11 Privatization of Tax Administration

Richard Bird, and Milka Casanegra de Jantscher
Published Date:
September 1992
  • ShareShare
Show Summary Details
Luis Fernando and Ramírez Acuña 

To speak of privatizing tax administration is risky, despite the winds of reform blowing in that direction in this early part of the 1990s. At first sight, the title of this study may possibly suggest a major contradiction in terms. The administration of a country’s tax system, along with the administration of justice and national defense, are probably the most frequent examples of what ought not to be privatized. Or at least they ought not to be privatized in the sense that, being functions inherent to the state, there should be no loss or delegation of autonomy. However, this does not mean that certain administrative tasks should not be contracted out to the private sector where this can be clearly shown to improve efficiency or effectiveness. The difference is that the state loses no autonomy if certain administrative activities are delegated. A company or a person is not free to establish a tax; this must be done by law or by a government decision.

Consequently, the topic is approached from the standpoint that the tax laws are within the sole purview of the state and come into being as a result of actions taken by the competent authorities. Even so, it is worth noting that countries are increasingly entrusting the construction of certain public works to private sector entities and authorizing them to collect certain taxes on that account. Thus, for instance, it is common now for the construction of a highway or a harbor to be awarded through public bidding to private entities (national or foreign) on condition that they make the initial investment and then collect the tolls, taxes, or improvement charges needed to pay for building the project.

Looked at from this angle, government divests itself of (privatizes) certain functions related not only with collecting taxes but also with building public works. To some extent, these are examples of privatization of tax and budget execution at one and the same time. The important point to note is that these arrangements are always made without the state losing its power of supervision. In other words, the private sector cannot decide on its own that it is going to build a highway and collect tolls. It is the state that adopts the decision and awards a concession to the private sector, under certain conditions and usually for a predetermined length of time.

One of the most important lessons we have learned in recent years may be that in order to have a strong state we do not necessarily need a large state. Indeed, in some cases a large state cannot be a strong state. It is for this very reason that, in the area of tax administration, having government entrust some tasks to the private sector instead of directly discharging them may lead to higher efficiency and effectiveness. In other words, such a measure may help build a smaller but stronger tax administration.

I. Tasks That May Be Assigned to the Private Sector

To examine the feasibility of contracting out certain tasks to the private sector, it may be helpful to classify each task under one of the three basic types of work performed by a tax administration, namely: (1) collection of tax receipts; (2) audit; and (3) collection of delinquent taxes.

Collection of Tax Receipts

The possibility of subcontracting some tasks involved in tax collection depends on the type of tax being administered, for example, whether it is a mass tax or one applicable only to a few taxpayers, whether it operates under a self-assessment system or must be assessed by the tax authority.

When dealing with a system of taxes massively applied and based on self-assessment, the following tasks may be contracted out to the private sector to a greater or a lesser degree, as has in fact been done in several countries.

Printing Tax Return Forms

The forms used for tax returns are frequently changed—nearly every year in some countries—either because the laws change or simply because an effort is made to streamline procedures. Some tax administrations purchase the equipment needed to set up their own printing shop directly within the tax agency. Other countries have a government printing house where all printing by government agencies must be done. Some—Bolivia, Argentina, and Colombia, for example—have their own official printing houses but turn to private companies in order to print their tax forms.

Experience suggests that the tax department should be permitted either to print its forms in its own printing shops or contract the work out to the private sector. As a rule, private companies have more flexibility than public ones when it comes to printing and designing the layout of forms, as well as making last-minute changes in them, always according to the instructions received from the tax authorities.

Distribution and Sale of Forms

In most countries, distribution of tax return forms takes place as follows: (1) the forms are picked up by taxpayers at tax administration offices, as in Argentina and Bolivia; (2) the forms are mailed to known taxpayers, as in the United States and Uruguay; (3) a specialized company is hired, usually the same company that prints the forms, to take care of their distribution, as in Colombia (although these specialized companies may in turn use the mail); (4) the forms are available at banks or may be purchased in stationery shops and bookstores. There are also combinations of these options. Another way, which has been used sporadically in Peru and other countries, is to distribute the forms in dailies or other newspapers.

Another issue related to tax forms is deciding whether the taxpayer should pay for them or not. Most countries have chosen not to charge for the form because this delays the distribution process, and have instead absorbed the total cost as an expense of tax administration. In Peru, however, the taxpayer is charged for the cost of the form. In Brazil, banks must bear the printing cost of forms used for payment of taxes through the banking system.

The option of distributing the forms to each taxpayer by mail, or of using for that purpose the services of a specialized company, makes it possible to send out each form with the data on the taxpayer preprinted on self-adhesive labels, thereby making it more likely that the “inputting” of the data after the taxpayer has filled in the information will include fewer identification errors capable of hindering subsequent data processing. If the preprinted form is sent out showing the name, the address, and the identification number of the taxpayer, a lower margin of error is guaranteed when the data are later recorded and faster data entry can be achieved.

Receipt of Forms Filled Out by Taxpayers

The receipt of tax returns filled out by taxpayers is one of the most delicate areas of tax administration. However, private companies have recently been entrusted with this task in some countries.

In countries such as the United States or Canada taxpayers usually mail their tax returns to the tax administration. In nearly all Latin American countries, however, taxpayers deliver completed tax forms directly to the offices of the tax administration, basically for two reasons: (1) there is a generalized perception that the mail does not work properly and (2) there is also the legal tradition of letting the taxpayer have a stamped copy of the form or a special receipt to show that the return has been filed.

In practice, accepting delivery of the forms at the offices of the tax administration has been a difficult business because it is a cyclical event for which the tax administration does not have the proper infrastructure. In recent years this task has been increasingly delegated, especially to banks, which are accustomed to receiving documents, waiting on the public, and engaging in a large volume of transactions every day. This has been the case in the last five years in Bolivia, Chile, Colombia, and Ecuador. As a rule, banks have been authorized to do this work. In principle, not only banks but also other entities in the financial sector can be enlisted for this purpose.

The receipt of tax returns is a delicate task because it sets in motion the process that will lead to accounting entries, collection, assessment of penalties, audit, rulings on appeals or claims filed by the taxpayer, and so on.

A first obstacle to be surmounted if a decision is made to have the private sector (banks) receive tax returns is the requirement of confidentiality.

In most tax systems the information set out in tax returns is confidential and may not be used for purposes other than tax determination. In some countries the information provided by the taxpayer may not even be used as evidence against him. Thus, it is conceivable that no penalties might be imposed for illegally obtained income so long as that income has been reported by the taxpayer.

Confidentiality of the information reported on tax returns is crucial for the success of a self-assessment system. However, when banks are placed under a duty to preserve the confidentiality of the data disclosed by the taxpayer and are barred from using that information for other purposes, the taxpayer usually finds that his banker is someone who, in view of their credit and commercial relationship, may be trusted with the information set out in the tax return.

Another major point to be addressed if the receipt of tax returns is to be contracted out is that banks must be under an obligation to receive returns submitted either with or without an accompanying payment. Filing of the tax return is an act wholly separable from payment, though the two may take place simultaneously. It is desirable to induce the taxpayer to report and pay taxes at the same time, so as to reduce the volume of operations; however, if the taxpayer wishes only to file a return and pay the tax later on, he should have that option and banks should have to accept his tax return.

It is important to specify in the contract with banks their obligation to receive from taxpayers tax returns and payments of all kinds, irrespective of the amount of taxes reported or paid (large or small taxpayers) and of whether they are or are not customers of the particular bank. This will give the taxpayer more options when filing his returns and will encourage competition among banks to provide better service.

Also advisable, as will be seen later, is that the compensation paid to banks should reflect the number of tax returns received by each bank and not merely the amount of taxes collected, so that banks may not be tempted to accept tax returns and payments only from “large” taxpayers.

When banks are given the task of receiving tax returns, one point that should be amply clear to the tax administration, to the banks, and to the taxpayers, is that the stage of filing returns should not involve any verification activities. In other words, the bank should not be given the authority to challenge the data submitted by the taxpayer; it should make no review other than to check the identification number and other information needed to make certain that each taxpayer’s payment will be posted correctly.

Processing Tax Return Data on Magnetic Media

It is not conceivable now that a system of taxes massively applied and based on self-assessment can be administered without the aid of computers.

One of the most serious difficulties faced by tax administrations in recent years has been their limited understanding of strategies in the area of automatic data processing. At first, most countries designed a centralized system with large data processing centers, the center being sometimes a unit of the tax administration and other times an independent agency. Later, decentralized systems were attempted, and more recently three countries, Colombia, Bolivia, and, later, Ecuador, have chosen to assign the “data entry” function to the banking system, reserving for themselves the processing of the data. In other cases, Chile for instance, a specialized agency (a service company) is employed to enter the data.

The availability of good, systematized, and timely information does not depend on who does this work but on the fulfillment of certain requirements. These are (1) a tax identification number that works properly and includes, insofar as possible, a check digit; (2) forms that are simple to complete and that call for relatively few data to be entered into the computer; and (3) a computer system within the tax administration that is capable of interacting with the system in banks or other entities commissioned to do the work. Needless to say, the banking system must have an infrastructure suited to the efficient performance of this work.

One of the advantages of commissioning banks to transcribe the data onto magnetic media is that it helps to unclog the bottleneck usually experienced by the data processing centers of tax administrations; each bank is responsible for processing only the tax returns received by it. This advantage should translate into a requirement of timeliness for banks transcribing the data and sending the magnetic media to the tax administration along with the tax returns. A time frame not exceeding 15 days after the tax return has been filed is a reasonable time for a payment or a tax return to have found its way into the computer of the tax administration.

Some basic quality control over the transcription of data by banks is imperative. Banks should be required to transcribe data twice, or to use some other system that guarantees a low margin of errors. Furthermore, a system of penalties must be devised for deficiencies in the quality of the data transcribed; banks should be penalized by means of monetary charges or a lowered compensation when they make mistakes beyond a certain margin of tolerance. It is also essential for the tax administration itself to design its own data entry system, which ought to set the standard and be open to no changes on the part of any bank. Only the tax administration should be authorized to change it.

One of the disadvantages of entrusting the task of data entry to banks may be that, over time, the ability of the tax administration to make structural changes in the design of forms or in the procedures for receiving tax returns is reduced, inasmuch as any such changes would entail not only retraining the staff of the administration itself but retraining or requiring the retraining of bank personnel.

As in the stage of filing returns, in the transcription stage the bank should not be authorized to engage in any verification activities. The bank should transcribe the data as submitted by the taxpayer even if it detects errors by the taxpayer. In some cases, to gain time, data entry programs have been designed jointly by banks and the tax administration, with a view to detecting arithmetical errors (addition, subtraction, multiplication, division, percentages, and so on) to be able to ask the taxpayer to correct them, but without any coercive power being exercised in the process. In practice, this could be an “ex ante” control offered by the bank as a service to customers.

Receipt of Cash, Checks, Securities, or Other Means of Payment Allowed for Payment of Taxes

Although many tax administrations still have their own infrastructure for receiving tax payments, the trend is clearly in the direction of transferring this function to banks, which do such work routinely and have the installed capacity needed to perform it.

If this function is entrusted to the banking system it may be discharged simultaneously with the receipt of tax returns, or separately. In some countries, payment may be made at the banks but returns must be filed at the tax offices.

Some administrations have chosen to receive payments from large taxpayers directly at their offices (as in Peru and Uruguay) or at an official bank or a national bank (as in Argentina and Bolivia); payments from other taxpayers are usually received throughout the banking system. All manner of combinations are now being used. However, if compensation paid to banks is linked to the amount of tax payments they receive, then banks would not be interested in joining the system if payments from large taxpayers are excluded from the system of payment through the banks.

When a tax administration that has been collecting taxes directly at its offices decides to let banks do this work, there is a strong tendency to give the taxpayer the option of paying at one or the other location. This, however, is seldom advisable for two reasons: (1) for an administration to preserve its collection infrastructure while it lets banks take over this function means forfeiting the possibility of using the personnel from that infrastructure for more pressing or important business, such as enforcement activities, and (2) if “input” problems arise in the processing of data it becomes hard to tell with any degree of certainty in which system they originate. For this reason, once the decision is made to let collection take place through banks, this work should be entrusted to them exclusively; in other words, the tax administration should relinquish this function and should redefine the duties of staff formerly assigned to do that work.

Thought should be given to the possibility of letting banks accept cash or other widely used means of payment. Insofar as possible, no additional requirements should be placed on personal checks—such as prior clearance or certification of funds—because these hinder the work considerably. If a personal check is returned because of insufficient funds, the payment is simply regarded as never having been made, aside from any penalties that may apply. The tendency to place additional requirements on personal checks is typical of countries where there are mechanisms such as the paz y salvo (tax clearance certificate), which is issued against the tendering of a check, even though the check may later bounce.

In designing procedures for filing tax returns and making tax payments in banks, one key element to be considered is the control that must be exercised over banks with respect to their turning tax receipts over to the tax administration (or the treasury). An important lesson is that there must be penalties (interest on arrears) for the late surrender of tax monies collected. This control must obviously be coupled with control of the tax returns filed.

Furthermore, timely updating of taxpayer data in the computers of the tax administration is essential if bank errors are to be detected in time. Most bank errors can be detected when the tax administration sends a notice to the taxpayer advising him that his tax return or tax payment has not been received. When dealing with taxes for which the administration has no single register of taxpayers, the system for monitoring banks must be different: for instance, asking certain taxpayers to check their tax payments, as recorded by the tax administration, against what they have actually paid, based on their own records.

Withholding at Source as a Way of Privatizing Payments

Withholding at source has been one of the most revolutionary steps taken by tax administrations in their efforts to privatize, by making agents in the private sector responsible for withholding taxes and turning them over to the government. It has turned into an assignment given to the private sector to collect taxes in advance.

Although the advantages of withholding as a form of privatization of tax administration tasks are countless, I have excluded this topic from this review because it deserves a separate study. However, attention should be drawn to a side effect caused by the increasingly widespread use of withholding at source, namely, the additional simplification inherent in reducing the number of persons who file taxes without reducing the number of taxpayers. In some countries, the number of tax returns to be processed fell by as much as 50 percent after salaried employees were excluded from the obligation of filing returns, thanks to a legal provision under which the tax they owe is equal to the amounts withheld from their salaries.

Furthermore, it is evident that withholding at source has been generally effective in reducing tax evasion. Some experts have nevertheless voiced the concern that this tool is being used as a cure-all without full awareness of the fact that a widespread system of withholding at source requires monitoring of withholding agents as well.

Tax Refunds

As withholding at source becomes the rule, the volume of tax refunds tends to rise, which is not bad in itself so long as there is an efficient tax refund system in place. A speedy system of tax refunds is conceivable provided the data on taxpayers are reliable. Thus, banks themselves could refund the surplus or balance in favor of taxpayers after the tax returns of, say, large taxpayers, have been filed. Banks would turn over to the tax administration the net tax receipts after deducting the amount of tax refunds. This, of course, would not be in the nature of a definitive reimbursement and would not bar the tax administration from subsequently looking into whether the tax refund was properly made.

This sort of swift, direct refund is conceivable only for large taxpayers or for other taxpayers who take out a bond or insurance policy from the same bank or from an insurance company in order to guarantee to the tax administration that the proper amounts will be reimbursed to it if, upon review, it is shown that the taxpayer was not entitled to the refund.

When it comes to tax refunds, however, most administrations prefer to be cautious and in most cases make a review to establish the amount of the refund. In some countries, however, the taxpayer has to wait longer than a reasonable length of time for his tax refund, while the matter is being reviewed. This is more serious when the amounts to be refunded are used by the treasury to finance its own operations, and there are no provisions allowing indexing of late refunds or payment of interest on them.


The most controversial side of private sector involvement relates to tax audit, with good reason. Deciding which taxpayers are to be audited, for how long, and in what ways, is properly the responsibility of the tax administration, for the fulfillment of which it is granted discretionary authority, the power to impose penalties and the power to enforce the law.

The most significant step taken to privatize the audit function may well have been the introduction of self-assessment. From a historical perspective this was an enormous innovation, the changeover from a system in which the tax and the taxable base were established by the tax authorities, to one in which both are calculated by the taxpayer himself. Under this system the taxpayer must not only pay the tax but must determine it, compute it, and report it himself; if he should fail to do so, penalties will be applied. The tax authorities confine themselves to reviewing the reported data and, of course, computing and collecting underpayments as well as any applicable penalties.

Some taxes, though they may be massively applied, require that the tax base be computed by the tax administration, as in the case of real estate taxes for which an official valuation of the property is made in order to calculate the tax. Some administrations are assigning private companies the task of measuring the tax base by using objective means, such as aerial photography, to determine the size of properties; others are trying out self-assessment by the taxpayer himself, and a trend toward self-assessment of property values is beginning to emerge in countries such as Bolivia and Colombia. Such self-assessment and reporting by the taxpayer may have positive results where the law provides (as in the case of Colombia, for example) that any entity in the public sector may, by a unilateral decision, acquire any piece of real estate for the value reported by the taxpayer plus 25 percent to cover any underassessment on the part of the taxpayer. The possibility that the government can buy the property at the price reported by the taxpayer cannot be regarded as confiscation because the price is set by the owner. So long as this possibility is considered credible, the outcome may be that self-assessment of real estate will approach real values in order to forestall the possibility of having to sell for a lower price.

In the area of customs duties, some countries such as Bolivia and Indonesia have hired specialized companies to check the value of imports, owing usually to the absence of proper control systems at customs. This type of service basically involves checking the value reported by the importer against reference prices in order to establish the duty payable. Countries wishing to embark on such schemes should weigh the costs of the service against the benefits provided.

The audit function, when dealing with massively applied taxes based on self-assessment, may include certain features intended to encourage compliance or to simplify administration, which cannot strictly speaking be considered “privatization,” but which reflect a private sector mentality. Some countries have introduced legislation regulating such features, which merit discussion because they aim at decreasing the audit work load by introducing some elements of a private sector approach to tax administration. Below we discuss some of them.

The Signature of an Accountant on Tax Returns

By requiring the signature of an accountant on tax returns, a good deal of the responsibility for the veracity of the information reported is placed on a professional accountant, thereby making the task of audit easier and improving at the same time the quality of the data reported by the taxpayer.

The obligation to have the tax return signed by a public accountant is usually mandatory for larger taxpayers (companies, businesses, and so on), because they are commonly under an obligation to keep books of account and consequently have an accountant or a tax advisor. The fact that the signature of an accountant is required to file a tax return does not mean that the return will not be looked at by the tax administration. The idea is merely that having an accountant check the return raises the odds against tax evasion. Of course, this requirement is usually accompanied by severe penalties when violations attributable to the accountant or tax advisor are detected, so that he may be said to share to some extent the responsibility of the taxpayer.

Audit Insurance

Audit insurance guarantees to taxpayers who increase their tax payments by more than a certain percentage over the immediately preceding tax period that they will not be audited. This device is usually resorted to on a purely temporary basis (a particular tax period or tax year) when there is an awareness that the tax administration does not have even the minimum audit capacity for reasonable enforcement. Accordingly, the administration makes a first selection and excludes from audit procedures all those taxpayers who register a substantial increase in their tax payments. This policy is announced beforehand, so as to encourage taxpayers to increase their payments by the established minimum and thereby win immunity from audit. To illustrate this mechanism, let us take a country with an annual inflation rate of 20 percent, real economic growth equal to 5 percent (which means that nominal economic growth is 26 percent), and a tax system with unitary elasticity (which would imply that nominal tax revenues should increase by 26 percent). The audit insurance would then be guaranteed to taxpayers who increase their tax payments by, say, 36 percent as a minimum. Actually, audit insurance may—and insofar as possible should—be set up on a differential basis, that is to say, in line with the behavior of relatively homogeneous economic sectors.

The problem with this procedure is that taxpayers whose potential tax increase is higher than what is required for giving them insurance against audit will tend to keep their tax payment to the minimum needed for immunity from control. Even so, the idea is that the administration should tolerate these possible violations for the sake of being able to devote its limited resources to auditing those taxpayers who fall short of the minimum required for audit insurance.

Expert Certification

Certification by an expert is used, for example, in Mexico, and goes farther than requiring the signature of a public accountant on tax returns. The taxpayer who files a return with a tax certification from a public accountant has, in the eyes of the tax administration, a “guarantee” that his return meets current tax regulations. Accordingly, any liability for inconsistencies found in the return rests with the accountant who signed the tax certification, the taxpayer escaping any penalty for mistakes that can be laid to the former.

Two essential characteristics are noteworthy in this procedure: (1) the system is voluntary; in other words, it applies only to taxpayers who wish to avail themselves of it, in which case they must bear the cost of tax certification; and (2) besides being voluntary, the system allows the taxpayer to choose his own accountant, which is meant to ensure the confidentiality of the information.

Legitimizing Previous Tax Returns

A mechanism that has already been introduced in the legislation of several countries—the Dominican Republic being probably the first—guarantees to the taxpayer that his tax returns for earlier tax periods will not be audited so long as his most recent return is correct. It is certainly attractive for the taxpayer to have the option of barring audits for earlier tax periods. For this option to operate properly and protect him from audits, however, he will in practice always have to file correctly in the future. Otherwise, the possibility of auditing him will apply once again. If the goal of having the taxpayer file accurate returns is attained, the tax administration will have gained, because ensuring proper future compliance is more important than preserving the possibility of reviewing the past.

Reducing Penalties if the Taxpayer Does Not Appeal

In line with the idea of simplifying the work of tax administration and improving compliance by taxpayers, legislation has been enacted in some countries to discourage controversies between the government and the taxpayer, without, however, disregarding due process if the taxpayer believes he is entitled to challenge a decision of the tax administration. The taxpayer is given the right to a gradual reduction of his penalty in proportion to the reduction in the work that the tax administration or the courts have to do in connection with his case.

The idea of gradually reducing penalties may be illustrated by the following example:

(1) Let us suppose that a taxpayer has underreported certain incomes and thus underpaid $1,000 in taxes.

(2) Let us further suppose that the tax administration is able to detect 25 percent of all unreported income, which in this case would mean $250 in taxes (in many tax administrations this ratio is substantially lower).

(3) Finally, let us suppose that the penalty applicable to the taxpayer is 200 percent of the amount of the unreported tax detected, in other words, $500 ($250 x 200 percent).

In the end, the administration would be able to collect a total of $750 ($250 in taxes and $500 as the penalty). Of course, after imposing the penalty the tax administration would probably have to institute proceedings for collection, particularly in view of the stiff penalty involved.

A system that substantially reduces penalties might succeed in collecting a higher amount. The outcome will naturally depend on the assumptions made in the example set out below. The purpose of that example is to illustrate what kind of results might be expected from a strategy of reducing penalties while increasing the resources devoted to audit.

(1) Once the unreported income is detected by the tax administration, instead of imposing the penalty described above the administration allows the taxpayer to correct all his errors, holding out the possibility of reducing the penalty to, say, 20 percent in lieu of 200 percent. The condition for reducing the penalty is that the taxpayer must actually pay the unpaid tax, along with the reduced penalty.

(2) If the taxpayer calculates the odds and does not know exactly how much of his unreported income has been detected by the tax administration, he will probably correct his tax return provided the penalty is reduced considerably. It is important to point out that at this stage the administration should let the taxpayer know that it has evidence regarding the unreported income, without, of course, telling him what sort of evidence.

(3) If we assume that the taxpayer declares the entire unreported income, the penalty will have been reduced to only $200 (20 percent of the unreported $1,000), but on the other hand, a tax of $1,000 will have been recovered. In other words, the tax administration will actually collect $1,200 and will save itself collection and litigation proceedings against the taxpayer.

(4) If we assume that the taxpayer does not—as is likely—declare all of the income initially unreported, the tax administration will nevertheless collect the portion accepted by the taxpayer, plus the reduced penalty, but in this case it will retain the option of imposing the 200 percent penalty on the tax payable on the income that the taxpayer has failed to include in his amended return. In other words, it is preferable to give the taxpayer the opportunity of correcting his return and of changing his behavior, with the least possible expenditure of resources on the part of the tax administration.

(5) On the penalty imposed by the administration (for the uncorrected portion of the taxpayer’s return) a smaller reduction should be offered—say a 40 percent penalty instead of the 20 percent penalty mentioned above—against the taxpayer’s commitment not to challenge the administration’s version of the facts and to pay the tax due plus the penalty thus reduced. This would discourage litigation and would show taxpayers that the longer they persist in trying to put off the penalty through evasion or litigation, the costlier it will be for them.

Through this type of measure, combined with broader audit coverage, some countries have successfully put in practice the notion that pursuing improvements in compliance by all taxpayers is better than achieving perfect compliance by a few or exhaustively auditing a handful. From a different standpoint, this policy reflects a private sector attitude in that the less litigation there is the more resources will be freed for the administration to reinforce audit work.

For the policy of reducing penalties to be successful, however, the coverage of audit (ratio of taxpayers audited to total taxpayers) must be expanded; in other words, the likelihood of detecting tax evasion will need to increase.

Requesting Information from Third Parties

The attempt to control tax evasion has led tax administrations to require from taxpayers additional information concerning their financial or commercial dealings, in order to make sure that taxes have been correctly assessed. In addition, both taxpayers and nontaxpayers must provide the administration with information intended for checking the accounts of other taxpayers. This is a way of privatizing control, by relying on third parties for information used for audit purposes. To expedite the process of matching the information received with information reported by taxpayers, it is advisable that the data be supplied on magnetic media.

In the last two decades tax administrations have tended to ask for too much information, especially when they began to use computers. At present, many administrations are rethinking this strategy, after learning that having the computers and asking for the information was not enough to control tax evasion. In fact, taxpayers themselves were the first to realize that by giving the administration information “indigestion,” its attention would be distracted over purely formal matters and it would never get to the heart of the problems.

Recently, there has been a widespread trend toward simplifying forms and asking only for such information as the computer is able to process, but providing in the law that the taxpayer is required to keep records and to make them available to the tax administration on request.

Collection of Delinquent Taxes

In principle, it should not be very difficult to turn over to companies or persons (preferably lawyers) the task of collecting the delinquent tax portfolio of a tax administration. The practice is not very widespread, but tax administrations are increasingly considering a move in this direction.

If this task is to be successfully accomplished by the private sector, the tax administration must provide reliable data on delinquent taxes; the legal system must provide for administrative collection procedures instead of always requiring involvement of the courts; the interest charged on delinquent taxes must be higher than the prevailing market rate, so that taxpayers will not be tempted to use their delinquent taxes as working capital; and/or collection costs must be paid by the taxpayer on whose account the collection proceedings had to be undertaken.

Other key points to be considered include the system itself for turning listings of taxpayers over to companies or persons charged with collecting delinquent taxes, the compensation arrangements, and the incentives established to encourage speedy collection. Care must be taken not to turn over the portfolio of large taxpayers exclusively to private companies (in fact, the tax administration should begin by commissioning the collection of delinquent taxes owed by small taxpayers), and care should be taken to evaluating the number of cases of successful collection as well as the geographical location of delinquent taxpayers.

The choice between turning collection over to private companies and paying a commission (a variable salary) to the tax administration’s own officials engaged in collecting delinquent taxes is an interesting one, because in the first case the task is privatized and in the second a private sector attitude is applied. In Argentina, for example, the officials of the tax administration engaged in this work have a variable salary, determined by the results obtained.

Assigning a variable salary to tax administration employees according to the amount of delinquent taxes they recover does, of course, generate debate. Kept within reasonable bounds, however, this policy is beginning to make inroads in a good many tax administrations.

II. Remuneration for Work Assigned to the Private Sector

How to pay for the services of the private sector once a function of the tax administration has been privatized may be even more controversial than the decision to privatize.

Probably the first question that needs to be answered is whether the service can be simultaneously entrusted to several persons or entities or whether it requires exclusivity. In either case, but especially in the latter, the most advisable course of action is to award the contract after calling for bids. In both cases, competition should be stimulated among the various entities providing the service.

When the private sector is called upon to perform a task previously discharged by the public sector, compensation for that service will in some cases be a fixed amount payable by the tax administration under a contract; but in other cases it will be necessary to regulate the amount that may be charged directly by the entities newly responsible for providing the service.

When it comes to receipt of tax payments, and collection of delinquent taxes by the private sector, tax administrations use different remuneration methods. These range from paying for each staff hour of work involved to paying a percentage of the amount of taxes collected.

In the case of collection through the banks, one way to remunerate them is to let the banks keep the taxes collected for a reasonable length of time that will cover their costs. Of course, compensation systems may have some more sophisticated features aimed at stimulating competition among the banks themselves, in order to provide better service to the taxpayer, handle the largest number of persons, and reward processing of good quality performed by the banks.

In some countries, competition has meant that the profitability of these operations has fallen off over time, because banks, in their effort to attract large taxpayers, have opted for sharing their profit with the customer (the taxpayer) by offering him loans at preferential interest rates to pay his taxes, thereby sharing with him the profit derived from the tax collection work.

When the choice is made to compensate banks by letting them use the tax receipts for a certain period, the time specified varies across countries. In Bolivia, Colombia, and Ecuador, for example, this period is about 15 days. In other words, the compensation received by the banks is equal to the rate of interest they are able to obtain for 15 days on the amount of taxes collected. In the countries mentioned, the service provided by the banks includes receiving tax returns, processing the data onto magnetic media, accepting payments, and ultimately turning over, within an equally brief time frame, properly filed copies of the forms received, copies of the magnetic tapes, and the money collected.

Penalties are usually provided for banks that fail to meet their obligations. Aside from the interest applied when tax receipts are not turned over on time, the systems in force in Bolivia, Colombia, and Ecuador impose separate penalties based on the frequency of delay in submitting documents and the number of mistakes made in transcribing the data onto magnetic media. It is advisable to have a penalty-free margin of tolerance and, in particular, more flexible regulations during the early months of implementation of the system, when banks do not yet have enough experience with it. Another useful feature is to reward, through temporary monetary incentives, the bank that registers the highest degree of compliance.

Finally, it is important to underscore that banks should be compensated in such a way as to encourage each one to handle the largest number of taxpayers. In other words, the factors to be considered for setting their remuneration should be not only the amount of money collected but also the number of documents received and transcribed onto magnetic media. Absent this, banks may concentrate on receiving only payments from large taxpayers that involve few documents and large amounts of taxes.

As for the compensation of agents withholding taxes at source, attempts have been made to give them a tax credit or a tax discount, but in practice the normal procedure in countries where the rate of inflation is not very high is to give the withholding agent a reasonable length of time before he must turn over the money withheld, which is the equivalent of paying the agent the rate of interest prevailing during that period.

Turning now to what might be called “taxpayer assistance” or “taxpayer services,” an enormous field opens up for private sector participation. Data base systems that make it possible to make computer inquiries about current tax legislation; software designed for preparation of tax returns; private audits before filing returns; publication of specialized literature; and training seminars on tax matters; these are all examples of private sector activities aimed at assisting the taxpayer in his dealings with the tax administration. Of course, the tax administration will need to step in and provide these services, or share in their provision, where they have not yet been developed by the private sector.

III. The Private Sector Approach to Tax Administration

One aspect that has been stressed in this study is that privatization is not always a way to improve a service that is being provided poorly by the tax administration. Indeed, many tasks are better performed by the tax administration itself acting with private sector mentality, which means investing resources to train employees, improving job conditions, devising financial and other incentives to encourage the best employees, and so on.

It is better to have fewer well-paid employees than a larger number who are poorly paid. The tax administration must pay a competitive salary in order to keep its best employees.

Happily, this private sector managerial approach has begun to be accepted by many tax administrations. It is now common to observe specialized units devoted to large taxpayers, “red carpet” treatment for the largest taxpayers, different management and audit strategies for various types of taxpayers—all of which should not be viewed as a form of discrimination. In fact, discrimination could be said to exist if taxpayers who are different are treated in the same way.

IV. Conclusions

Privatization of tax administration does not mean giving up government control over tax policy in a particular country. Quite the contrary, it means letting the private sector perform certain tasks under the supervision of the tax administration, in order to improve efficiency or effectiveness.

When a tax administration delegates certain logistical functions to the private sector it must have a firm hold on its managerial ranks to enable it to monitor, to introduce any necessary changes in the delegated services, and to react adequately to any problems that may arise. This reaction may include reclaiming for the tax administration any functions that the private sector may be performing poorly at a particular time.

Experiences in privatizing tax administration functions have not necessarily followed a sequence by starting out with tax collection, then moving on to audit, and finally dealing with collection of delinquent taxes. Instead, some of these functions have been turned over to the private sector whenever this was thought advisable in order to improve efficiency or effectiveness.

A policy of gradually reducing penalties while substantially raising the likelihood of detecting tax evaders is preferable to a policy of imposing stiff penalties and exhaustively auditing only a few cases. Stiff penalties encourage litigation and make it necessary to devote resources to fighting the taxpayer in court, rather than to exercising better enforcement.

Differential treatment of large taxpayers has enabled tax administrations to set better priorities in the area of services to the taxpayer and to improve their control over tax compliance.

Imposing penalties on taxpayers, issuing final interpretations of tax legislation, auditing returns to ascertain the true tax base, and reaching settlements with taxpayers in an effort to change their behavior in the area of tax compliance are all functions that should remain in the hands of tax administrations. Even so, certain penalties are beginning to be entrusted to the private sector when they involve a low degree of subjectivity, as when private sector agents collect not only the delinquent taxes but also the interest or penalty owed for late payment, which is standard for all taxpayers.

Most tax administrators complain that they do not have enough inspectors or resources to adequately discharge the audit function. Recently, the possibility has emerged of letting the private sector (basically banks) do a large part of the logistical work involved in processing payments and returns, thereby freeing resources to strengthen audit activities, where the administration is called upon to play a more important role.

Privatizing does not always mean greater efficiency, but delegating certain tasks to the private sector, as some countries have successfully done, offers an attractive alternative for tax administration in the decade ahead of us.


Martinez-Vazquez Jorge

Ramírez Acuña’s paper is interesting and innovative. The advanced state of the privatization of tax administration in Latin America surprises me. Ramírez Acuña has performed a valuable task in guiding us through the current privatization efforts in a number of countries in several areas of tax administration.

My comments are designed as a supplement to Ramírez Acuña’s work and are divided into two parts. I shall start by taking a little more time to frame the privatization process in tax administration within the wider movement to privatize the public sector in the last decade. The fundamental idea here is to see that there are lessons for tax administration in the experience with privatization in the rest of the public sector. The second part of my comments will examine some of the propositions and experiences described by Ramírez Acuña. For reasons of space I shall refer specifically to cases I think should be emphasized or require further study before definitive conclusions can be offered on them. I shall also refer to some other areas of tax administration, in addition to those mentioned by Ramírez Acuña, which I think are suitable for privatization.

Lessons for Tax Administration from the Experience with Privatization in the Public Sector

The nature of the process of privatizing tax administration differs in important aspects from the process of privatizing the rest of the public sector. Analysis of some of these differences leads us to some interesting conclusions regarding the nature, objectives, limitations, and problems of privatizing tax administration.

The impetus for the privatization of the public sector in the past decade, in both developed and developing countries, sprang largely from ideological opposition to the growth of the public sector and to the general belief that the private sector may be more efficient than the public sector, especially with respect to organization and incentives. This privatization of the public sector has taken three distinct forms: complete devolution of the public function, maintenance of control of the public function while turning over implementation or production to the private sector, and deregulation. Only the second form of privatization is relevant for purposes of tax administration. But even here the differences between tax administration and the rest of the public sector are important. The government can in principle transfer to the private sector any kind of activity that represents a quid pro quo or compensation for services. Only the state, however, has the right to collect taxes. This coercive power to require the surrender of income without a direct payback is the characteristic which perhaps best defines the state and as such is inalienable.

These premises provide an appropriate guideline for what should and should not be privatized in tax administration. Tax administration involves a number of functions such as tax collection, audit, and investigation of taxpayers (if they fail to file or have ceased to file), and action in legal procedures and administrative appeals. All these functions include, in one form or another, an information component and a decision component. The latter contains the principle of public authority and must therefore not be privatized. Examples include the (administrative) interpretation of the law in an audit or the execution of an eviction. But all the information components, including the collection of data, documents, and the like, may, from a constitutional or legal point of view, be privatized.

Naturally, some aspects of tax administration could legally be privatized but should not be for reasons of efficiency. For example, from the theory of business organization we know it is more efficient to keep certain activities within the organization and not subcontract them, when the costs of recontracting with third parties—because of, for instance, lack of information or a very changeable context—are high.

Another important difference between privatization of tax administration and of the rest of the public sector is the political element. Often the privatization of certain public sector activities is a result of pressures of a political nature, for example, for the containment or reduction of public spending. Frequently, such proposals enjoy the support of a majority of the population and are promoted by politicians seeking election or re-election. In contrast, we would expect only in very special cases to find the average citizen exerting political pressure in favor of more efficient tax administration or political parties making improved compliance with tax requirements a central theme of their election campaigns. On the contrary, many taxpayers may be more satisfied with tax systems that permit a certain amount of evasion, provided it does not pose a serious threat to basic services. Politicians may favor less tax evasion because they see evasion as working against equitable taxation and they want the revenue. But it can also be argued that politicians prefer systems with more administrative discretion. An interesting conclusion from these observations is that, contrary to the situation in the rest of the public sector, the impetus for the privatization of tax administration has to come from within the public administration itself.

What are or ought to be the goals in privatizing tax administration? In the light of the Latin American experience so far, the most fundamental of these goals seems to have been that of raising the tax authority’s operational capacity to a level which had proved unattainable in the public sector. The need to turn to the private sector in basic aspects of administration seems to be linked to rigidities and other problems in the recruitment and compensation of civil service personnel. From this perspective, privatization can be interpreted as a substitute for an adequate personnel policy. An assumption to be verified empirically is that an inverse ratio exists between the degree of privatization of tax administration and personnel problems in the public sector.

However, it may also be that even when the public sector can recruit and retain employees of the same caliber as the private sector, the latter is more efficient than the private sector. The greater efficiency may derive from better use of resources owing to such factors as incentives and organization. Other goals in addition to efficiency may be considered when evaluating the privatization of a particular tax function. An example of another important goal is that of lessening the taxpayer’s cost of compliance. It might be, for instance, that the cost of collecting the tax internally through its own agencies would be the same for the government as collecting through the private banking system. However, the latter route would be considered superior if it is more convenient to the taxpayer, that is to say, if it makes the process of compliance less costly.

What are the most significant problems we can expect in the process of privatizing tax administration given the experience with privatization in the rest of the public sector? First, where public sector employees are organized in unions and reports of planned privatization emerge, labor opposition to privatization can be expected, especially if privatization actually threatens to cut the number of jobs in the public sector. A second problem may be the low quality of services performed by the private sector following privatization. One way to tackle this problem is to invite bids from service providers or use other means that might increase competition among them and protect the quality of service offered by the private sector.

An additional obstacle to the privatization of certain facets of tax administration may arise from a faulty cost-benefit calculation by the public sector. For example, methodologies frequently used in cost-benefit analysis do not internalize all the costs of the public sector. While most public sector agencies, including the tax authority, receive subsidies of many kinds, private firms typically have to pay all their own costs. Often the costs of buildings, furnishings, and equipment, as well as employee pensions, are ignored in cost-benefit calculations in the public sector.

Corruption can also be an important problem accompanying the privatization of tax administration. This problem may be greater for small countries where the number of businesses capable of performing the privatized tasks is small and those businesses are more likely to be controlled by or have some kind of link to government officials. The history of tax administration, with the practice of awarding the right to collect taxes to the highest bidder—common in the old civilizations of the Mediterranean basin and current in Europe up to the Age of Enlightenment—clearly shows that privatization of tax administration and corruption and abuse of power can go hand in hand. The most important lesson to learn from this historical experience is that the government must not lose control, whether direct or indirect, of the private enterprise providing the tax services. An indirect way of exercising this control is to ensure a certain degree of competition within the private sector. A good example is competition among banks to have taxpayers pay their taxes at bank branches.

An important alternative to the privatization of tax administration mentioned by Ramirez Acuna—and which has also been experimented with in other areas of the public sector—is individual incentive systems. On this I would like to inject one note of caution. Incentive systems might be an effective weapon in attacking the problem of efficiency in the public sector, but only if the public sector is able to attract and retain the same quality of personnel as the private sector. Such systems would be less effective if salaries in the public sector are markedly below those of the private sector. In this case the most fundamental problem is the lack of qualified personnel and not the lack of incentives.

Privatization of Tax Administration in Latin America

Ramírez Acuña’s paper emphasizes, and rightly so, areas of administration that already have experience with privatization. The bulk of the privatization of tax administration in Latin America seems to have taken place in the area of collection. It seems that almost nothing has been privatized in the audit area and nothing in the legal and inspection areas (inspection of nonfilers and of those who have ceased to file). These last two areas are not even mentioned in Ramírez Acuña’s paper. But even with respect to collection, almost all privatization experiments have been in a narrower area, the collection of current taxes. Much less has been done with the privatization of the collection of delinquent accounts. It is not surprising that privatization began in the areas of tax administration that are easier to privatize or less apt to compromise the principle of patria potestas embodied in public administration.

Although a good many aspects of tax collection have been privatized, the giant step taken in recent years in this area has been to transfer to the private banking system the receipt of returns and payment of taxes. Given the unreliable mail service in many Latin American countries, the return and payment traditionally had to be completed in person in offices of the ministry of finance. Frequently, this practice meant a significant increase in compliance cost for the taxpayer. An irony of many Latin American tax systems is still that the refrain “in this country everyone is a tax evader” can be heard while long lines of taxpayers patiently wait their turn for hours to pay their taxes in finance ministry offices. The transfer to private banks of the receipt of the return and collection of the tax has cleared one of the greatest tax administration bottlenecks in many countries. El Salvador, Guatemala, and Panama can be added to the countries mentioned by Ramírez Acuña.

Given the high probability that other countries will adopt similar privatization measures, one lesson to be learned from experience is the need to ensure that the tax authority retains personnel qualified in data processing and has the equipment necessary to verify the tapes of data sent by the private banks. An independent system for checking payments is also important to eliminate the possibility of abuse on the banks’ part. As Ramírez Acuña points out, the banks’ omissions would come to light if the tax authorities were to send taxpayers a notice of failure to pay. However, for this procedure to function properly, a system providing an up-to-date list of taxpayers and their current accounts would be necessary. Unfortunately, this is a problem area for many Latin American tax administrations. In spite of international technical assistance in this area during the last decade, several Latin American countries still do not have an operational individual register or current account system.

Should the treasury turn all tax collection activities over to the private sector? Transferring them to the private banking system, as Ramírez Acuña notes, could free resources for other areas of tax administration. Nevertheless, it may be desirable to keep personnel qualified and experienced in tax collection within the administration.

Another big step in the privatization of tax collection is, as Ramírez Acuña indicates, withholding at source. In addition to making the task of the administration easier, withholding of tax at source has other advantages, for example, decreased tax evasion. Although reporting systems are not perfect substitutes for withholding systems, they have also contributed indirectly to the privatization and simplification of tax administration in some countries. However, reporting systems have seen much less development in Latin America and are not mentioned in the paper. Ramírez Acuña highlights another effect of withholding systems that he considers very important: the possibility of reducing the number of filers without reducing the number of taxpayers. He mentions the example of some countries in which the number of returns was cut 50 percent following exclusion of wage-earners subject to withholding. For these taxpayers the amount withheld is the final tax.

This subject—the waiver of the requirement to file a return for taxpayers subject to withholding at source—merits a little more discussion, given its swift acceptance in Latin American countries. The advantages inherent in the elimination of a universal filing requirement are clear; the most important is that it simplified tax administration. However, there are also disadvantages to this procedure. Aside from the political and philosophical appeal of the universal return and citizen participation, the possibility of increased evasion is perhaps the most important disadvantage. For example, in countries where the personal income tax system leaves it up to the taxpayer to identify and claim various types of deductions and rebates, the elimination of universal filing is an open invitation to falsify deductions. Elimination of the filing requirement also makes it possible for the withholder to defraud the treasury, since there is no independent proof that the tax was paid and that this was done in the amount later filed by the withholding agent. The rejoinder to this line of reasoning has been that, after all, Latin American tax authorities do not have the ability to audit all the returns of wage-earners subject to withholding and that it is unrealistic to pretend otherwise. However, no tax administration is able to audit all taxpayers. It is a known fact that even in the most advanced countries the tax authorities audit only a tiny percentage of the taxpayers and clearly lack the capacity to audit all of them. However, the possibility of being audited appears to play an important role in the “voluntary” compliance of taxpayers. The same logic applies to withholding. One way in which eliminating requirements for filing returns with respect to wages earned may not lead to tax evasion is if the law allows the taxpayer absolutely no discretion: for example, if it grants the taxpayer the right only to a single overall rebate. However, the problem of controlling the withholding agents remains.

In the audit area, Ramírez Acuña correctly points to the self-assessment system as an important step in privatizing this function. The author also notes that some countries have begun to use self-valuation for property taxes. One form of property self-assessment was actually used in Guatemala in the 1987 tax reform, but with little success. The amounts filed represented a very low increase over the real estate values established decades earlier, and a large number of private businesses boycotted the registration and self-valuation process.

Ramírez Acuña also mentions several measures that tax authorities have adopted to facilitate audit activities. These include the signature of the accountant or professional preparer of the return and the fiscal certificate (the preparer not only signs but also assumes responsibility before the treasury for his own errors in computing the tax and interpretation of the legal provisions). Three other measures described by Ramírez Acuña as simplifying inspection—audit insurance, the acceptance of the preceding years if the last tax period is filed correctly, and the reduction of sanctions if the taxpayer does not challenge or appeal the audit—bear similarity to tax amnesties and for that reason are more controversial. It is true that these three types of measures can be more effective than the traditional tax amnesties because they try to change future taxpayer behavior. For example, in the case of audit insurance, the tax authority guarantees to taxpayers that they will not be audited if taxes filed in the current year exceed those filed in the preceding year by a certain percentage. This offer to accept previous years is revoked in the event of failure to meet tax requirements in subsequent years.

Even so, the three types of measures share certain disadvantages with tax amnesties. Their worst feature is that they undermine the public’s confidence in the tax authority. Taxpayers grow accustomed to expecting a new form of amnesty or negotiation with the authority. Another serious objection is that those who have paid their taxes may resent that other individuals with the same or higher incomes are officially permitted not to pay their fair share. It is essential to respect the principle of equity and the perception that everyone must pay his or her “fair” share when attempting to secure a high degree of taxpayer compliance. For this reason, the tax authority must take pains to cultivate an image of impartiality in its dealings with taxpayers. From this perspective I do not believe it is a good idea to give special or “red carpet” treatment to large taxpayers as Ramírez Acuña suggests.

I would like to end these comments with a brief summary of other functions of tax administration that lend themselves to privatization and with which one could experiment. Something mentioned at the beginning of these comments should be recalled here: all functions of tax administration have, in different proportions, a decision-making component and an information or implementation component. Only the first component, decision making, involves the principle of patria potestas wielded by the state as embodied in the tax authority and as such cannot be privatized. That leaves a number of activities in the area of audit which could be privatized. The tax authority could engage domestic or foreign experts to document areas that require technical expertise not possessed by the authority. For example, the authority could engage private actuaries to analyze the finances of insurance companies and prepare a file prior to an audit of such companies. Later, a decision on all legal interpretations, the application of fines, and the like would be carried out solely by the tax authority.

The same logic could be applied to two other important functions of tax administration: the monitoring of taxpayers and legal action. The tax authority could, for instance, contract with the private sector for a survey designed to identify nonfilers. The survey could be limited to certain groups of professionals. Once the nonfilers were identified, the authority, using its own resources, could act with respect to those taxpayers or subject them to a presumptive tax. In the legal area the public authority could engage private attorneys to represent the authority in court cases involving appeals by taxpayers. These two examples provide a glimpse of what a more imaginative view of the privatization process might lead to in the realm of tax administration.

Taxation, like politics, is the art of the possible-yet most public finance texts ignore the critical role played by tax administration in restoring macroeconomic balance and promoting equity and efficiency. This volume fills a gap in the literature by:

  • Linking tax policy and tax administration reform, and

  • Exploring ways to improve taxpayer compliance.

The papers included in the volume were prepared for a symposium sponsored by the Instituto de Estudios Fiscales of the Ministry of Finance of Spain. The editors are authorities on tax policy and administration and have published extensively on tax issues. Richard M. Bird is Professor of Economics, University of Toronto, and Milka Casanegra de Jantscher is Assistant Director of the IMF’s Fiscal Affairs Department.

I am grateful for comments from Richard Bird, Jorge Martinez-Vazquez, and Teodoro Nichtawitz.

    Other Resources Citing This Publication