5 The Introduction of a Value-Added Tax in Trinidad and Tobago

Richard Bird, and Milka Casanegra de Jantscher
Published Date:
September 1992
  • ShareShare
Show Summary Details
John F. Due and Francis P. Greaney 

Trinidad and Tobago is a member of the Caribbean Community (CARICOM), one of the Community’s “big four” (together with Barbados, Guyana, and Jamaica). Following independence and the formation in 1973 of CARICOM, the four countries had many features in common: unusually high dependence on income taxation,1 with steeply rising marginal rates, a uniform (with minor exceptions) external tariff that the countries could not unilaterally raise under terms of the CARICOM agreement, and the traditional excises applying only to domestic production.2 However Trinidad and Tobago differs from the others in that it is an oil exporting country—petroleum exports accounted for 84 percent of its exports in 1965 and 72 percent in 1987.

Table 1 shows Trinidad and Tobago’s major tax sources for selected years as a percentage of GDP. Oil revenues rose sharply through the 1970s as the dominating source of revenues; all other taxes declined as a percentage of tax revenue, especially levies on domestic goods and services. As a percentage of GDP, however, individual income taxes increased substantially, whereas customs duties fell.

Table 1.Trinidad and Tobago: Major Tax Sources(As percentages of GDP)
Direct taxes
Oil companies2.523.56.89.2
Other corporate2.
Indirect taxes
Domestic goods and services4.
Import duties3.
Property and other0.
Source: Ministry of Finance.
Source: Ministry of Finance.

Development of Interest in the Value-Added Tax

As long as Trinidad and Tobago had substantial oil revenue, it was less concerned than the other three major CARICOM countries about modifying the indirect tax structure to increase revenues. But with the decline in world oil prices, the Government turned increased attention to the question. In his budget speech of 1983, the Prime Minister and Minister of Finance of Trinidad and Tobago stated, “We have been giving consideration for some time to the introduction of a general sales tax. I am advised that there are numerous problems in the administration of such a tax. I therefore propose to seek the assistance of the International Monetary Fund as to the form which would be most appropriate in our circumstances.” This request resulted in a report by the IMF entitled Trinidad and Tobago: Sales Tax and Other Options of Indirect Tax Reform (May 18, 1983). No immediate action was taken on this report, however, because it recommended a sales tax to be imposed at the manufacturers level and the existing purchase tax was already being imposed on the wholesale value of goods.

In his 1986 budget speech, the Prime Minister indicated that changes in that year in the purchase tax represented movement toward a full implementation of a general sales tax. The Acting Permanent Secretary of Finance requested the Board of Inland Revenue to develop a proposal for a general retail sales tax, and to this end the Board appointed a committee headed by Michal Christian, Assistant Commissioner of Inland Revenue, to prepare such a plan. During 1986, the Committee held a number of meetings, gathered relevant information, and prepared a report. The Committee concluded that with a retail sales tax, a rate of 21 percent would be required to raise the same revenue as the purchase taxes, which had relatively high rates. A 40 percent rate would be required if food were exempted.3 If various service sectors were included, the necessary rates would be 13 percent (food taxed) and 19 percent (food exempted). The Committee did not regard a tax with such rates as feasible, given rates in other countries, and sought to avoid multiple rates, which would allow a lower basic rate.

Preliminary Measures

In the 1988 budget speech, the Government announced that it had established in 1987 a Tax Performance Committee, with Steve Ferguson as Chairman, to further consider plans for a general sales tax, and indicated that a sales tax would be introduced in 1988. The establishment of a tax committee, comprised of representatives from labor, business, and the Government, to examine the tax system and recommend changes to the Government was not new. This approach had often been taken in the past, with formal committees established in 1967, 1981, and 1986 to review the tax system and recommend reforms in the system. The terms of reference of the Tax Performance Committee were not limited to the indirect tax system, but rather included an examination of the entire system of taxation.

In early 1988, the business community invited to Trinidad R. Watson of Price Waterhouse and Co. (U.K.), an expert on value-added tax (VAT) operations in the United Kingdom, for discussions about the feasibility of a VAT. Following a meeting on January 16, 1988, at which the participants were Mr. Watson and representatives of the private sector and the Government, a draft report on the introduction of a VAT was prepared by a committee headed by Mr. Ferguson.4 This brief report considered the relative merits of a retail sales tax and a VAT and indicated that the business community favored the VAT. A schedule of work was set up, with the hope that the tax could be introduced January 1, 1989, and steps in the operation of the tax were outlined. The report noted the necessary educational program, favored administration by Inland Revenue rather than Customs and Excise, and stressed the need for simplicity and for care in designing the new tax.

In May 1988 the Policy Economics Group of KPMG Peat Marwick formally entered into an agreement with the Government of Trinidad and Tobago to assist the Tax Performance Committee and the Ministry of Finance in developing a comprehensive tax reform program. The scope of work of the Policy Economics Group was divided into three phases and included the issuance of various reports which are referred to below.

Phase I of the study consisted of two principal components: (1) a review of the current tax system covering indirect taxes, individual income taxes, corporate income tax, petroleum taxes, and tax administration; and (2) the development of preliminary recommendations for directions for reform.

In phase I, particular emphasis was to be placed on (1) simplifying and expanding the system of indirect taxes; (2) improving the structure of the individual income tax with special attention paid to the equity of the system; and (3) reforming the corporate income tax to broaden the base and lower the tax rate to improve the environment for economic growth.

Phase II was focused on the development of models to analyze the entire tax system. Three microsimulation models were developed for the individual income tax, the corporate income tax, and the system of indirect taxation.5 These models, based on detailed tax and economic data, were designed to estimate the revenue and distributional effects of proposed changes in the tax system. The models provided the Government and the Tax Performance Committee, for the first time, with a reliable means of quantifying the effects of proposed tax changes.

Phase III called for developing specific recommendations for a comprehensive tax reform program along with a detailed plan for implementing the reforms. The aim was to have a final report of the overall tax reform program completed by the end of 1988, together with a detailed implementation plan.

Peat Marwick assembled a team for the project, and team members visited Trinidad and Tobago on various occasions in 1988 and 1989. The team worked closely with the Tax Performance Committee, Inland Revenue, Customs and Excise, and other government agencies and private sector firms, and held a series of discussions concerning their findings and recommendations with the Tax Performance Committee and the Government. To assist in the work of the tax reform team, a specially appointed technical committee was formed. This committee included representatives from the Ministry of Finance, the Board of Inland Revenue, Customs and Excise, and the Central Statistical Office. Of great benefit was the assignment of two government officials to assist the members of the team working on indirect taxes.6

Analysis of the Existing Indirect Tax System

A detailed review of the existing indirect tax structure and administration was the first step on the part of those members of the team who were involved with policy on indirect taxes. This work resulted in a report entitled Indirect Taxation in Trinidad and Tobago: Initial Report, May 12, 1988, which was combined with analysis of other aspects of the tax system and was formally presented to the Tax Performance Committee and the Ministry of Finance on June 1, 1988. A report entitled Draft Phase I Report: Initial Overview of Trinidad and Tobago Tax System was also prepared. The Ministry and the Committee provided considerable feedback on the direction of change, and the team developed more specific recommendations concerning the design of a tax reform program for Trinidad and Tobago.

Indirect taxes yielded about 32 percent of total tax revenue. Of this, import duties plus a stamp tax on imports, purchase taxes on various goods applying to both imports and domestic production, and excises (applying only to domestic production), each yielded about 8 percent, or 24 percent in total. As of 1988, there were four ad valorem rates of purchase tax (at 20, 40, 60, and 85 percent), plus specific rates on tobacco products. Tax applied to importation and to sale by manufacturers. The total purchase tax coverage was broad, only a few basic necessities being excluded plus various inputs of industry. Over one third of the revenue from purchase taxes was obtained from alcoholic beverages, tobacco products, food and soft drinks, and building materials. The purchase tax did not apply to services. There were about 600 active manufacturers paying tax. The tax was administered by Customs and Excise.

While the tax had some desirable characteristics, there were serious limitations. The designation of taxable items by tariff number resulted in unnecessary restriction of coverage. The relatively high rate—40 percent on many commodities—inevitably resulted in misclassification and adverse economic effects. While relatively few inputs in production were taxed, the lack of a general rule for excluding them resulted in some cascading and restricted the ability to extend the coverage of the tax to goods used both as production inputs and for consumption. The levies did not constitute a broad-based sales tax.

There were also serious inadequacies in administrative requirements, particularly the lack of penalties for failure to file and pay. There was no trained audit staff. Failure to cover any part of the distribution sector restricted potential revenue. Overall, the purchase tax was unnecessarily complex, restricted in scope, and applied unusually high rates on many transactions. Compliance requirements were inadequate.

In the excise field, petroleum products had become the major revenue source, yielding two thirds of the total excise revenue. Excises applied only to domestic production under the traditional British Commonwealth practice, and all had specific rates. There were in addition several special levies on sales of motor vehicles and on hotel, telephone, and electric services.

The overall evaluation was that the indirect tax system was in urgent need of reform, to lessen complexity, broaden the scope, reduce the number of separate taxes, introduce computerization, establish an audit program, improve equity and buoyancy of revenue, end adverse economic effects, and strengthen administration.

Development of Proposals for Change

The report prepared by the team initially proposed the establishment of a general sales tax, merging in the existing levies (except customs), with the suggested title of General Indirect Consumption Act.7 The report also reviewed alternative forms of general indirect taxes. It recommended the value-added technique, but stressed the need for further review before making the key decision on the extent to which the tax should go forward beyond the manufacturing sector. It also noted other issues to be resolved, such as exemptions, taxation of services, choice of single versus multiple rates, the tax rate, exclusion of small firms and location of administration, and presented a list of steps to be taken in developing the tax.

Following further work in May and June 1988, a supplement to the initial report was prepared. A substantial portion of the supplement discussed how far the tax should be extended forward to the retail sector. Consultations with various trade and industry groups, three major firms involved in importing, manufacturing, wholesale distribution, and retailing, the Chamber of Commerce, and other business groups, plus examinination of data from the Central Statistical Office, led to the recommendation that the scope of the VAT not be defined in terms of sector (manufacturing, wholesale, retail) but in terms of size of establishment. Given the complexities in the distribution system, many firms were carrying on business in several sectors, and the inevitable influence of the sector approach would encourage firms to shift functions forward beyond the point of impact of the tax. The decision was partly influenced by the high literacy rate of the country and the adequacy of record systems; the same reasoning does not necessarily apply to other developing countries. Trinidad, which ranks thirty-sixth from the highest level of human development among the 130 countries listed in Human Development Report, 1990 is vastly different from the typical developing country.8

The discussions showed substantial support in the business community for increased reliance on indirect taxation, as well as the value-added approach, primarily for enforcement reasons. There was no sharp division in the business community on the preferable form of the sales tax, as there was in New Zealand and Australia.

Several major issues remained, however, including the treatment of agriculture, the selection of the threshold figure above which registration, collection, and payment of tax would be required, and exemptions and zero rating. Further investigation by the team and discussions with various persons involved in the private sector, ministries, the Agricultural Development Bank, and faculty at the University of the West Indies, St. Augustine, facilitated making recommendations on these issues.

Indirect Taxation in Trinidad and Tobago: Second Report was completed in July 1988. This report incorporated the various recommendations, noted the issues that remained, and made a number of more specific recommendations, based on equity, administration, and economic efficiency considerations:

  • The threshold figure should be TT$50,000 (about US$12,500).

  • Tax credit should be allowed to registered firms for tax paid on all business inputs.

  • Exclusion from tax should be limited to a very small number of unprocessed foodstuffs and a few other commodities. This recommendation was based in part on analysis of data from household budget studies.

  • Taxation of all services was not recommended, for equity and operational reasons; a limited number of services considered suitable for taxation were listed.

  • A single rate for the VAT was strongly recommended. Only the excises (except a few minor ones) and one special levy would be retained—a simplified ad valorem rate on sales of motor vehicles.

  • Administration should be placed with Inland Revenue, because of the importance of audit and the fact that the staff of Customs and Excise did not have the appropriate training and skills for this function. It was also thought to facilitate the integration, to some degree, of VAT and income tax audit.

  • Farmers, comprising a small portion of Trinidad’s economy, in general should not be registered for the tax, but major farm inputs would be zero rated to relieve their produce of tax.

This report was supplemented with additional analysis of the revenue and distributional effects of the current tax system and of alternative tax policies. This analysis was undertaken with the help of a computer-based indirect tax model designed specifically for that purpose.

The indirect tax model was constructed around a detailed data base of tax and economic data that captured all the important flows of goods and services in the economy. This approach provided a single tool with which to analyze a wide variety of indirect tax policy alternatives and allowed for the explicit modeling of the operation of the VAT.

The data base was constructed from detailed data compiled from trade statistics, national accounts, surveys of business establishments, household budgetary surveys, and actual tax collections, all of which were generally available for 1987 from the Central Statistical Office and the Ministry of Finance. After the data was integrated into a comprehensive and consistent form for the base year 1987, it was extrapolated to future years using the Government’s forecast of future economic activity.

The model used this data to replicate or simulate the operation of the entire indirect tax system. Taxes on international trade (that is, import duties, stamp tax, and purchase taxes on imports) were calculated based on the detailed data on the value and type of imports, categorized by tariff heading. Domestic indirect taxes (that is, purchase taxes, excise taxes, and other special indirect taxes on domestic production) were calculated based on domestic production data and the tax base implied by current tax collections. The modeling of the VAT was more involved. Not only were import and domestic production data used to determine the tax base, but data on import destinations and interindustry flows were used to model explicitly the crediting mechanism of the VAT. This approach enabled the model to account explicitly for the differential effects of exemption and zero rating at various stages of the production and distribution process. It also captured the revenue effect of zero rating exports. The model took into account not only VAT paid directly by consumers but also the indirect effects of VAT paid by exempt firms, such as small businesses, that was assumed to be passed along to consumers in the form of higher prices.

The distribution of the tax burden by household income class was estimated, based on detailed expenditure data from the household budgetary survey that was included in the model’s data base. An overview of the indirect tax model is presented in Annex I.

The report’s recommendations on both indirect and direct taxes were formally presented to the Cabinet and the Tax Performance Committee in separate meetings held on September 7 and 8, 1988. While the recommendations were well received, both the Government and the Tax Performance Committee offered numerous comments and alternative proposals that required further analysis. Throughout the entire process, a heavy emphasis was placed on the quantitative analysis of the tax reform proposals. This emphasis served to focus on developing a sound tax reform program that met not only the structural objectives of good tax policy but also satisfied the revenue and distributional goals of the Government. On October 17, 1988, the proposed tax reform program, with appropriate modifications, was submitted to the Government and the Tax Performance Committee.

Finalizing the Tax Reform Package

The revised tax reform program was presented to the Government in a formal report entitled Tax Reform Program for Trinidad and Tobago, Final Report (December 1988), which covered all the taxes. While primarily presenting the recommendations of the previous reports, this report contained a major addition, proposing a rate of 18 percent for the VAT. This figure was based upon estimates of revenue necessary to make up for that lost from the repeal of the purchase taxes, repeal of a number of miscellaneous special levies (but not the excises), and a reduction in the individual and corporate income taxes.

The tax reform program was formally announced in the budget speech of the Minister of Finance in December 1988 and a Provisional Collection of Taxes Order was issued to effect the changes. The first phase of the reform took effect in January 1989 with the introduction of extensive income tax reforms.

The overall income tax reform called for:

  • Substantial reductions in marginal income tax rates for both individuals and corporations. The number of rates was reduced from 11 to 4 and overall tax rates were reduced. The top individual income tax rate was reduced from 55 percent to 35 percent, with the proposed rate schedule being 5, 15, 25, and 35 percent.

  • Raising the threshold for imposition of the tax significantly, the allowance being replaced by credit against tax—TT$600 for the taxpayer and spouse and TT$100 for dependent children. The base was broadened by eliminating various special allowances.

  • Cutting the corporate income tax rate from 49.5 percent to 35 percent. A wide range of allowances (investment, initial, and so on) were repealed.

The VAT Implementation Plan

Continued work resulted in a report entitled Proposed Program for Implementation of a Value Added Tax in Trinidad and Tobago (January 10, 1989), which defined in detail the responsibilities of the consultants and the Government and outlined the steps to be taken in 1989 so that the tax could become operational on January 1, 1990, the date indicated in the 1989 budget speech. The program provided a detailed month-by-month statement of the objectives to be attained in four areas: policy and planning; organizational analysis, staffing and training; public education and publicity; and computerization.

To facilitate the implementation of the VAT program, experienced tax administrators from Customs and Excise in the United Kingdom and Inland Revenue in New Zealand were recruited to assist in policy decisions, framing of legislation, and training. Additional personnel were also recruited to work with the Port of Spain office of Peat Marwick to complete the computerization program. A detailed checklist of issues on which decisions had to be made was developed, and discussed at length among persons on the consulting team, government officials, and the Tax Performance Committee.

A VAT Implementation Team was appointed by the Ministry of Finance to carry out the implementation program for the VAT. This team was drawn from the Board of Inland Revenue and Customs and Excise and included the following: an assistant commissioner of Inland Revenue, who was designated head of the VAT office, a tax lawyer, a tax officer, a tax training specialist, a computer systems analyst, and a customs and excise tax officer. Members were assigned specific responsibilities in their area of expertise. The group met frequently to formulate key policy and administrative recommendations for the Government.

The consultants held weekly meetings with the VAT Implementation Team to review the accomplishments of the previous week, to plan activities for the following week, and to address any problems that arose. They also met periodically with the Tax Performance Committee and the Minister of Finance to review and discuss detailed policy and administrative recommendations. In the course of these meetings, unresolved issues were reviewed and decisions were made—in the end by Inland Revenue and the Ministry of Finance—on almost all issues, although some new ones appeared later. The Government decided to call the tax the value-added tax, rather than general consumption tax (as in Jamaica) or goods and services tax (as in New Zealand and Canada). It is not feasible to list all of the issues that were considered, but their general nature can be summarized briefly:

(1) Final decision on services: specified, or all services except those specifically excluded.

(2) Restrictions on credits for tax on business inputs.

(3) Exclusions: zero rating versus exemptions for food, medicine, clothing, and so on.

(4) Actual coverage of the services category.

(5) Special rules on imports; zero rating of exports.

(6) Sales to and by the Government and various organizations.

(7) Sales of used goods.

(8) Tax treatment of farming and fishing.

(9) Exact rules for small firms.

(10) Rate structure: single versus multiple.

(11) Rate level.

(12) Quotation of tax; tax invoices.

(13) Elements in taxable price.

(14) Time of tax liability.

(15) Building construction, rentals, and sale of buildings.

(16) Allowance of refund of excess input tax credits.

There were likewise a number of issues relating to operation of the tax, legal aspects, and operating procedures, on which decisions were made as noted below.

Final Decision on the Structure and Operation of the VAT

During March-July 1989 work intensified on all aspects of the tax, preparatory to the introduction of legislation in the summer period. Major activities included the following:

(1) Final decision on policy issues, which resulted in the legislation subsequently drafted, as noted below. This involved cooperative effort between the consulting team and the VAT Implementation Team. In addition, the Tax Performance Committee and the Minister of Finance were involved in discussions concerning the major VAT policy issues.

(2) The selection of a director of the VAT office and the recruiting of personnel. Michal Christian, Assistant Commissioner of Inland Revenue, was named VAT Commissioner and was also appointed to the Board of Inland Revenue. For several years, she had been involved in consideration of a general sales tax. The VAT Administration Center, as it came to be called, proceeded to recruit personnel, largely from Inland Revenue, but some from the outside.

(3) An extensive program of training persons in VAT operations.

(4) Development of the computer system, assisted by personnel of the Port of Spain office of Peat Marwick, technical consultants from outside the country, and personnel within government.

(5) Establishment of an informational and educational program, with guidance from an official of the New Zealand Government and the contracting of a local advertising firm.

At this time, a number of key decisions were made about the operation of the tax:

(1) The tax would be administered by a separate unit in Inland Revenue, with its own enforcement and audit staff, rather than under a functional system, in which audit, enforcement, and other activities would be administered by units involved with all taxes. It was agreed by all that the tax should be handled by Inland Revenue, and not by Customs and Excise.

(2) Three local offices, in Port of Spain, San Fernando, and Tobago, would be established.

(3) A bimonthly system of returns would be adopted, with half of the firms filing in each two-month period.

The organizational structure would consist of five units: (1) operations, to deal with registration and returns processing, records and office management, and accounts; (2) the compliance unit, including two sections: enforcement, which would have jurisdiction over the regional offices and which would ensure that returns are filed and taxes are paid, and audit, which would be concerned with the payment of correct amounts; (3) the information systems, with control over the computer system (which would be wholly under the jurisdiction of the VAT administration); (4) the policy unit, which would be responsible for audit selection, policy, and research, as well as liaison with Customs and Excise; and (5) the legal unit, which would be involved with interpretation and changes in the act.

The Tax Reform Program: The Second Phase

The policy discussions and decisions resulted in the preparation of two policy documents. The first, A White Paper on Tax Reform for Trinidad and Tobago: The Second Phase, placed the introduction of the VAT in the context of the overall tax reform effort. The first phase of the tax reform, put into effect in January 1989, focused on reforms of the income tax system, broadening the base through eliminating a wide variety of special exemptions and allowances, reducing taxes, and generally simplifying the system. The second phase focused on implementing the indirect tax reforms, principally the elimination of the purchase tax and numerous other indirect taxes and the introduction of the VAT. Also, because the tax mix changed, the second phase led to a further reduction in income tax rates.

The second paper. The Design of a Value-Added Tax for Trinidad and Tobago, presented in some detail the structure and operation of the proposed VAT, including the following:

  • A rate between 15 percent and 20 percent would be used, with a single rate to avoid operational complications.

  • The threshold for small business would be raised to TT$75,000.

  • Specified services would be included in the tax; financial, insurance, and professional services would not be taxed.

  • Zero rating would be limited to exports and a few basic foods and medicines requiring a prescription.

Late in March 1989 the consulting team presented the recommendations to the Minister, the Permanent Secretary, and the Tax Performance Committee, and with minor exceptions the recommendations were accepted, approved by Cabinet, and served as the basis for ultimate drafting of the legislation. The check list for policy issues was replaced by a statement of proposed decisions on the tax, and this aided the preparation by the Legal Counsel, in cooperation with consultants, of the statement on which the draft of legislation would be prepared.

Drafting the VAT Legislation

The Chief Parliamentary Counsel drafted the VAT legislation. It was completed in July 1989, and a number of meetings were held between members of the consulting team and the Minister of Finance and the Permanent Secretary. Several major decisions were made by the Government:

  • The VAT rate was set at 15 percent, instead of 18 percent as proposed originally; the lowering was made possible by an expansion of the base and the maintenance of the same level of taxation on traditional excise tax goods.

  • The threshold annual sales figure for registration of firms was raised to TT$120,000 (about US$30,000).

  • Crude oil and natural gas were zero rated, to avoid liquidity problems in the petroleum industry. (Since natural gas is used solely by a few particular industries, there would be basically no revenue effect from the change.)

  • A decision was made to tax construction services, but not the sale of newly constructed buildings. Since this change reduced the tax on new housing (many small contractors would be below the threshold), an initial provision for a refund of tax on new housing was eliminated.

Subsequently, as a result of debate in Parliament, the structure of the VAT was changed to tax all services except those explicitly exempted.

The legislation, the passage of which required a three-fifths majority of all members, was passed by the House on August 18, 1989, and by the Senate on September 8, 1989. By no means were all the issues resolved, however. In subsequent meetings between the VAT officials and the consultants, various decisions were made, though issues continued to arise through 1990.

Following the enactment of the legislation, booklets and other instructions for the business community and the public were prepared and the services of an advertising firm were used to provide substantial newspaper and television publicity. The Minister of Finance traveled throughout the country giving presentations to various community and business groups explaining the objectives of the overall tax reform program, its anticipated revenue and distributional effects, and many of the specifics on the structure and operation of the VAT. Again, the overriding theme was that the VAT must be viewed as part of the overall tax reform program, which was designed to provide a stable basis for economic growth.

At all three local offices, staff members were available to provide information. Forms for registration were prepared and distributed, and tax return forms were developed. Recruitment and training continued. Registration was required by October 31, 1989.

Response to the Tax and Criticism

No new tax is ever popular—but on the whole the tax was received relatively well. The business community favored it—unlike the reaction in many countries—partly because the tax allowed income taxes to be reduced and partly because of the unsatisfactory nature of existing indirect taxes. Cooperation between the Government and the business community obviously facilitated a favorable reaction. The downturn in the economy and the sincere desire of both business and the Government to take possibly difficult actions for the good of the country as a whole played an important role in reaching consensus on the appropriate tax reform program.

Inevitably, the opposition in Parliament criticized the new tax, arguing that the shift to VAT reduced income tax and luxury tax burdens on higher-income groups and increased the burden on lower-income and middle-income groups. The Government did make a strong effort to minimize the burden on the poor, but some shifting between upper and middle-income groups was inevitable, given the general objectives of the overall tax reform. Furthermore, there had undoubtedly been considerable evasion or avoidance of income taxes at the higher-income levels. Those objecting to shifting burden to the poor argued for excluding all medicines, not merely prescription ones, from the tax, and for excluding books, or at least textbooks. But driven by important revenue goals and with special attention paid to the distributional effects of the tax system, the Government kept to several politically difficult decisions concerning the appropriate coverage of the VAT. The Government also committed itself to increasing the funding of certain transfer programs, such as old-age pensions, food subsidies, and public assistance, aimed at assisting low-income and elderly persons.

In everyday discussions, much of the criticism was against the 15 percent rate; 10 percent was often mentioned as a reasonable figure. But the purchase taxes being removed had rates of at least 20 percent. Revenue needs and the desire to increase reliance on indirect taxes largely dictated the rate level.

Concern was expressed about inflationary effects of the tax and fear that the tax would be added to prices without downward adjustments reflecting the elimination of the purchase taxes. There was also concern that small firms would adjust prices upward by the amount of the tax rate applied to sales even though they were only taxed on inputs. Some of the criticisms included outright misinformation—the argument, for example, that registered firms would shift forward the taxes on their inputs even though these amounts were deductible against tax due on sales.

People involved with various activities were fearful of adverse effects. For instance, promoters of various Carnival activities were afraid that they would have to pay tax on goods purchased in 1989 for delivery in 1990 at Carnival time; this potential problem was alleviated by the introduction of transitional provisions. The hotel industry was concerned that the 15 percent tax on hotel charges was out of line with taxes in neighboring countries—and thus would hamper the tourist trade that the country was trying to develop. However, the evidence showed that the major impediments to the development of the tourist industry related more to infrastructure and marketing decisions, including the quality, availability, and accessibility of resort hotels, than to the rates of existing hotels. Farm groups sought zero rating of additional farm inputs.

There was concern over the lack of clarity in defining zero-rated commodities and various taxable services—a problem that was gradually lessened by interpretations of the provisions. But questions continue to arise about the meaning of “unprocessed” food. Further, the issue of taxing second-hand goods has not been resolved.

Finally, there were concerns about firms advertising to buy early to avoid the tax (ignoring the fact that the purchase taxes would come off), and after the tax became effective, advertising claims by firms that they were not shifting the tax to the consumer.

Experience with VAT Operation, 1990-91

As scheduled, the tax went into effect on January 1, 1990. By June 1990, the revenue yield from the VAT was in line with the original estimates; the estimates proved to be reasonably accurate. As of June 1991, there were 10,145 registered firms, somewhat lower than the original estimate of 12,000. Of these, 66 percent were in Port of Spain district, 31 percent in San Fernando, and 3 percent in Tobago. During the initial registration period, 52,144 registration packets had been mailed out, the list based on a number of sources of potential taxpaying firms. Many of these names were duplicates, and a significant number of firms had gone out of business but had never been removed from the various registers. By October 10, 1989, 15,929 of the applications were completed and returned to the VAT Center. Of these, 6,343 were found to have sufficient sales to be subject to VAT. Of the remainder, 17,104 were returned as duplicate, unclaimed, out of business, and so on; 18,478 never responded; and 633 applications were returned incomplete.

By December 12, 1989, 8,215 registrations were completed; by August 8, 1990 the number of registrants was 9,729, by December 1990, 9,688, by June 1991, 10,145. The figure for population divided by the number of registered firms, 118, compares favorably with countries in a similar situation; it is unlikely that the number is substantially less than should be expected.

The nonfiling or delinquency rate—that is, the number of registered firms not filing—was initially very high; it had fallen to 25 percent by the bimonthly period ended May 25, 1990, but was still 24 percent in November 1990. Roughly half of these delinquencies were cleared a month later.

By January 1991, the chief deficiency was the failure of the computer system to perform all desired functions. There were delays in developing the computer system along the way, and for reasons by no means obvious the completed system was inadequate. The most serious failure was the initial inability of the system to identify nonfilers and prepare a list and notices for mailing. The department responded quickly to develop a substitute approach manually. Ultimately, the system was brought into full operation. Meanwhile the delinquency rate remains relatively high.

The audit program was under way by mid-1990, but it is not adequately staffed because of training requirements and illnesses. As of July 1991 there were 16 active auditors. During March 1990-June 1991, there were 330 audits of refund cases and 222 audits of payable cases. Of the total, 140 cases involved no change. Additional tax assessed was TT$5 million.

The budget estimate of revenue from the tax for the calendar year 1990 was TT$863 million. The actual yield after refunds was TT$899.7 million, approximately 4 percent higher than anticipated. The gross yield was TT$1,213.1 million, of which TT$581.7 million (48 percent) was collected at customs, TT$630 million on domestic sales (52 percent), and a negligible TT$1.4 million from government departments. The total figure of refunds for input tax credit was TT$313.4 million, or 26 percent of the gross yield.

For the 1991 budget year, the tax is expected to yield 25 percent of total tax revenue; in 1987, the replaced purchase taxes yielded only 8.8 percent of total revenue.

The general situation after more than a year of operation was reasonably satisfactory. The VAT office has effective leadership and competent staff, a potentially good computer system, excellent physical quarters, and the support of the top level of the Government. The overall program of design and introduction of the tax proved to be satisfactory. Some difficulties arose, as is almost inevitable.

There were considerable delays in staffing and obtaining physical resources. The rigidity of public service administration and procedures caused much of the staffing delays. As in many countries, both developed and developing, more importance was placed on seniority than capability. On the whole, however, the VAT office was successful in obtaining exceptionally capable personnel.

Implementation efforts fell several months behind the original schedule because of a late start and the other unavoidable delays. Delays in staffing, training, development of legislation, and decisions on many basic policy issues adversely affected both information for the public and staff training. The difficulties with the computer system should have been avoidable. Striking a balance between authoritative leadership and committee decision making is always difficult, and it may be argued that too much emphasis was placed on the committee approach, in terms of utilization of time. But these problems should not be exaggerated: the tax became operational on schedule without serious difficulties; it has proved to yield the desired level of revenues; and it is now an important feature of the Trinidad and Tobago tax system. The main problems that remain are a delinquency rate that is still excessive and an inadequate audit program.

The act, as enacted in September 1989, proved for the most part to be satisfactory, but some changes became necessary. Several were made in December 1989, mostly minor, but some of substance: baby formulas and baby milk substitutes were zero rated, for example, and travel between Trinidad and Tobago zero rated instead of being exempted. In early 1990, tax invoices were simplified for specified types of business, requiring much less information. During the spring, a number of rather detailed changes were developed, primarily to schedules in the act, to explain the meaning of various categories of goods and services. Clarifying zero-rated goods such as unprocessed food and exempt services in a workable manner is a serious problem in any VAT. Other provisions of the act could be questioned—for example, the various alternative “schemes” for simplified accounting for tax, which had their genesis in the U.K. VAT and are themselves complicated and may be regarded as unnecessary. The most significant one allows a firm to calculate taxable sales on the basis of the ratio of taxable inputs to total inputs. The explanation of this rule in the act is unnecessarily complex.

Because firms selling to unregistered persons had the choice of quoting the tax separately or quoting tax-inclusive prices, some confusion arose. This, again, was the result of the compromise between the U.K. and New Zealand approaches to VAT. Typically, prices were readjusted to a tax-included basis, and some stores closed down for one or two days to make the changes. These firms were fearful of confusion at the tax register if the tax were added separately.

Data on price increases showed that prices of goods subject to tax rose between 5 percent and 11 percent between December 1989 and January 1990, while nontaxed goods rose by only 2 percent. Even by April 1990, none of the commodity groups had risen in price by as much as 15 percent, an indication that some downward adjustments were made for the repeal of purchase taxes. Prices were relatively stable in 1989; the tax was not introduced in a strongly inflationary period.

While refunds are being paid, there have been some delays. However, the Government is required to pay interest on refunds delayed for more than six months.

Summary of the Tax as of 1991

The major features of the VAT as introduced and modified slightly in 1990 are as follows:

Nature. The tax is a value-added levy of the usual tax credit or invoice type. It applies to imports, as well as to domestic sales. Credit is allowed for tax paid on inputs for business use by registered firms, without exception. The tax extends through the retail level, although small firms in all sectors are free of the requirement to register and collect and remit tax; they pay tax on their purchases and are not eligible for input tax credit.

Coverage. The tax applies to all sales of goods and to all services, except as otherwise specified. The sale of land and buildings is not included in the scope of the tax, although the rental of business property is taxable. Commodities specifically excluded from application of the tax are zero rated; thus, the tax does not apply to sales of these goods, and sellers receive input tax credit for tax paid on inputs to produce them.

Zero rating is limited to (1) unprocessed food and a few basic processed foods, such as flour, bread, milk, and margarine; (2) medicines sold by prescription; (3) live animals, livestock feed, seed, fertilizer, and farm machinery; (4) water sold through pipes; (5) exports and certain export-related activities; (6) natural gas and crude oil; and (7) veterinary and pest control services.

Specified services are exempted, and thus while not subject to tax, suppliers do not receive input tax credits. The principal exempt services are (1) medical, dental, hospital, and other health-related services; (2) most education; (3) rental of residential property; (4) bus and taxi service and postal service; and (5) real estate brokerage, insurance, banking, and stock brokerage.

Small firms. The threshold for registration of firms is set at TT$120,000 (about US$30,000) in annual sales. This figure appears to be satisfactory. Voluntary registration is permitted for firms under the threshold in certain important industries such as farming and manufacturing.

Operations. Returns and payment are required on a bimonthly basis, due by the twenty-fifth of the following month.

Tax invoices are required, with a simplified form for use by gas stations, car parks, and supermarkets. Firms selling to unregistered buyers have the choice of quoting the tax separately from the price or quoting tax-inclusive prices, although the latter is increasingly common.

The tax is administered by the Value-Added Tax Administration Center of Inland Revenue, under the VAT Commissioner, who is a member of the Board of Inland Revenue.

Reasons for the Successful Introduction of the Tax

Despite delays in the full operation of the computer system and problems in acquiring adequate staffing for the audit function, the overall introduction of the tax has been quite successful. There are several reasons for this:

(1) The careful planning that went into the development of the tax structure and administration. Heavy emphasis was placed on the quantitative analysis of the proposed changes.

(2) The close cooperation between business and the Government from the beginning of consideration of the tax. This cooperation was facilitated by the formation and participation of the Tax Performance Committee in the development and review of the VAT proposals.

(3) The extensive publicity program to acquaint the public and taxpaying firms with the operation of the tax.

(4) The introduction of the tax in conjunction with repeal of the purchase taxes and significant reduction in income tax rates.

(5) The selection of competent persons in key positions. The VAT Implementation Team appointed by the Minister of Finance included some of the most capable members of the Board of Inland Revenue.

Lessons for Other Countries

The introduction of the VAT in Trinidad and Tobago in the context of an overall tax reform program enabled the Government to formulate a consistent set of policies with an appropriate mix of winners and losers. The reduction in income tax rates greatly facilitated the introduction of the VAT by providing the benefit of lowering or eliminating income taxes for the lower- to middle-income groups, and increasing benefits to the poor lessened complaints of burden on these persons.

Based on the experience in Trinidad and Tobago, the following suggestions are offered for countries considering a VAT:

(1) If at all possible, the rate should be kept down to 10 percent when the tax is introduced. Much of the complaint in Trinidad was against the 15 percent rate.

(2) If the tax is to yield substantial revenue at a tolerable rate, it is imperative that most food be taxable.

(3) Delineation of exempt from taxable food is difficult in a workable and acceptable fashion; the exemption of unprocessed food, plus a few other items such as milk and bread, was the approach used in Trinidad, and this seems to have been accepted although there are definitional problems, especially about what constitutes “processing.” This approach provides adequate protection to the lowest-income group while maintaining a broad base and lower rate by taxing processed foods, much of which is consumed by middle- and high-income families.

(4) If services are to be taxed, it appears desirable to tax all services other than those exempted or zero rated, in the interest of simplicity. It is desirable to include types of services provided mainly to businesses (which is highly undesirable in the usual sales tax) given the political repercussions of not doing so, and because most businesses will be able to obtain input tax credit for the tax on these services.

(5) Studies in many countries have demonstrated the difficulties of applying a VAT to financial and insurance services, and it is better not to attempt it; the concept of value added is not clearly definable in these fields. Taxation of rentals of nontransient housing is not feasible because of the consequent discrimination against the tenant compared to the homeowner.

(6) It is important to accept zero rating rather than exemption for categories designed to be free of tax; otherwise, the major inputs used to produce the exempt goods must be exempted also, and firms producing both exempt and taxable goods must segregate inputs between the two. Exemption is desirable when there is justification for not taxing the item, but it is considered appropriate for some tax burden to rest on the purchasers of it. Zero rating has to be controlled carefully to prevent fraudulent requests for refunds.

(7) Registration of all farmers is clearly not feasible, in terms of compliance and enforcement. Accordingly, major farm inputs—fertilizer, feed, seed, livestock, and farm machinery—should be zero rated, but not hand tools and items widely used also for nonfarm use, such as batteries for farm and other tractors.

(8) One of the most confusing issues in Trinidad was the question of how firms should indicate tax to unregistered buyers. Clear instructions are necessary for the firms about their policies. There is merit in a requirement for separate quotation, in the belief that this facilitates exact and uniform shifting, as is intended, but the tax-included approach appears simpler.

(9) A major issue is the criterion for excluding small firms from the registration requirement. The aim is to draw the line at the level above which most firms have records sufficient to ensure control of the tax. This figure cannot be determined scientifically, but must be based on the views of persons familiar with the business community, particularly retailing and crafts.

(10) Another important issue relates to allowing credit for previous consumption or purchase taxes on stocks of goods on hand when the tax is introduced. No allowance was given in Trinidad, primarily because that would have had major negative effects on revenue in the first year of operation, as the experience of other countries shows.

(11) Use of consultants from countries that have a VAT obviously has merit. They bring a significant amount of experience to the task at hand. However, one must take care to understand the background, experiences, and points of view of the consultants. Advisors with experience in a number of countries are usually preferable to those whose experience is limited to one country. VAT advisors whose background is in Customs and Excise, such as in the United Kingdom, will have a different approach from those whose background is in Inland Revenue, such as in New Zealand.

(12) There are a number of issues that must be resolved, many of which often escape attention when the tax is first considered. For example:

  • Are hotel service charges to be included in the taxable charge for hotels?

  • How are transient rentals, usually taxed, to be delineated from permanent housing, which is usually not taxed?

  • What sort of invoices must retailers give to registered customers, showing tax paid? These must not be too complex, as many taxable purchases by registered firms will be small in amount.

  • If fertilizer and seed are to be exempt for farm use, how about small packages sold primarily at retail for nonfarm use? Attempts to make the latter taxable in Trinidad were not successful for operational reasons.

  • Are trade-in allowances to be deductible from the taxable price? If not, are the subsequent sales of goods taken in as trade-ins to be taxable?

  • Are sales of second-hand goods to be taxable?

    190 Value-Added Tax in Trinidad and Tobago

  • What is the appropriate treatment of construction of real property for business purposes? How should residential construction be treated?

(13) Implementation schedule. An initial study of the tax structure and operation, computer facilities, and so on, will require about 6 months. Then, a further 6 months are needed from the completion of this study until basic decisions on tax reform are made, and at least another 6 months until implementation. Thus, 18 months from the start is a minimum time; two years is more realistic. Once these decisions are made, several aspects require careful planning and scheduling:

Legislation. Early enactment of the new act is necessary. Plans should be made early for regulations and for instructional booklets. Use can be made of other acts but adapted to circumstances in the country.

Staffing. If a new tax such as a VAT is to be introduced, it is important to acquire staff immediately, particularly a suitable director of the new tax, which typically should be located in Inland Revenue.

Training. One possibility is to send for two months two key officials to a country using a VAT to become well acquainted with the tax and its operation. New Zealand is one of the best, from this aspect; Ireland is another, and Canada once its goods and services tax is in operation for a time. These officials will then direct the training of others when they return. The alternative is to bring a person familiar with the VAT from another country to direct the training. Obviously, this person must be selected with great care.

Public information. Publicity for the new tax must be developed early in the process and made widely available. In this field, both New Zealand and Trinidad have done very well. Both the public and the taxpaying firms need to be instructed about the new tax.

Computer systems. The development of a new computer system is a major effort and can be a lengthy process in itself. It includes analysis of requirements, system design, hardware and software procurement (often done through competitive bidding), software development testing, installation, documentation, and training, and each step can give rise to substantial delays. In Trinidad, the use of advanced software and the temptation to be too ambitious in the initial design resulted in a delay in delivering a fully operational computer system prior to the effective date of the VAT. However, the use of modern data base management software allowed the flexibility to develop isolated applications to address the major needs of the VAT administration, particularly in carrying out the important function of taxpayer registration on a timely basis.

ANNEX I Overview of the Trinidad and Tobago Indirect Tax Model

Trinidad and Tobago’s indirect tax model was designed to estimate the revenue yield and distributional impact of alternative structures of indirect taxation. The model treats the indirect taxes as a complete system and explicitly takes into account the interaction between each of the taxes. Import duties, for example, are included in the base of the purchase tax; therefore, if import duties are increased, the base on which the purchase tax is levied is also increased.

Data Base Construction

A primary focus of the model was to calculate the potential yield from a broad-based VAT. Therefore, substantial effort was devoted to developing a data base detailed enough to estimate the potential revenue yield from alternative rate and exemption policies under the proposed VAT.

The data base was constructed from a variety of sources that included national income and product accounts; surveys of business establishments; import and export data, by tariff heading; household expenditure data; interindustry flow data; tax collection data; and an economic forecast.

These data were used to construct an integrated data base that captured all the important flows of goods and services throughout the economy. An overview of the data base construction is presented in Figure 1.

Figure 1.Overview of Indirect Tax Data Base Development

Structure of the Model

The model was designed to estimate the effects of alternative policies. These included changes in tax rates; changes in tax base (including the effects of cascading in the system); changes in exemption criteria; and changes in the overall structure of the indirect tax system (for example, single-stage taxes versus VAT).

An overview of the operation of the model is presented in Figure 2. The model uses the detailed economic data to simulate the operation of the entire indirect tax system. Taxes on international trade (that is, import duties, stamp tax, and purchase taxes on imports) are calculated based on the detailed data on the value of imports by tariff heading. The value of dutiable imports is then determined (by accounting for the share of exempt imports) and the statutory tax rates are applied to calculate the duty paid. The duty-paid value of imports then forms the base of the other indirect taxes levied on imports, including the VAT.

Figure 2.Overview of the Indirect Tax Model

Domestic indirect taxes (that is, purchase taxes, excise taxes, and other special indirect taxes on domestic production) are calculated based on domestic production data and the implicit tax base of the current system. Essentially the same procedure is used as with imports—the tax base is determined and then the statutory tax rates are applied to calculate taxes paid on domestic production. The tax-inclusive value of domestic production then forms the basis of the taxable value for the VAT.

The modeling of the VAT not only uses import and domestic production data to determine the tax base but also incorporates data on import destinations and interindustry flows to model explicitly the crediting mechanism of the VAT. Parameters specifying the coverage of the VAT are used to determine the base of the VAT. Taxable imports are specified by type of good, and taxable domestic production is specified by subsector. The model applies the statutory VAT rate or rates to the tax bases to calculate the gross VAT paid. It then uses information on the destination of the goods and services to determine how much tax will be credited by taxable firms later in the production and distribution process. The remainder of the tax is either paid directly by consumers or by exempt firms, which will in turn pass the tax along in the form of higher prices.

After calculating the indirect taxes paid on each type of good and service, the model uses information from the household budgetary survey to allocate indirect taxes across income classes according to their patterns of consumption.

Model Results

Each execution of the model calculates the revenue and distributional effects under two alternative scenarios, generally referred to as Plan X and Plan Y. The results of each simulation are presented in a series of output tables. Tables 2 and 3 present two types of output from a simulation of the indirect tax reform program in Trinidad and Tobago. In this simulation, Plan X is current law (pretax reform) and Plan Y is the revised tax reform proposal.

Table 2.Trinidad and Tobago: Indirect Tax Model
Revenue Yield of Indirect Taxes
Type of Indirect TaxPlan X1 (Million TT$)Plan Y2 (Million TT$)Change (Million TT$)Change (In percent)
Taxes on international trade
Import duty203203000.0
Stamp duty173173000.0
Special consolidated levy530- 53- 100.0
Subtotal429376- 53- 12.3
Taxes on goods and services
Purchase tax (imports)10812- 96- 88.7
Purchase tax (local)33893- 246- 72.6
Excise taxes (spirits)424200.0
Excise taxes (beer)373700.0
Excise taxes (petroleum)31331300.0
Excise taxes (tobacco)8961- 28- 31.7
Excise taxes (edible oil and matches)000- 100.0
Motor vehicles tax525200.0
Wireless licenses40- 4- 100.0
Electricity tax120- 12- 100.0
Telephone tax250- 25- 100.0
Hotel room tax10- 1- 100.0
Airline ticket tax140- 14- 100.0
Domestic stamp duty282800.0
Miscellaneous taxes0000.0
Subtotal1,065638- 426- 40.0
Total indirect taxes1,4931,88038725.9
Source: Policy Economics Group of KPMG Peat Marwick.

Plan X: Current law; all indirect taxes.

Plan Y: Revised proposal 15 percent VAT; reduced purchase tax on alcohol and tobacco.

Source: Policy Economics Group of KPMG Peat Marwick.

Plan X: Current law; all indirect taxes.

Plan Y: Revised proposal 15 percent VAT; reduced purchase tax on alcohol and tobacco.

Distribution of Indirect Taxes
Income ClassPlan X1 (Million TT$)Plan X1 (In percent)Plan Y2 (Million TT$)Plan Y2 (In percent)Change (Million TT$)Change (In percent)
0 < 10,0001077.21367.22826.4
10,000 < 20,0001157.71427.62723.7
20,000 < 30,00019112.823012.23920.4
30,000 < 40,00018412.323112.34625.2
40,000 < 50,00018212.223212.34927.0
50,000 < 75,00032721.941221.98526.1
75,000 < 100,00017912.022712.14826.9
100,000 < 200,00019012.724713.25830.9
200,000 and above181.2231.2528.2
Source: Policy Economics Group of KPMG Peat Marwick.

Plan X: Current law; all indirect taxes.

Plan Y: Revised proposal 15 percent VAT; reduced purchase tax on alcohol and tobacco.

Source: Policy Economics Group of KPMG Peat Marwick.

Plan X: Current law; all indirect taxes.

Plan Y: Revised proposal 15 percent VAT; reduced purchase tax on alcohol and tobacco.

Table 3.Trinidad and Tobago: Indirect Tax Model—VAT(In millions of Trinidad and Tobago dollars)
TTSNA CategoryVAT on ImportsCredit VAT on ImportsCalculated Domestic VATCredit VAT Domestic ProductionRefund VAT ExportsTotal
0 Export agriculture000000
1 Domestic agriculture000000
2 Sugar000000
3 Petroleum000000
4 Food, drinks, and tobacco000000
5 Textiles000000
6 Printing paper000000
7 Furniture000000
8 Chemicals000000
9 Assembly000000
10 Miscellaneous manufacturing000000
11 Electricity and water000000
12 Construction000000
13 Distribution000000
14 Hotels000000
15 Transportation000000
16 Business and financial services000000
17 Government000000
18 Education000000
19 Personal services000000

The top part of Table 2 presents the revenue yield of each of the indirect taxes under the two alternative plans, and the dollar and percentage change in revenues between the two. The taxes are grouped into three categories: taxes on international trade, taxes on goods and services, and VAT. The simulation shows the revenue effects of eliminating a number of indirect taxes and introducing a new VAT.

The bottom part of Table 2 shows the distribution of indirect taxes by income classes for the two plans, again including the change in dollars and percent between the two plans.

Table 3 presents a summary of the operation of the proposed VAT by Trinidad and Tobago system of national account (TTSNA) category. Plan X, the prereform law, is presented at the top of the table, and Plan Y, the revised tax reform proposal, at the bottom of the table. The major steps in calculating VAT revenues is summarized in the table:

(1) The first two columns deal with VAT paid on imports. The first column, “VAT on imports” presents the VAT collected on imports originating from the corresponding TTSNA sectors abroad. The second column, “Credit VAT on imports,” presents the amount of VAT collected on imports that is credited when purchased by taxable domestic firms. The credit appears as a negative value in the table and is a loss in revenue remitted to the Government.

(2) The next two columns deal with VAT on domestic sales. The third column, “Calculated domestic VAT,” presents the VAT collected on taxable sales by taxable domestic firms. The fourth column, “Credit VAT domestic production,” presents the amount of VAT paid on domestic purchases that will be credited by taxable domestic firms. Again, the credit appears as a negative value in the table.

(3) The fifth column, “Refund VAT exports,” presents the amount of VAT that will be refunded on exports.

(4) The final column, “Total,” contains the total net VAT revenues to the Government.

TTSNA CategoryVAT on ImportsCredit VAT on ImportsCalculated Domestic VATCredit VAT Domestic ProductionRefund VAT ExportsTotal
0 Export agriculture000000
1 Domestic agriculture3- 10001
2 Sugar5- 298- 30- 1655
3 Petroleum67- 41947- 230- 607135
4 Food, drinks, and tobacco27- 2179- 48- 11144
5 Textiles30- 1130- 4- 343
6 Printing paper29- 1135- 13- 633
7 Furniture17- 360- 118
8 Chemicals51- 3074- 34- 1348
9 Assembly143- 9381- 37- 2469
10 Miscellaneous20- 816- 9- 317
11 Electricity and water0057- 28029
12 Construction14- 717- 5018
13 Distribution0081- 38043
14 Hotels0011- 506
15 Transportation00153- 470106
16 Business and financial services00110- 28082
17 Government19000019
18 Education000000
19 Personal services000000
Total425- 2211,894- 557- 685866
Source: Policy Economics Group of KPMG Peat Marwick.
Source: Policy Economics Group of KPMG Peat Marwick.


Christian Michal Y.

The paper presented by John Due and Francis Greaney gives a very accurate account of the introduction of the value-added tax (VAT) regime in Trinidad and Tobago, albeit as seen through the eyes of consultants who were contracted by the Government of Trinidad and Tobago to provide it with the specialized technical expertise and support necessary to implement such a VAT system. Each party to this contract had clearly defined responsibilities—the consultants were required to take the lead in the technical and administrative aspects of the implementation effort and the Government had the responsibility for final approval of all decisions relating to matters of policy, administration, and work output. The Government also had the critical responsibility for supplying the necessary resources, including manpower, materials, and facilities, to enable the timely implementation of the VAT.

I would like to highlight some of the significant practical issues that we, as administrators, had to address during the implementation stages of the VAT, and share with you some of my observations on the paper.

Requirements for a Successful VAT Implementation

During the initial stages of our implementation effort, concerns were expressed as to whether a small country like Trinidad and Tobago could really introduce a VAT system, since this system is viewed as the most sophisticated type of sales tax so far designed. As the paper points out, one of the earlier studies done by the IMF recommended that Trinidad and Tobago adopt a sales tax at the manufacturing level rather than a VAT system. These fears were based mainly on the fact that Trinidad and Tobago had no experience in operating a sales tax, whereas most countries that had successfully implemented a VAT system had previously operated some type of simple sales tax and thus both traders and administrators in these countries were able to assimilate a VAT system more easily. In spite of this lack of experience, Trinidad and Tobago was able to implement, in the view of the experts, a very successful VAT and is able to dispel the perception that without such prior experience, countries ought not to attempt a VAT system.

I am firmly of the view, however, that one of the key factors to the successful implementation of the VAT system in Trinidad and Tobago, and one that is not included in the paper’s list of “reasons for the successful introduction of the tax,” is that of the strong commitment to the implementation displayed by our Government. This commitment was a genuine one and was not just articulated to satisfy the requirements of an international organization. The Minister of Finance in particular was very dedicated to the project and was also knowledgeable about the system, and this allowed him to resist effectively a great deal of lobbying that took place to extend the range of zero-rated and exempt goods and services. Most countries embarking on a VAT system initially conceive of a well-designed system. However, in my view, it is the ability to withstand the inevitable lobbying that determines the extent to which the system that is finally implemented represents a good VAT system. From an administrator’s point of view, the Government’s commitment is also of vital importance, since it is needed to ensure that all government agencies that have a role to play in the implementation of the system are made aware of the importance of the tax and the need for their cooperation if the established goals are to be achieved.

In my opinion, one of the more difficult areas in the implementation process was staffing for the VAT Administration Center. Because the Inland Revenue Division of the Ministry of Finance was selected to administer the VAT, the first decision that had to be made was whether a separate category of officers should be created for the VAT administration. Two things militated against this. First, there was the problem of time. Traditionally, it takes a very long time to create a new category of officers in the public service since a number of agencies are associated with this process. For example, it involves the Director of Public Administration, who is responsible for determining job specifications, and the Chief Personnel Officer, who is responsible for determining the pay classification, as well as the terms and conditions of employment. Second, it was felt that a special category of officers would make the VAT Center a closed division and would limit promotional opportunities for staff. Since only 76 technical officers were to be assigned to VAT, this number was considered to be too small to have a closed office. After much deliberation, it was decided to use the same classes of technical officers as were used by Inland Revenue, thereby allowing promotional opportunities for existing officers. However, since technical officers for the Inland Revenue are recruited only at the lower levels, a number of experienced officers from the Inland Revenue Division had to be transferred to the VAT Center to fill middle- and higher-level staff positions. This resulted in severe hardship to the other Inland Revenue sections, given the length of time it takes to train replacements.

Another area of concern for us was the time given to implement the system. It is very important to provide adequate time for implementation in order to ensure that the infrastructure is in place. The paper suggests that at least 18 months are needed for this process, and I certainly agree. As you may know, Trinidad and Tobago implemented its VAT system in only 10 months, and during this relatively short time a great deal of planning had to be done. It certainly was not an easy task, and I would not suggest that you attempt to do likewise since there were a number of areas which, in my view, could have been improved upon if more time was available.

I would like to add another note of caution with regard to the timetable. I would urge you to resist setting a time frame that is too far removed from the present. The adage “work expands to fill the time to do it” is found to be true in this situation, since if too long a time period is allowed, the earlier months are often not used productively and the administrators still end up having to rush to complete the implementation within the time allocated. It should also be noted that, regardless of what length of time is given to prepare for the implementation, when the appointed day approaches some traders will still not be ready. This, in my view, will occur especially where the government is not fully committed to the project and traders, recognizing this, adopt a wait-and-see attitude in the hope that the implementation would be postponed and they would be saved from having to learn the system. My advice therefore would be that, provided the administration is at least 90 percent ready, the implementation should be carried out as planned, because in practical terms it is virtually impossible by the due date to have all aspects of the system in place and to have all traders registered and adequately briefed.

The date chosen for the implementation should also be considered carefully. Like Trinidad and Tobago, many countries have implemented their VAT system from January 1. At first glance this date seems to be a good choice, because it is at the start of the calendar year and any laws that need to be abolished to coincide with the introduction of the VAT can be accommodated in a tidy fashion. However, my experience has been that this date is a bad choice from the trader’s point of view. The period from September to December is one of the trader’s busiest periods, when he is preparing for the Christmas season and end-of-year sales and during this time he is also required to do most of the preparatory work on VAT to be ready for the appointed day. Thus, he is often either unable or unwilling to devote as much time as is necessary to implement the system properly. I would therefore suggest that the period chosen for the implementation be one that is slow for businessmen since their cooperation is vital to the success of the venture.

Training of VAT Staff

I would now like to mention some specific aspects of the paper. The paper refers to the extensive program that was introduced during March-July 1989 to train personnel in VAT operations. During that time we trained a core team of eight officers in the VAT law on the understanding that they would act as trainers and instruct all the other technical officers who join the VAT administration. Initially, this proved to be a somewhat frustrating exercise, since at that time all that was available was an outline of the proposed VAT law. When the law was finally enacted, a lot of time had to be spent in retraining the officers in those areas of the law that had subsequently changed.

I would therefore advise that it is critical to enact the VAT legislation early in the schedule. So many things hinge on this that if the law is delayed it can seriously jeopardize the implementation date. Apart from the training aspect, unless the law is in place a full-fledged publicity campaign cannot be begun, nor can the prescribed forms required for the system be designed and published since, more often than not, relevant sections of the law will be quoted on the forms.

We received advice from our New Zealand counterparts on another aspect of training. They suggested that we learn from their mistakes and commence our VAT audits as early as possible. Their audit program was not begun until the year after the implementation date and in the interim they discovered that a number of fraudulent practices had occurred. Bearing this in mind, we hurried to start our audits three months after the implementation date, but because our auditors had initially been conducting advisory visits to registrants we did not have enough time available to train them properly in conducting VAT audits. This lack of adequate training manifested itself in the results of the early audits, which were conducted with minimal effectiveness. We are only now in the process of continuing this training, some 15 months later.

The paper further suggests that to obtain the expertise for a VAT system “one possibility is to send for two months two key officials to a country using a VAT system to become well acquainted with the tax and its operation.” It further suggests that “the alternative is to bring a person familiar with the VAT from another country to direct the training.” I would like to suggest, however, that as a prerequisite to either of these alternatives, key personnel should be exposed in a formal classroom setting not only to the theory and classical operations of a VAT system but also to the variations employed in different countries.

It should be noted that adjustments to a country’s VAT law may be a direct result of influential lobbying, and the systems and operations of the VAT may be tailored to accommodate these representations. When examined closely, the particular operations employed in that country may thus in fact have little, if any, relevance to your own country.

The danger, in my view, of the paper’s suggestion is that, without some general background, the persons selected for training may be inclined to adopt wholesale the systems to which they have been exposed. If, on the other hand, these persons had attained a wider knowledge of VAT systems they would be in a better position to understand fully the background, experiences, and points of view of any foreign consultants and thus be able to examine their proposals constructively and ensure that the consultants deliver what is appropriate for the country introducing the VAT system.

A point which I would like to share with you and which is not mentioned in the paper, but from which we benefited a great deal, is that it may be worthwhile to also visit a country whose VAT system, to all intents and purposes, was not successfully implemented. Such a country may have experiences to share and might give insights into what mistakes should be avoided.

Treatment of Stocks on Hand at VAT Introduction Date

I would now like to highlight some of the practical legal issues of the VAT system that administrators also have to deal with.

Most countries, prior to the introduction of a VAT system, have in place some type of indirect tax that will be eliminated with the introduction of VAT such as the purchase tax in Trinidad and the manufacturers tax in Canada. These taxes are usually imposed at a multiplicity of different rates on various commodities. One of the major decisions that will face policymakers and administrators during the design of the VAT system is whether to give some kind of relief to traders who, on the commencement of the VAT system, have stocks on hand that have borne the existing tax. Most countries agree with the very valid arguments cited by traders, that to give no relief for such stocks would mean that on the implementation of the VAT these stocks will bear tax twice and the prices of goods will have to increase. To address this issue some countries introduce complex mechanisms to allow traders to credit against VAT payments the amount of the previous tax that is contained in the stocks on hand. In my opinion, such a system has endless opportunities for abuse and, since purchase-tax-type taxes are cascading in nature, it is also very difficult to determine precisely what percentage of the final purchase price of the goods relates to these taxes.

Trinidad and Tobago did not succumb to this type of pressure, principally because we benefited from the experiences of one of our Caribbean neighbors which had gone that route and paid dearly for it, in that for the first six months of the implementation of their VAT, the country received very little revenue and, as a result, had to keep modifying its VAT system to the point where today it bears little resemblance to a classical VAT system. Another factor that diverted us from this route was that, because of some deficiencies in our purchase tax law, arrears had accumulated and thus it was not considered prudent to give credit for taxes which we were not sure we had collected.

The strategy used by Trinidad and Tobago to avoid having to give such credits was to establish very early that there would be no refunds for purchase taxes previously paid and to advise traders to run down their stocks at the end of the year so that they could take advantage of the lower VAT rate on implementation. Experience also shows that where a substantial change occurs in a tax system, such as the introduction of a VAT system, traders take the opportunity to raise their prices and, thus, in such a scenario credit may be given to the trader with no real benefit accruing to the consumer. It was also felt that, at worst, any increases would be short term and, at best, the market forces and aggressive competition would prevent excessive increases. As far as we are aware, this strategy worked, and, in a report published by the Central Bank of Trinidad and Tobago, it was reported that the increase in the consumer price index attributable to VAT was only 4 percent during January-March 1990. In light of our experiences, I would therefore strongly recommend that, especially in a small economy, one should avoid giving relief for stocks on hand which previously bore purchase-tax-type taxes.

How VAT Should Be Indicated

An issue on which I entirely agree with the paper is the question of how firms should indicate the tax to final consumers. I am of the view that the tax-inclusive approach is simpler and is the better system. Under our VAT law, a trader has a choice and may show either VAT-inclusive prices or VAT-exclusive prices. In such a scenario, the element of VAT is always an issue. Every time someone goes to buy an item where the VAT is not included in the price but is added at the cash register, it is a source of annoyance and the issue of the VAT is brought to the fore on each such occasion. Contrast this for example with the United Kingdom where all prices are quoted VAT-inclusive, so although one is aware that there is a VAT, which incidentally was recently raised to 17½ percent, it is not an issue in the final price.

Level of VAT Rate

Another important question is the choice of a rate for the VAT. The paper lists this as number one in the suggestions relating to the introduction of a VAT and states that, if at all possible, the rate should be kept down to 10 percent when the tax is introduced. This is a very valid recommendation. Trinidad was not able to do this because of revenue considerations since its whole reform package was designed to be tax-neutral and was predicated on a shift from direct taxation to indirect taxation. The first part of the tax reform exercise was implemented in 1989 and involved substantial reductions to the income and corporation tax rates. The second part dealt with the introduction of the VAT in 1990. When the VAT was in the design stage, the consultants had originally proposed an 18 percent VAT to compensate for the reduction in direct taxes and the elimination of certain indirect impositions, but word of this high rate caused such a hue and cry from the population that policymakers were forced to re-examine the rate. Taking into consideration that revenue had already been forgone with the reforms on the direct tax side, the lowest rate that could have been sustained was that of 15 percent. On hindsight, however, I would recommend that, where VAT is to be introduced as part of a package, the VAT rate should be decided on first and any other reforms adjusted to accommodate a relatively low rate of, say, 10 percent. Experience will show that taxpayers soon forget the benefits they received from a reduction in income and other taxes, but they constantly have to face the rate of VAT set, and when this is too high it becomes a source of contention and inhibits general acceptance of the VAT system.

Other Implementation Issues

Further to the paper’s point on our bimonthly filing of VAT returns, our computer system was designed to allocate registrants randomly to alternate periods. If one is contemplating a similar system, care should be taken to ensure that registrants who make significant contributions to the VAT revenues are not all placed in the same reporting category as this could result in an imbalance of revenue flows.

There is also sound merit, as the paper suggests, in creating a different name for the tax. A new name affords you the opportunity not only to distance yourself from any adverse connotations that the name VAT may carry, but also to create the perception that this tax is unique to your country. Reference is made to the goods and services tax of Canada and New Zealand and the proposed general consumption tax of Jamaica. Since Trinidad and Tobago had already been using the term “VAT,” it was decided that there would be no advantage in changing the name at the implementation stage.

Problems Outstanding

The paper points out that our nonfiling or delinquency rate remains relatively high. Several factors contributed to this situation initially and we are only now in a position to correct them effectively.

The noncompletion of the design of our computer system has not only seriously hampered most of our operations but also has had dire effects on our enforcement procedures. A number of manual checking systems had to be introduced to counteract these deficiencies, but, as expected, these were a lot slower and thus the monitoring function was not as efficient as it should have been.

Further, because most of the offenses under our law, such as failure to file a VAT return, had to be prosecuted through the Magistrates Court, this was a lengthy procedure and we have only recently been able to obtain some judgments against persons who failed to file their returns on time. As a result of the latest amendment to the VAT Act (Finance Act, 1991), however, the Board is now able to impose administrative penalties for some of these offenses. It is anticipated that this measure will go a long way in lowering the delinquency rate since monetary penalties can now be imposed automatically for these offenses.


Finally, I would like to give a word of encouragement to those countries who are at present in the throes of implementation and are getting the full brunt of bad press and public outcry. The experience has been, in a number of countries such as New Zealand and more recently in Canada, that once the system is implemented the majority of these criticisms quickly subside. We have a local saying in Trinidad which is applicable to this situation; we say, “it boils down like badji.” (Badji, otherwise known as spinach, is a green leafy vegetable which, in its raw state, is very bulky but during the cooking process shrinks considerably.) This was also our experience, and, provided that certain key elements of your legislation are perceived to be acceptable, I am sure that this will be your experience as well, and after implementation you can quickly get down to the serious business of administering the system.

In this regard, I would like to extend our cooperation and goodwill to any country that may wish to avail itself of the knowledge and experience gained by Trinidad and Tobago in the implementation of its VAT system.

Alberto H. J. Radano

According to the paper by Due and Greaney:

  • Total indirect taxes accounted for about 32 percent of tax revenues in Trinidad and Tobago.

  • About 600 manufacturers paid purchase tax; the Customs and Special Taxes Department was responsible for administering and auditing it.

  • Oil and its derivatives accounted for two thirds of total revenues from excises.

  • While oil was a substantial source of revenue, scant interest was shown in reforming the indirect tax structure; as world oil prices declined, the Government turned its attention to introducing a general sales tax.


Several studies were conducted for introducing a broad-based sales tax. In fact, the Prime Minister and Minister of Finance had already referred to that tax in his presentation of the 1983 budget and the IMF referred to the same tax in a report dated May 1983.

In 1986 a committee chaired by Mrs. Michal Christian was established to prepare a reform plan that would include a general retail sales tax. The committee concluded that the tax rates would have to be very high to generate revenues equal to that obtained from the purchase tax, a fact that made its introduction hardly feasible.

In 1988 members of the business community brought a representative of Price Waterhouse & Co. to Trinidad and Tobago. Following a meeting attended by him and representatives of the private and public sectors, a report was drafted on the adoption of a value-added tax, specifying that the business community was in favor of introducing such a tax.


The groundwork for the tax reform was launched in 1988 and involved three stages: (1) Review of the current tax system and preliminary recommendations for reforming it; (2) the development of analytical models for assessing the changes to be introduced in the tax system; and (3) recommendations for a comprehensive reform program.

The basic proposal for change involved the introduction of a general sales tax, using the value-added method. The tax would apply to activities generating an annual turnover of over US$12,500 (the amount was ultimately set at US$30,000) and would have a single rate.

The name finally given to the tax was “value-added tax” (VAT). The responsibility for monitoring it was assigned to the VAT Administration Office within the Department of Inland Revenue, which has three local offices located in the country’s major cities (Port of Spain, San Fernando, and Tobago).

A detailed work plan was prepared. This included an intensive information campaign. The law, which came into force on January 1, 1990, requires that returns be filed and payments made bimonthly.


During the first year in which the new tax was levied, the results were generally positive. Revenues for the first six months were equal to the estimates made. As of December 1990, 9,688 enterprises had registered. On the other hand, 24 percent of registered taxpayers failed to file returns. Perhaps the main shortcoming was the failure of the computer system to carry out all the desired operations. The audit program was also inadequate.


While it is extremely difficult to analyze a country’s tax reform experience without visiting the country and witnessing the prevailing situation, a few comments can nevertheless be made.

According to the paper, all measures usually suggested were taken to introduce the VAT. Thus, the experience of other countries was assessed. International consultants were recruited who have contributed experience gained in countries in which they had worked. Exhaustive studies were carried out regarding the methodology to be applied. Computerized microsimulation models were applied to study the impact of alternative tax policies on collection and income distribution. The situations of specific economic sectors and activities were studied. Finally, detailed programs were launched to publicize the new tax.

A single rate (15 percent) was adopted, which simplifies administration of the tax. Tax administration was entrusted to the VAT Administration Office which, although a separate unit, is part of the Department of Inland Revenue. This decision appears to be more reasonable than assigning such responsibility to the Customs and Excise Department.

The failure of the computer system to carry out all the desired operations is a source of concern. Efforts should be stepped up to identify the causes of this failure and to make the necessary corrections or to change the design of the system for monitoring filing and payment requirements. It would appear that the number of enterprises registered is sufficiently low to permit the use of methods similar to those applied in some other countries (for example, Argentina, Bolivia, and Uruguay) for monitoring compliance with filing and payment requirements by large taxpayers.

The establishment of inspection offices that could each handle 5,000 taxpayers, receiving and checking information from returns and payments in real time, would make it possible to know which taxpayers were delinquent the day after every deadline. This option should be studied by the tax authorities of Trinidad and Tobago.

The alternative “plans” for a simplified accounting method for the VAT stem from the British VAT. They are complicated and unnecessary, and illustrate how care must be taken when incorporating technical features from other countries in which they may operate smoothly, only to lose their effectiveness when transferred to countries where they are inappropriate.

The decision that returns should be submitted on a bimonthly basis may benefit the tax authorities because it halves the work load involved in receiving and processing returns and payments. It may, however, also result in financial problems, particularly in the event of high rates of inflation. The widespread trend in CIAT countries is to collect indirect taxes in the month following that in which the taxable transaction takes place.

The option given to enterprises selling to unregistered persons of itemizing the tax on sales slips or of including it in the price is not recommended. Ideally, there should be a single method so as to avoid confusion that might encourage tax evasion.

Twenty-one percent to 27 percent of total revenues from individual income taxes, and 37 percent to 75 percent from all income taxes.

The experience is described by Roy T. Gobin, “An Analysis of the System of Sales Taxation in the Caribbean Common Market,” Public Finance, Vol. 35, No. 2 (1980), pp. 269-88, and “The System of Indirect Taxation in the Caribbean Common Market,” Bulletin for International Fiscal Documentation, Vol. 33, No. 6 (June 1979), pp. 252-57.

Report of the General Sales Tax Committee, by its Chairman, Mrs. Michal Christian to Mr. Randolph Kong, Chairman, Board of Inland Revenue, September 17, 1986.

Draft Report on the Introduction of a Value Added Tax, 1988.

The models are called “micro” because they use a stratified sample of income, expenditure, and demographic data for individuals or businesses as the foundation for the analysis. The effects of alternative tax policies are calculated for each individual or business in the data base and then aggregated to obtain the results for the entire country. This method of analysis provides aggregate results for the entire country as well as results by various categories of taxpayers (for example, by income class and economic activity). It also provides information on the number of tax units experiencing increases and decreases in tax liability because of changes in the tax law.

The members of the team would like to express great appreciation to T. Boodoosingh of the Budget Office and Ken Superville, Superintendent of Enforcement for Customs and Excise, for their assistance.

The experience with developing the proposals for change is reviewed in the paper by Michal Y. Christian, VAT: The Trinidad and Tobago Experience. It was presented to the Caribbean Community Secretariat’s Seminar on Value Added Tax held in Barbados on June 18-22, 1990.

United Nations Development Program, Human Development Report, 1990 (New York: Oxford University Press, 1990).

    Other Resources Citing This Publication