permission, prohibit, or re-quire action under designated circumstances”(p. 514). Incentive tools “rely on tangible payoffs,positive or negative, to induce compliance or en-courage utilization” (p. 515). Capacity tools “pro-vide information, training, education, and resourcesto enable individuals, groups, or agencies to makedecisions or carry out activities” (p. 517). Symbolicand hortatory tools “assume that people are moti-vated from within and decide whether or not to takepolicy-related actions on the basis of their beliefsand values” (p. 519). Learning tools are used “whenthe basis upon which target populations might bemoved to take problem-solving action is unknownor uncertain” (p. 521).

Doern and Phidd (1992, p. 97) argued that thereare five broad categories of policy instruments:(1) self-regulation, (2) exhortation, (3) expenditure,(4) regulation (including taxation), and (5) publicownership. This typology assumed that as one movesfrom the first category to the last, one moves alonga continuum of legitimate coercion. Their argumentwas that all governance in a liberal democracy in-volves some degree of imposition or coercion, andthat politicians generally prefer to use the least co-ercive instrument possible. Within these broad cat-egories, Doern and Phidd identified as many as 26finer “graduations of choice” such as grants andsubsidies, guidelines, and speeches (p. 112).

This brief review of the dominant typologies ofpolicy instruments demonstrates that, despite dif-ferent approaches and levels of detail, there is a



modest consensus that the core instrument catego-ries appear to be exhortation (information), expen-ditures, regulation (including self-regulation), andtaxes. We will use these categories for the inven-tory. The review also shows, however, that the studyof instruments has been somewhat stagnant in re-cent years. Several typologies have been offered,based on different approaches, but there has beenrelatively little effort expended in actually applyingthe categories in some systematic way in order tounderstand what is going on in a policy field. Ourinventory methodology does precisely this.

In the policy inventory, then, for the topic “pilot-age,” for example, the entry for policy instrumentmight read: regulation, Canada Pilotage Act. Forsome topics governments will have chosen to donothing. The entry in the policy instrument columnmight then read: laissez-faire or deliberate inaction.There might also be topics of which governmentsare not aware, i.e., they will not have deliberated onthem and not even the no-action policy instrumentwill have been applied. The policy instrument entrycould then read “topic not considered.”

CREATING A POLICY INVENTORY

A major challenge is to ensure completeness of thecatalogue of topics. How can one be certain that nocrucial element has been ignored? Research meth-ods from the field of history seemed useful at first.Historians describe and analyze the nature of socie-ties for which they have no first-hand experience andfor which only limited sources are available. Yet suchanalysis requires that the description of the studiedsociety is reasonably complete. Wiedman (1986,p. xii) outlines the historical method’s four elements:

• the collection of the surviving objects and of theprinted, written, and oral materials that may berelevant,

• the exclusion of those materials (or parts thereof)that are unauthentic,

• the extraction from the authentic material of tes-timony that is credible, and

• the organization of that reliable testimony intoa meaningful narrative or exposition.

The method really refers to steps in a selection proc-ess and selection is classification. Lipset andHofstadter (1968) note:

The way in which one looks at data depends inlarge measure on the questions asked, the theoryemployed, and the classifications used. It is ob-vious that in dealing with complex alternativeinterpretations, no matter how rigorous the meth-odology employed, most historians and sociolo-gists basically present an argument which theythen “validate” by showing that there are morepositive than negative data available. (p. 50)

To postulate historical theories, one must have firstclassified relevant data. How can one ensure thatthe system of classification is complete? Similarly,anthropological methods assume the existence of aninitial frame of reference, a set of categories fromwhich the researcher begins. Trigger (1968) alludedto a method that seemed promising. He describedhow prehistorians assemble an “inventory of traits”to make conjectures about the culture upon whichthey have happened. However, the inventory of traitsassumed a set of categories in which artifacts canbe slotted. In the end, these categories rely on thereasoned creativity of the researcher (see Childe1956, pp. 124-31). The completeness of the catalogwould only be assured through trial and error, byapplying and revising categories several times.

Boundary AnalysisWilliam Dunn, in Public Policy Analysis (1994),proposes an interesting and potentially more effi-cient method to approximate a complete set of top-ics or problems. For Dunn, boundary analysis is amethod to structure so-called metaproblems, i.e.,“problem-of-problems” that “[are] ill-structuredbecause the “domain of problem representations



held by diverse stakeholders seems unmanageablyhuge” (p. 148; see also Dror 1971, p. 63). However,the method seems equally appropriate for problemsthat are moderately or even well-structured. Dunnoutlines three steps: saturation sampling, elicitationof problem representations, and then boundary es-timation (ibid., p. 164). Saturation sampling intendsto ensure that all stakeholders are identified fromwhom “ideas, basic paradigms, dominant metaphors,standard operating procedures, or whatever else wechoose to call the systems of interpretation by whichwe attach meaning to events” (ibid.) are elicited.These views, opinions, and ideas are subjected toboundary estimation. Boundary estimations are aform of content analysis: “The analyst constructs acumulative frequency distr ibution wherestakeholders are arrayed on the horizontal axis andthe number of new problem elements—ideas, con-cepts, variables, assumptions, objectives, policies—are plotted on the vertical axis” (ibid., see Figure 1).As the number of new elements declines, the curverepresenting “additions” begins to flatten out, untilfurther sampling contributes virtually nothing “new.”At this point, one has estimated the “boundary” oftopics in the field.

Dunn’s method can be adapted for policy inven-tories. The array of problems which defines theboundary can serve as the foundation of the policyinventory. It represents the catalogue of topics. Asecond step is to assemble the policy instrumentsassociated with the catalogue of topics.

This method can also be applied to text sources,whether they be “scientific literature, library hold-ings, languages, literary works, consumer prefer-ences” (Dunn 1994, p. 165) either for a given pointin time or for a time period.

Assembling the List of Associated PolicyInstrumentsOnce a list of topics has been established, a list ofassociated policy instruments can be assembled. Theprocess is straightforward. A survey of acts and regu-lations pertaining to each topic would cover the moreintrusive policy instruments, i.e., expenditure, taxa-tion, regulation, and public enterprise. To incorpo-rate the less intrusive policy instruments, i.e.,laissez-faire, exhortation and symbolic policy out-puts, speeches, news releases, budgets, and othergovernment documents would have to be examinedfor evidence of government activity regarding eachtopic. Budgets would, for example, provide infor-mation on government programs that exhort the pub-lic toward a goal or generate symbolic policy out-puts. Speeches and news releases could give indi-cations as to why a government has chosen a courseof inaction.

THE CASE OF SHIPPING POLICY

The concept of a policy inventory and the associ-ated methodology will be applied to shipping policy.From a deductive or linguistic approach, shippingpolicy refers to the displacement of people andgoods using waterborne vehicles. Contemporary Ca-nadian public discourse uses the term “marinepolicy” to label this field (see, e.g., Transport Canada1995). However, the term marine policy is often usedto include issues such as fisheries management or

FIGURE 1Boundaries of a Metaproblem

Source: Dunn 1994, p. 165.

25

20

15

10

5

0 2 4 6 8 10

Boundaries ofMetaproblem

NEW CONCEPTS

POLICY STAKEHOLDERS



deep-sea mining (see, for example, Hennessey1981). Bibliographic indexes usually employ theheader “shipping policy” for the policy area to beexamined here. We follow that convention, andexclude natural resource management from theanalysis.

Boundary AnalysisTo obtain a complete set of “ideas, concepts, vari-ables, assumptions, objectives, policies” (Dunn1994, p. 164) — the list of topics for the policy in-ventory — the subject indexes of several bibliogra-phies were surveyed. A bibliography is an “inclu-sive, continuing record of all contributions to theliterature” (Freides 1973, p. 139). As Freides notes,bibliographies can provide researchers with the“main outlines of a topic” (ibid.). They usually con-tain a subject index which focuses “on only one di-mension of the work, the subject matter of thedata...” (ibid., p. 143). A policy inventory’s list oftopics is a list of subjects, thus a survey of generaland specific bibliographies should provide a com-plete list.

Six indexes were consulted for this analysis. Theyare the

• ABI/Inform (ABI), a periodical index,

• Canadian Research Index (CRI), a periodicalindex,

• Library of Congress Subject Headings (LC),

• Public Affairs Information Service Bulletin(PAIS), a periodical index,

• Transport Canada Library’s catalogue (TC), and

• WINSPIRS TRANSPORT (TRAN), a periodi-cal index on transportation.

The four periodical indexes are available on CD-ROM. These sources offer significant advantages,but also require some qualifications. Because a CD-

ROM issue covers numerous years (e.g., PAIS be-gins in 1972), its subject index provides longer rangecoverage and thus a more substantial list of topics.Moreover, the Library of Congress Subject Head-ings and the Transport Canada Library catalogue are“timeless” as they are intended to cover those li-braries’ entire holdings. The library catalogues helpavoid missing a topic because it was not addressedin a given time period. However, they do introducea spatial bias toward North America. On the otherhand, PAIS covers material in all forms — books,periodical articles, government documents, and awide range of pamphlets and reports by a large andvaried group of commercial and voluntary agencies— published anywhere in the world in the Englishlanguage (Freides 1973, p. 167). Also, TRAN is pro-duced by the Organisation for Economic Co-operation and Development (OECD), the US Trans-portation Research Branch, and the European Con-ference of Ministers of Transport, and covers docu-ments published in those jurisdictions.

The thesauri and dictionaries of the periodicalindexes and the TC and LC listings were scouredfor sub-headers as well as the narrower, broader, andrelated terms of the search keys “shipping,” “ships,”“transportation,” and “navigation.” This choice ofsearch keys was intended to provide a full listing ofall the topics pertaining to shipping, including top-ics that are related to the policy field in a peripheralway. For example, shipping companies in Europeare embracing the concept of “door-to-door” serv-ice. Under this concept, the shipping company isresponsible not just for the delivery of goods fromquay to quay but also for bringing the cargo to thequay from its point of origin and moving it to itsfinal destination. The infrastructure required by theshipping industry thus extends beyond port facili-ties and potentially involves railway connections andtrucking. Such topics can have implications for pub-lic policy, and may not be adequately covered bythe term “shipping.” “Transportation” was thus in-cluded to incorporate a broader view of the indus-try. Similarly, the term shipping might notadequately cover some policies pertaining to the



operation of ships, such as construction regulationsfor ships. To include these elements, the terms“ships” and “navigation” were also used as searchterms. The four terms represent a progression ofspecificity — navigation, ships, shipping, transpor-tation — and move from lower to higher levels ofabstraction (Sartori 1977, p. 40).

Several pragmatic strategies were used to makethe analysis manageable. The terms were appliedonly in English, thus limiting international cover-age to some degree (Freides 1973, p. 142). How-ever, given the predominance of English in academeand commerce, this constraint should not unreason-ably bias the search results. The search was also lim-ited to the first level of sub-headers. For example,for the entry “shipping — communication systems— financial aspects,” only “shipping — communi-cation systems” was used. This restriction was in-tended to limit the size of the data file. However,many second-order entries were captured by first-order entries elsewhere. In the example above,financial aspects were be covered by “shipping —finance.” Similarly, first-level sub-headers to geo-graphic areas or jurisdictions such as “Transporta-tion — Canada” or “Transportation — EuropeanEconomic Community” were skipped. Finally, first-level references to bibliographic works, e.g., “Ships— Bibliography,” were also omitted.

Even with these constraints, an initial list of 692potential topics was generated. However, this listincluded topics that clearly have no relation to ship-ping policy. For example, the high-level search term“transportation” generated topics pertaining to bi-cycles and urban transport. The removal of redun-dant terms, of hyphenation (e.g., decision-makingand decision making), of typos, and of terms in thesingular for which plurals are present (e.g., buoyand buoys, though not operation and operations)yielded a list of 473 different topics (see AppendixA). The list includes entries which may be very simi-lar to others in the list, e.g., routes and routing.However, routes refers to the trip trajectories thatcould be used whereas routing refers to trip

trajectories that are used by shipping companies.These nuances reflect different ways of looking at aproblem. They are “alternative problem representa-tions” provided by the sources, the elicitation ofwhich is precisely the purpose of boundary analy-sis (Dunn 1994, p. 164).

The topics listed are non-duplicate problem rep-resentations. The list of topics generated providesan outline of the policy field by topic. It representsthe boundaries of the “policy space” of shippingpolicy. Using a database program, it was possibleto determine in which database each “new and non-duplicative element” first appears. The results areplotted in Figure 2.

Figure 2 approximates the shape of the curve inthe theoretical illustration provided by Dunn. How-ever, a flat part indicating a complete set of prob-lem representations was not attained. More sourceswould likely have to be consulted to reach the outerlimit. 3

FIGURE 2Boundaries of Shipping Policy

500

400

300

200

100

0

TRAN LC CRI PAIS ABI TCSource

Hits



Classifying the TopicsWith 473 entries, the list of topics is too unwieldyto be useful. An analysis of shipping policy basedon this inventory would entail determining how eachpolicy instrument (and as will be seen, their numberis substantial, too) pertains to each of the 473 top-ics. As Rose notes, the “more distinctions made, themore complex the resulting typology — and themore improbable it is that social scientists will everbe able to find data to fill all the resulting cells”(1973, p. 89). The principal purpose of a policy in-ventory, however, is to give analysts and policymakers an analytic tool that can contribute effec-tively to research and decision making. The list oftopics must be narrowed through classification.

Classification categories should fulfill four re-quirements: exhaustiveness, disjointness, consist-ency, and hierarchical distinctiveness (Dunn 1994,p. 166; see also O’Shaughnessy 1972, p. 26).Exhaustiveness means that all subjects in a policyarea should be covered. Disjointness means that thecategories are mutually exclusive. Consistency re-quires that only one classificational principle pro-vides the basis for each category. Hierarchical dis-tinctiveness refers to the idea that the “meaning oflevels in a classification system must be carefullydistinguished” (Dunn 1994, p. 167). While theseprinciples provide good general guidelines, one can-not always fully adhere to them. For example, Webernotes that “... having one attribute does not logi-cally exclude the possession of a second attribute”(1985, p. 34). The topic of dangerous goods pro-vides a good example. Dangerous goods rules andregulations cover the handling of goods that can posea threat to the individuals handling them, the vehi-cles used for their transport, as well as the environ-ment. In a higher level classification system, thetopic dangerous goods could thus be classified as asafety topic or as an environmental topic. Danger-ous goods actually falls under both; however, thiswould compromise the disjointness principle. Thedecision on how to classify a particular topic thusdepends on the purposes of the classification and/or the bias of the analyst (O’Shaughnessy 1972,

p. 25; see also Dunn 1994, p. 166; Kingdon 1984,p. 117; and Rose 1973, p. 89). It also requires sub-stantive knowledge of the policy area to appreciatethe nuances of different terms generated throughboundary analysis (see O’Shaughnessy 1972, p. 25).

To create a narrower list of topics, the topics ap-pearing most frequently in the six sources were tal-lied. They are listed in Appendix B. These gener-ally low-level topics were reclassified. For exam-ple, accounting was subsumed under finance, portsunder infrastructure, data processing under commu-nications, ice under climatic factors, etc.4 This gen-erated the final, revised list presented in AppendixC. This list represents a more manageable grid thatcan be applied to the shipping policy area.

Policy InstrumentsThe compilation of policy instruments proved to berelatively easy. The Transport Canada World WideWeb site5 offers a list and the full text of acts andregulations for which the minister of transport isresponsible or partially responsible. It also providesseveral of the minister’s speeches, departmentalnews releases, the Speech from the Throne, the 1996budget, and documents pertaining to safety stand-ards and voluntary guidelines. In short, it is a re-pository of documents covering the entire range ofpolicy instruments pertaining to that department. Forillustrative purposes, the discussion of policy inven-tories and shipping policy will be limited to thissource and only to the acts, regulations, and volun-tary guidelines listed there. The list of regulationson the website is not complete yet. Also, the Cana-dian Coast Guard is now under the jurisdiction ofthe Department of Fisheries and Oceans. The CoastGuard is responsible, among other things, for theprovision of icebreaking and aids to navigation. Thussome of the policy instruments relating to shippingpolicy may not be identified in this paper.

Appendix D represents the policy inventory gen-erated through the methodology outlined above. Itlists the acts, regulations, and voluntary guidelines,i.e., the policy instruments employed to address the



topics. It was unavoidable to attach a second topicto the classification of policy instruments, i.e., it wasnecessary to infringe on the classification principleof disjointness. For example, the regulations per-taining to radio communications under the ArcticWaters Pollution Prevention Act in the first instancerefer to pollution prevention. However, they alsopertain to communications. Moreover, the radiocommunications regulations do not have the samedegree of “environmental content” as, for example,oil pollution prevention measures, which deal onlywith oil pollution. A second topic supplies morenuanced information. Also included in the list is anindication of the nature of the policy instrument. Theclassification of type of policy instruments is fairlyunambiguous. Voluntary guidelines and standardsare classified as exhortations/symbolic outputs be-cause adherence to them is voluntary (though thereis an implicit threat that non-adherence will resultin binding regulations). Items that provide for thecollection of fees and tariffs fall under the headingtaxation. All non-fee, non-tariff regulations fall un-der the heading regulations.

Not included in Appendix D are policy instru-ments found in documents such as ministers’speeches, departmental news releases, budgets, thatis, documents that would provide information on theless intrusive policy instruments. It was found thatthe classification of such documents would requirea detailed content analysis of each of the items basedon the narrow list of topics. To validly analyze thecontent of each document would require an effortwell beyond the scope of this article.6 In this con-text it must be noted that the topic classification ofthe acts, regulations, and voluntary guidelines andstandards presented in Appendix D is based on thetitles of these items as well as the authors’ judge-ments. It could well be that a close reading of all ofthese documents would result in a different classi-fication. This represents a serious problem and hasimplications for the usefulness of policy inventories.

A policy inventory — even one as incomplete asours — might serve to point out gaps in existing

and proposed policy, and act as a guide for futurepolicy activity and as a basis for comparisons acrosstime and jurisdictions. For example, a crude indica-tion of the degree of “coverage” of any topic can beachieved by tallying its relevant policy instruments.A graph can than provide a visual indication of gaps.Appendix E demonstrates this. The valleys repre-sent topics in which few or no policy instrumentshave been applied, the peaks indicate topics withmore activity. Four principal areas of activity canbe discerned in the case of shipping policy: safetyand health, registration/classification/standards, in-frastructure, and environment. Lesser areas of ac-tivity deal with manning, finance, equipment andsupplies, communications, coastwise shipping(cabotage), and accidents. No activity was detectedaround issues such as taxation, automation, and en-ergy. A complete and more in-depth survey of alldocuments from government departments (e.g., in-cluding the Department of Fisheries and Oceans),and of symbolic outputs, might yield different re-sults. However, as is, the valleys do show where thepolicy community may (or may not) want to becomeactive in the future to address legislative and regu-latory shortcomings. Alternatively, peaks may indi-cate areas in which too much legislation and regu-lation limits the performance of the policy frame-work, a situation which the policy community maywant to redress.7

CONCLUSIONS

The daunting question posed at the outset of thisarticle was “If you had to completely describe thecontents of policy field X, how would you do it?”For example, if you had to reform Canada’s marinepolicy, what acts, regulations, and issues would youneed to consider during the reform process?

The methodology has been successful in severalregards. First, the methodology of boundary analy-sis has generated a broad list of topics which listsall or virtually all elements of the policy area. Assuch, it provides a good overview of the policy area.



In terms of the definition of public policy, it liststhe various elements in which conditions can beobserved, and for which preferred alternatives canbe formulated by governments. Second, the meth-odology also generated a substantial list of policyinstruments that have been used by the Canadiangovernment for marine transportation. By listingtopics and instruments, the methodology has gener-ated a good overview of the anatomy of the ship-ping policy area. Third, we hope that this approachhas taken the study of policy instruments forwardin a modest, empirical fashion. The literature to datehas been rich in theory but poor in application. Wethink that the inventory approach makes an explicitconnection between policy content and policy-instrument categories. It provides a way of portray-ing, with some greater confidence, the key instru-ments used in a given policy domain.

The limitations of our experiment are also pain-fully evident. It takes enormous time and resources,relies on judgements and hunches throughout, andto be truly thorough would require countless hoursof careful content analysis. The list of policy instru-ments, for example, is biased in favour of the moreintrusive policy elements because a content analy-sis of sources which would reveal information ofless intrusive policy instruments was not attempted.Ironically, the best inventories will be compiled bypeople who already know the policy field well —that knowledge will help them make the judgementcalls upon which complex classification schemeslike this depend.

However, our experiment does demonstrate thatthere are ways to completely describe a policy field.At the very least, the illustration we have presentedreminds us that it is important to map out policyfields as clearly as we can, so that policy analysts,policy managers, and policy makers, as well as thegeneral public can understand the full scope of gov-ernmental activity, its intensity and character, andthe possible consequences of policy reform.

NOTES

1There is no difference between a policy area and apolicy field, and the two terms can be used as synonyms.

2A problem can be defined as the discrepancy betweena condition and a preferred alternative. This definition isbased on a survey of definitions of policy problems givenby Pal (1992), Brewer and Deleon (1983), Dunn (1981),and Dery (1984).

3It is not clear though that a completely flat slope couldbe obtained. Given the versatility of the English languagethere will always be words that could be used by othersources to describe an exactly identical problemrepresentation.

4These reclassifications were based on substantiveknowledge of the policy area gained by one of the au-thors (Hosseus). The list generated in this fashion wascompared to a list that the author had casually assembledbased on newspaper articles and conversations during awork term with a German shipping company.

5http://www.tc.gc.ca

6Manheim and Rich, in their discussion of contentanalysis, suggest, for example, using numerous contentcoders (“as many as possible”) to ensure coding reliabil-ity (1995, p. 196).

7The table (Appendix D) on which the graph is basedalso illustrates the task the reformers of the Canada Ship-ping Act have at hand. Regulations flowing from that actcover topics such as distressed seamen, pilotage ladders,nominal horsepower computing methods, and shipfumigation.

REFERENCES

Brewer, G.D. and P. Deleon (1983), The Foundations ofPolicy Analysis (Homewood, IL: The Dorsey Press).

Childe, G.V. (1956), Piecing Together the Past: The In-terpretat ion of Archaeological Data (London:Routledge & Kegan Paul).

Dahl, R.A. and C.E. Lindblom (1953), Politics, Econom-ics and Welfare (New York: Harper).

Dery, D. (1984), Problem Definition in Policy Analysis(Lawrence, KA: University Press of Kansas).

Dror, Y. (1971), Design for Policy Sciences (New York:Elsevier).



Dryzek, J.S. (1990), Discursive Democracy: Politics,Policy and Political Science (Cambridge: CambridgeUniversity Press).

Dunn, W. (1981), Public Policy Analysis: An Introduc-tion (Englewood Cliffs: Prentice Hall).

____ (1994), Public Policy Analysis: An Introduction, 2nded. (Englewood Cliffs: Prentice Hall).

Doern, B. and R. Phidd (1983), Canadian Public Policy:Ideas, Structure, and Process (Toronto: Methuen).

Fischer, F. and J. Forester, ed. (1993), The Argumenta-tive Turn in Policy Analysis and Planning (Durham,NC: Duke University Press).

Freides, T. (1973), Literature and Bibliography of theSocial Sciences (Los Angeles: Melville).

Hennessey, T.M (1981), “Evaluating Marine Policy: Cri-teria from Two Models and a Comparative Study,” inComparative Marine Policy: Perspectives from Eu-rope, Scandinavia, Canada, and the United States(New York: Praeger).

Hood, C. (1984), The Tools of Government (London:Macmillan).

Kingdon, J.W. (1984), Agendas, Alternatives, and PublicPolicies (Toronto: Little, Brown and Company).

Kirschen, E.S. et al. (1964), Economic Policy in Our Time,3 vols (Amsterdam: North Holland).

Linder, S. and G.B. Peters (1989), “Instruments of Gov-ernment: Perceptions and Contexts,” Journal of Pub-lic Policy 9:35-58.

Lipset, S.M. and R. Hofstadter, ed. (1968), Sociology andHistory: Methods (New York: Basic Books, Inc.).

Manheim, J.B. and R.C. Rich (1995), Empirical PoliticalAnalysis: Research Methods in Political Science(White Plains, NY: Longman Publishers).

McDonnell, L.M. and R.F. Elmore (1987), AlternativePolicy Instruments (Santa Monica: Center for Public

Research in Education).O’Shaughnessy, J. (1972), Inquiry and Decision (Lon-

don: George Allen & Unwin Ltd.).Pal, L.A. (1992), Public Policy Analysis: An Introduc-

tion. 2nd ed. (Scarborough, ON: Nelson Canada).____ (1997), Beyond Policy Analysis: Public Issue Man-

agement in Turbulent Times (Toronto: ITP Nelson).Rose, R., ed. (1973), The Dynamics of Public Policy: A

Comparative Analysis (Beverly Hills: Sage Publica-tions).

Salamon, L.M. (1981), “Rethinking Public Management,”Public Policy 29:255-75.

Sartori, G. (1977), “Concept Misinformation in Compara-tive Politics,” in Comparative Politics: Notes andReadings, ed. R.C. Macridis and B.E. Brown, 5th ed.(Homewood, IL: The Dorsey Press).

Schneider, A. and H. Ingram (1993), “Social Construc-tion of Target Populations: Implications for Politicsand Policy,” American Political Science Review87:334-47.

Stone, D.A. (1988), Policy Paradox and Political Reason(New York: Harper Collins).

Transport Canada (1995), National Marine Policy (Ot-tawa: Transport Canada).

Trigger, B.G. (1968), Beyond History: The Methods ofPrehistory (New York: Holt).

Weber, R.P. (1985), Basic Content Analysis (BeverlyHills: Sage Publications).

Wiedman, D., ed. (1986), Ethnohistory: A Researcher’sGuide. Studies in Third World Societies, no. 35(Williamsburg, VA: College of William and Mary,Department of Anthropology).

Wildavsky, A. (1979), Speaking Truth to Power: The Artand Craft of Policy Analysis (Boston: Little, Brownand Co.).



APPENDIX AL IST OF TOPICS*

abandoning ofactaeronauticsaids to navigationallowanceand pierastronauticsautomationautomotiveberthbuildersbuildingsbuoyscareerscathodic protectioncharacteristicsclassificationcoastwise shippingcollisions at seacommon carriersconferencesconsolidationcontainerizationcontracts, maritimeconvoyscorporationscostsdata processingdecksdepartmentderegulationdeveloping countriesdisadvantageddivisiondomainearth stations

educationefficiencyelectronic equipmentemergency employeeengineeringenvironmental aspectsfacilitiesfendersfinancingfisheries navigationfoulingfuelfumigationgovernment aidguideshazardshighwayhullshydrographyhydrostaticsimpact forcesindustriesinformationinformation systeminsulationinterioriron and steellegislationliabilitylighthousesload linelossesmaintenance and repairmaneuverabilitymanningmaneuvering

maritimemathematical modelsmedicinemodelsmovementnautical paraphernalianetworkof disabled personsofficialsoperationoperatorsoptimum ship routingpaintingpassenger listsperiodicalspilotspneumatic equipmentservicespriceproducingproductivitynavigationradar in navigationrailroad transportationratesratproof constructionreformregistrationregulationrepair facilitiesreplacementroutesrouting and schedulingrules of classificationrural transportationsafety

sailing cardssatellitesscrappingsectorship handlingshippingshuttlessignals and signalingsimulatorssoundproofingspeedsponsoredstate aidstorage and movingstranding of shipsstudiesstudy and teachingsubsidiessurveyssurvivabilitytaxationtechniquestelecommunicationstest centers — USto transittransfers to foreigntransit time economicstransportationtruckingunderwater navigationunitized cargouser fees

vehicles

* This list has been shortened considerably for publication. A complete list is available from the authors on request.



APPENDIX BL IST OF MOST FREQUENTLY APPEARING

TOPICS

Number ofAppearances Topic

10 costs8 energy8 finance7 rates7 research7 safety measures6 accidents6 economic aspects6 environmental aspects6 statistics5 communications5 freight5 fuel5 ice5 management4 automation4 data processing4 equipment4 equipment and supplies4 hydrodynamics4 law and legislation4 mathematical models4 planning4 related4 study and teaching3 accounting3 electronic equipment3 fares3 information services3 maintenance and repair3 manning3 noise3 pilots3 ports3 railroads3 regulation3 safety regulations3 security measures3 taxation3 technological innovations3 transportation

APPENDIX CFINAL , REVISED L IST OF TOPICS

abandoning of ships/scrappingaccidentsautomationclimatic factorscoastwise shipping (cabotage)communicationsenergyenvironmentequipment and suppliesfinancefreightinfrastructuremaintenance and repairmanningplanningratesregistration/classification/standardsresearch/innovationsafety and healthsecurityshipbuildingstatisticssubsidiestaxation



AP

PE

ND

IX D

: TH

E P

OLI

CY

IN

VE

NTO

RY

1

Prim

ary

Topi

cSe

cond

ary

Topi

cPo

licy

Inst

rum

ent

Act2

Type

of I

nstru

men

t

acci

dent

sNo

neCo

llisi

on R

egul

atio

nsCS

ARe

gula

tion

acci

dent

sNo

neSh

ippi

ng C

asua

lties

App

eal R

ules

CSA

Regu

latio

nac

cide

nts

None

Ship

ping

Inqu

iries

and

Inve

stig

atio

ns R

ules

CSA

Regu

latio

nco

astw

ise s

hipp

ing

(cab

otag

e)sa

fety

and

hea

lthNo

n-Ca

nadi

an S

hips

Saf

ety

Orde

rCS

ARe

gula

tion

com

mun

icat

ions

infra

stru

ctur

eVe

ssel

Tra

ffic

Serv

ices

Zon

es R

egul

atio

nsCS

ARe

gula

tion

com

mun

icat

ions

None

VHF

Radi

otel

epho

ne P

ract

ices

and

Pro

cedu

res

Regu

latio

nsCS

ARe

gula

tion

envi

ronm

ent

com

mun

icat

ions

Ship

Sta

tion

Radi

o Re

gula

tions

AWPP

ARe

gula

tion

envi

ronm

ent

equi

pmen

t and

sup

plie

sSt

eerin

g Ap

plia

nces

and

Equ

ipm

ent R

egul

atio

nsAW

PPA

Regu

latio

nen

viro

nmen

tNo

neAi

r Po

llutio

n Re

gula

tions

CSA

Regu

latio

nen

viro

nmen

tNo

neGa

rbag

e Po

llutio

n Pr

even

tion

Regu

latio

nsCS

ARe

gula

tion

envi

ronm

ent

None

Grea

t Lak

es S

ewag

e Po

llutio

n Pr

even

tion

Regu

latio

nsCS

ARe

gula

tion

envi

ronm

ent

None

Non-

Plea

sure

Cra

ft Se

wag

e Po

llutio

n Pr

even

tion

Regu

latio

nsCS

ARe

gula

tion

envi

ronm

ent

None

Pollu

tant

Sub

stan

ces

Regu

latio

nsCS

ARe

gula

tion

envi

ronm

ent

None

Volu

ntar

y Gu

idel

ines

For

The

Con

trol O

f Bal

last

Wat

er D

isch

arge

sSS

IEx

horta

tion/

Sym

bolic

Out

put

envi

ronm

ent

safe

ty a

nd h

ealth

Ship

ping

Saf

ety

Cont

rol Z

ones

Ord

erAW

PPA

Regu

latio

nen

viro

nmen

tsa

fety

and

hea

lthSh

ips’

Dec

k W

atch

Reg

ulat

ions

AWPP

ARe

gula

tion

equi

pmen

t and

sup

plie

sNo

neSh

ip S

tatio

n Te

chni

cal R

egul

atio

nsCS

ARe

gula

tion

equi

pmen

t and

sup

plie

sNo

neSh

ips’

Ele

vato

r Re

gula

tions

CSA

Regu

latio

nfin

ance

com

mun

icat

ions

Ship

Rad

io In

spec

tion

Fees

Reg

ulat

ions

CSA

Regu

latio

nfin

ance

envi

ronm

ent

Ship

-Sou

rce

Oil P

ollu

tion

Fund

Reg

ulat

ions

CSA

Regu

latio

nfin

ance

infra

stru

ctur

eGr

eat L

akes

Pilo

tage

Tar

iff R

egul

atio

nsPA

Taxa

tion

finan

cein

frast

ruct

ure

Laur

entia

n Pi

lota

ge A

utho

rity

Dist

rict N

o. 3

Reg

ulat

ions

PARe

gula

tion

finan

cein

frast

ruct

ure

Port

War

dens

Tar

iffCS

ATa

xatio

nfin

ance

None

Boar

d of

Ste

amsh

ip In

spec

tion

Scal

e of

Fee

sCS

ARe

gula

tion

finan

cere

gist

ratio

n/cl

assi

ficat

ion/

stan

dard

sSh

ips

Regi

stry

Fee

s Ta

riff

CSA

Taxa

tion

freig

htNo

neGr

ain

Carg

o Re

gula

tions

CSA

Regu

latio

nfre

ight

None

Stea

msh

ips

Carr

ying

Car

go C

onta

iner

s Or

der

CSA

Regu

latio

nin

frast

ruct

ure

envi

ronm

ent

Shor

e Tr

affic

Reg

ulat

ions

SLSA

ARe

gula

tion

infra

stru

ctur

eNo

neBu

rling

ton

Cana

l Reg

ulat

ions

CSA

Regu

latio

nin

frast

ruct

ure

None

Ferr

y Ca

ble

Regu

latio

nsNW

PARe

gula

tion

infra

stru

ctur

eNo

neGe

nera

l Pilo

tage

Reg

ulat

ions

PARe

gula

tion

infra

stru

ctur

eNo

neGo

vern

men

t Pro

perty

Tra

ffic

Regu

latio

nsGP

TARe

gula

tion

infra

stru

ctur

eNo

neLa

uren

tian

Pilo

tage

Aut

horit

y Re

gula

tions

PARe

gula

tion

infra

stru

ctur

eNo

neM

inor

Wat

ers

Orde

rCS

ARe

gula

tion

infra

stru

ctur

eNo

neNa

viga

ble

Wat

ers

Wor

ks R

egul

atio

ns (

upda

ted

Nove

mbe

r 19

95)

NWPA

Regu

latio

nin

frast

ruct

ure

None

Paci

fic P

ilota

ge R

egul

atio

nsPA

Regu

latio

n

... c

ontin

ued



infra

stru

ctur

eNo

nePr

ivat

e Bu

oy R

egul

atio

nsCS

ARe

gula

tion

infra

stru

ctur

eNo

neSt

. Law

renc

e W

ater

way

Mar

ine

Traf

fic R

egul

atio

nsCS

ARe

gula

tion

infra

stru

ctur

era

tes

St. L

awre

nce

Seaw

ay W

harfa

ge a

nd S

tora

ge C

harg

es T

ariff

SLSA

ATa

xatio

nm

anni

ngNo

neCe

rtific

atio

n of

Abl

e Se

amen

Reg

ulat

ions

CSA

Regu

latio

nm

anni

ngsa

fety

and

hea

lthCe

rtific

atio

n of

Shi

ps’ C

ooks

Reg

ulat

ions

, Par

t I a

nd P

art I

ICS

ARe

gula

tion

man

ning

safe

ty a

nd h

ealth

Med

ical

Exa

min

atio

n of

Sea

fare

rs R

egul

atio

nsCS

ARe

gula

tion

man

ning

safe

ty a

nd h

ealth

Safe

Man

ning

Reg

ulat

ions

CSA

Regu

latio

nNo

t est

ablis

hed

yet

None

Boat

ing

Rest

rictio

n Re

gula

tions

CSA

Regu

latio

nre

gist

ratio

n/cl

assi

ficat

ion/

stan

dard

sNo

neCl

asse

d Sh

ips

Insp

ectio

n Re

gula

tions

, 198

8CS

ARe

gula

tion

regi

stra

tion/

clas

sific

atio

n/st

anda

rds

None

Publ

icat

ion

of S

tand

ards

Reg

ulat

ions

CSA

Regu

latio

nre

gist

ratio

n/cl

assi

ficat

ion/

stan

dard

sNo

neSh

ip’s

Drau

ght M

arks

Exe

mpt

ion

Orde

rCS

ARe

gula

tion

regi

stra

tion/

clas

sific

atio

n/st

anda

rds

safe

ty a

nd h

ealth

Load

Lin

e As

sign

men

t Aut

horiz

atio

n Or

der

CSA

Regu

latio

nsa

fety

and

hea

lthen

viro

nmen

tSa

fe C

onta

iner

s Co

nven

tion

Regu

latio

nsSC

CARe

gula

tion

safe

ty a

nd h

ealth

envi

ronm

ent

Tran

spor

tatio

n of

Dan

gero

us G

oods

Reg

ulat

ions

TDGA

Regu

latio

nsa

fety

and

hea

lthin

frast

ruct

ure

Char

ts a

nd N

autic

al P

ublic

atio

ns R

egul

atio

ns, 1

995

CSA

Regu

latio

nsa

fety

and

hea

lthNo

neFi

re D

etec

tion

and

Extin

guis

hing

Equ

ipm

ent R

egul

atio

nsCS

ARe

gula

tion

safe

ty a

nd h

ealth

None

Gene

ral L

oad

Line

Rul

esCS

ARe

gula

tion

safe

ty a

nd h

ealth

None

Grea

t Lak

es N

avig

atio

n Sa

fety

Reg

ulat

ions

CSA

Regu

latio

nsa

fety

and

hea

lthNo

neLa

rge

Fish

ing

Vess

el In

spec

tion

Regu

latio

nsCS

ARe

gula

tion

1 This

is a

n ab

brev

iate

d ve

rsio

n. A

mor

e co

mpl

ete

list i

s av

aila

ble

from

the

auth

ors

on re

ques

t.2 Ac

rony

mAc

tAW

PPA

Arct

ic W

ater

s Po

llutio

n Pr

even

tion

Act

CLC

Cana

da L

abor

Cod

eCS

ACa

nada

Shi

ppin

g Ac

tCT

AISB

ACa

nadi

an T

rans

porta

tion

Acci

dent

Inve

stig

atio

n an

d Sa

fety

Boa

rd A

ctGP

TAGo

vern

men

t Pro

perty

Act

NWPA

Navi

gabl

e W

ater

s Pr

otec

tion

Act

PAPi

lota

ge A

ctSC

CASa

fe C

onta

iner

s Co

nven

tion

Act

SLSA

ASt

.Law

renc

e Se

away

Aut

horit

y Ac

tSS

ISh

ip S

afet

y In

form

atio

n (s

et o

f sta

ndar

ds a

nd v

olun

tary

gui

delin

es)

TDGA

Tran

spor

tatio

n of

Dan

gero

us G

oods

Act

AP

PE

ND

IX D

: TH

E P

OLI

CY

IN

VE

NTO

RY

1 (C

ON

T’D

)

Prim

ary

Topi

cSe

cond

ary

Topi

cPo

licy

Inst

rum

ent

Act2

Type

of I

nstru

men

t



APPENDIX E“G AP ANALYSIS”

Secondary TopicPrimary Topic

taxation

subsidies

statistics

shipbuilding

security

safety and health

research/innovation

registration/classification/standards

rates

planning

manning

maintenance and repair

infrastructure

freight

finance

equipment and supplies

environment

energy

comm

unications

coastwise shipping (cabotage)

climatic factors

automation

accidents

abandoning of ships/scrapping

Documents

Anatomy of a Policy Area: The Case of Shippingqed.econ.queensu.ca/pub/cpp/Dec97/Pal.pdf · Anatomy of a Policy Area: The Case of Shipping 399 ... based on William Dunn’s methodology