|
|
||||||||||||||||||||||||
|
Discussion
Forum on “Improving Tax Administration: A New View from the Theory of
Tax Evasion in a Corrupt Regime[1] 14
November 2000 FORUM
HIGHLIGHTS[2] Opening
activities started at 9:25 a.m. Hon. Felipe M. Medalla, Secretary of
Socio-Economic Planning and Chairman of LEDAC Exec. Committee, gave the
opening statement. Dr. Mario B. Lamberte, President of the Philippine
Institute for Development Studies (PIDS), acted as the moderator during
the open forum. Opening
Statement of Hon. Felipe Medalla Let me inform everybody that the revised revenue targets this year will be revised downward again. In view of this, it is really important to do something about tax administration and the issue of tax evasion. Why is it that the BIR collection has been much worse than the performance of the economy? Perhaps we should take a look at our accounting system. It is time, I think, that our accounting system is carefully examined especially in relation to analyzing the relationship between the revenue and the GNP. Suppose that you produce a lot, thinking that it will be sold. However it turns out that it is hard to sell these things so you sold them at prices way below. The problem with the national income accounting is that the output goes up but the demand contracted. But since you already produced the units you will sell them anyway. The accounting system is capturing the outputs and the inputs but what a prim observer sees is that nothing has happened. What happens to the usual relationship between taxes and GNP? We may actually be mistakenly attributing what may be, in further analysis, attributable to a boom-bust cycle, to the worsening performance of the BIR. There must be way of evaluating tax administration and evasion. Unfortunately, the only measure we have now is the ratio between the collection of BIR and GNP. We must focus on improving tax administration because a substantial portion of this so-called market-friendly-pro-poor agenda cannot be delivered unless tax administration is improved. We are already behind the basic social services and infrastructure targets of Angat Pinoy 2004. Tax collection is falling behind the Medium-Term Plan’s targets. What do we do? Will we adjust pro-poor targets downwards or can we still do something about the BIR and BOC performances? Clearly we have to look at the role of evasion and administration. Tax evasion is much more than a moral issue; it is an issue of systems. For instance, with the kind of rewards and penalty systems that we have for BIR employees, why are we expecting much better performance? In addition, there are really opportunities for evasion. Lastly, the track record of the justice system in punishing people who evaded taxes is to be considered. So where does the solution begin? This is an extremely important issue. Perhaps in Dr. Manasan’s presentation, we can see some possible solutions. 1) “Improving Tax
Administration: A New View from the Theory of Tax Evasion in a Corrupt
Regime” by Dr. Rosario G. Manasan,
Senior Research Fellow, Philippine Institute for Development Studies (PIDS) Dr. Manasan presented estimates in terms of tax compliance and tax
evasion from the mid-seventies to recent years. She then presented recent
developments in the literature on tax evasion – the expansion of the
basic model of tax evasion to that of the second-generation models wherein
the respective motives of the taxpayers and tax collector and the
interplay between the two are considered endogenous. Finally, she looked at actual developments and results in the
Philippines, vis-à-vis the possible implications of certain provisions
and assumptions in the
second-generation models of tax evasion. Below are some details. Recent
Efforts in Tax Restructuring
In
1986, we had a series of tax reforms. We had a comprehensive tax reform
package that focused more on the efficiency goals rather than on the
revenue goals. Since 1987, more tax changes have been introduced to
respond to the need to raise more revenues in the context of a series of
fiscal adjustment programs.
Tax
measures installed in 1986 were meant to simplify the tax structure and
broaden the tax base. With
expanded coverage and stricter implementation of tax withholding schemes,
it should have improved tax compliance.
Significant
gains in tax effort have indeed been achieved.
The tax effort to GNP ratio rose from an average of 11% in
1975-1985 to a peak of 16% in 1997. However, with the Asian crisis, it
dropped to 14.8%. The deterioration has continued thereafter despite the
economy posting recovery in 1999. Tax
Evasion
Updated
estimates of tax evasion confirm improved tax compliance between 1985 and
1995.
However,
as noted earlier, from 1995 onwards, deterioration in tax evasion has
become pronounced both in terms of individual income tax and value added
tax. In short, while there were improvements in tax compliance and thereby
less tax evasion from 1985-1995, the years that followed were
characterized by deterioration in the tax effort and an increase in tax
evasion. These developments suggest that one cannot rely on modifications
in the tax structure alone to correct shortcomings in tax administration.
Certainly,
mechanical approach to tax administration based entirely on logistics and
which excludes the taxpayer and tax collector motivations may not be
useful. Theoretical
Literature on Tax Evasion Basic
Model
In
the basic model of tax evasion, what you have is an
income-maximizing taxpayer deciding how much of his income to declare for
tax purposes given some constant probability that evasion is detected and
penalized. How the taxpayer decides to act is, then, largely dependent on
the relative magnitude of the temptation and the punishment attendant to
tax evasion.
Implications
of the model: a.
Tax compliance improves as the penalty rate and the probability of
being caught increases. b.
Penalty rate is typically fixed in the tax laws. c.
The probability of detection is determined by the ability of
government to enforce tax laws.
In
the Philippines, we saw the introduction of stiffer penalties, creation of
special tax courts in the early 1990s and more vigorous efforts to
prosecute tax offenders during the Aquino and Ramos administrations. These
are steps in the right direction.
None
of the cases handled by the BIR between 1980-1984 resulted in a conviction
up to 1986. In contrast, the number in effectiveness of tax fraud
investigations increased from 1986 onwards with the prosecution of tax
fraud cases and the jailing of some individuals. However, from 1998, there
were no convictions despite numerous cases filed.
The
impact of increasing the tax rate on compliance is not straightforward. An
increase in the tax rate increases the temptation to cheat. At the same
time though, it increases the penalty for cheating especially if the fines
are imposed on an ad valorem basis.
If
the taxpayer’s aversion to the risk of being caught does not increase
with his income level, punishment effect will dominate and an increase in
the tax rate leads to better compliance. This is in contrast to the
conventional wisdom that follows from the so-called Laffer curve.
The
Philippines has reduced the tax marginal rate from 50% to 35-40% then down
to 32%. In terms of the corporate rate, this was reduced from 35% to 32%.
Yet, despite this, we don’t see much improvement in terms of collection
efficiency.
Imposing
maximum penalties for small amounts of evasion may induce taxpayers to
evade by large amounts. Penalties as well as the effort to prosecute must
be graduated with the scale of evasion. Second
Generation Models
In
the second-generation models, the probability of detection becomes
endogenous in the system because of the presence of corruptible officials.
Tax evasion thus depends not only on the incentives to the taxpayer but
also on the incentives to the tax collector and the interaction between
the two players.
One
aspect considered in the second-generation models is the probability of
evasion being detected is a function of the cost to the taxpayer of
under-reporting income depending on the penalty rate and the amount of
effort put in tax enforcement.
When
a tax collector discovers the evasion, he decides on whether to report it
or not. If he under-reports evasion successfully, he gets a bribe in
addition to his salary. If he gets caught, he is punished in the form of a
fine, transfer to another assignment or at worst, dismissal from the
office.
The
probability that the errant tax collector gets caught depends on the
quality of supervision over tax examiners and the third-party audit. If a
tax collector decides to report the delinquent taxpayer, he may be allowed
to retain, as a reward or bonus, some portion of the additional revenue
that is generated because of his vigilance.
The
tax collector’s decision is thus dependent on the relative strength of
the carrot and the stick, namely, 1) the amount of the bribe, 2) the
amount of the reward or bonus, 3) the amount of the fine or penalty, and
4) the probability of being caught.
If
a taxpayer decides to evade taxes, corruption will occur if the collective
benefit of the taxpayer and the tax collector is positive. In a corrupt
regime, the taxpayer will select the amount of tax to evade while the tax
collector will select the effort devoted to monitoring and thus, the
probability of detecting evasion.
For
a given class of taxpayers, audit frequency and declared income ought to
be negatively related. A tax administration system that unduly focuses on
high income tax returns encourages taxpayers to evade by large amounts.
Conversely, a system may deter taxpayers from evading “too much” by
threatening to audit low income tax returns just as intensively as high
income returns.
The
Philippines has offered 10 tax amnesties between 1972 and 1981 and still
another one in 1986. In addition, “last priority” audit was given in
1982, 1987, 1993 and 1999. It was supposedly a last priority audit but in
fact it was immunity from audit and investigations.
Some
model findings relative to tax amnesties indicate that if tax enforcement
is weak, the amnesty rate may have to at a very low rate in order to
induce a significant response from delinquent taxpayers. In such a situation, the expected cost of not disclosing
hidden income is small because the probability of said evasion being
detected is also small. Thus, expected compliance gain from amnesty is
minimal. Also, if the tax amnesty is anticipated, there will be an
increase in tax evasion prior to the amnesty. If tax enforcement is poor,
anticipated amnesties cause aggregate revenues to fall. A point of fact is
that the tax take from numerous tax amnesties offered by the Philippines
government has progressively gotten smaller.
The
second-generation models also look at the possibility of having
compensation incentives for tax collectors in terms of salaries that are
higher than the market rate and/or performance-based bonuses scheme. Can
this help prevent tax evasion? The finding is that where the
government’s ability to punish corruption is weak, it may be better not
to pay tax collectors higher than market salaries because higher
compensation may simply increase the bribe level. Another drawback of
performance-based incentives is the possibility that they may induce the
harassment of taxpayers. The higher the bonus rate and cost of taxpayer
appeal is and the lower the rate of success of taxpayer appeal is, the
greater the likelihood of taxpayer harassment. Conclusion
Piecemeal
incentive reform may increase corruption. Because reforms in incentive
systems and organizational procedures are complementary, reforms are more
likely to succeed if they are comprehensive rather than incremental.
List
of organizational reforms: 1.
Improving personnel quality – Increase flexibility with regards
to hiring and firing of personnel. More stringent criteria for screening
new recruits and for promoting existing personnel will enhance the quality
of tax administration. Clear procedures for dealing with reports of
corrupt behavior and even-handed application of sanctions for the same
should be strictly observed at all times. 2.
Institutionalizing independent third-party audits – Examining the
tax examiners or auditing the tax collectors. 3.
Limiting discretionary authority of tax examiners – Centralized
audit system could reduce the scope for discretionary authority of tax
examiners. For this you need a centralized information base and strict
procedure for audit selection. 4.
Introducing functional specialization in tax administration –
Organizing the tax administration agency along functional line will
minimize collusion of taxpayers and tax collectors. 5.
Increasing computerization – Makes possible the matching of tax
returns with third-party information. It enhances the ability of tax
collection agency to automate the generation of overdue notices which
enhances voluntary compliance. It also facilitates automated audit
selection, thus reducing the discretion of tax examiners.
6.
Improving the appeal mechanism -- Appeals process can be enhanced
by improving the procedures for the filing of complaints, setting of time
limits for judgments, and establishing appropriate burden of proof. 7.
Reducing high-level political interference -- Efforts to eliminate
bureaucratic corruption are quixotic at best in systems with strongly
entrenched political corruption. If tax evaders have recourse to political
influence such that tax audit exposes the tax examiner to disciplinary
action from higher-ups, then no reform of the incentive structure will
work. Epilogue
Widespread
consensus - simplification of the tax structure (in the direction of fewer
exemptions, less differentiation of tax rates across sectors and
activities) will ease tax administration. Many of the tax laws enacted in
the Philippines in the last 15 years were justified on this ground.
The
tax structure has been the subject of unnecessary fine tuning, at best,
and excessive flip-flopping, at worst (moving from schedular to global,
then global to schedular with respect to income tax and from specific to ad
valorem, then ad valorem to specific with respect to excise
tax.)
Frequent
changes in the tax code simply and necessarily impose an additional burden
on tax administration without addressing the complexities that naturally
accompany economic development. 2)
“Comments on Dr. Rosario G. Manasan’s Paper on Improving Tax
Administration” by Hon. Gil Beltran,
Assistant Secretary, Department of Finance (DOF)
The
problem of flip-flopping efforts at tax restructuring is obvious. Former
USEC Milwida Guevara is a witness to how tax reforms are butchered beyond
recognition. For instance, in the last CTRP, we wanted fewer income tax
brackets, a presumptive income tax for hard-to-tax groups, broad VAT
coverage with few exemptions, an MCIT based on net assets, a fiscal
incentive system devoid of tax holidays, etc., but we were unable to get
these.
The
evasion estimates of Dr. Manasan for domestic VAT are very close to
DOF’s estimates although the DOF has a higher individual income tax
estimate because they have imputed the informal economy in their income
figure.
Agrees
that audit rates are very low; at 1.5%, the Philippines has a poor record.
Other countries, despite their lower evasion rates, generate revenues from
audit in the range of 4%-5% of total revenues.
The
audit program of the BIR is still at its infant stage. It still has to
develop industry benchmarks and even has not developed its data sources.
Its planning office has not been aggressive enough in making use of survey
data from NSO, NIA from NSCB, data from SEC or PSE or industry
associations or supervising agencies, or information provided by bidders
in government projects to run after dishonest taxpayers.
BIR
has a poor record on tax cases. Its commissioner is looking into the
possibility of hiring private lawyers to handle BIR tax cases but this
might go against the confidentiality provision in the National Internal
Revenue Code.
We
have had too many tax amnesties such that taxpayers now look forward to
them as periodic events. While tax amnesties generate revenues in the
short run, compliance in the long run is compromised.
Under
Sec. 285 of the NIRC, BIR enjoys an incentive not given to other
government offices. Five percent of the excess collection over the target
of BIR goes to a fund from which bonuses could be allocated for their
personnel (BIR received bonuses from this collection only thrice for the
last ten years).
A
centralized audit selection system is already being developed in the BIR.
An independent third-party audit is being proposed but this proposal is
being objected to by the BIR because it says that under the NIRC, all data
in tax returns are confidential.
The
BIR is already organized along functional lines. Separate units handle
assessment, collection, audit, administrative appeals, and litigation and
prosecution.
Full
computerization is already in the BIR’s medium-term program. Along with
it, the third party information program is being expanded. Despite the
good computerized database of some revenue districts, such information has
not been analyzed with the objective of increasing collections.
Penalties
are imposed whenever bureaucratic corruption is discovered. Every quarter,
the BIR reports about 20 employees being removed or suspended for
violations.
Most
important reform not mentioned in the study is a change in attitude of its
officials and employees. They hold intra-agency seminars and occasions
emphasizing on their crucial role in nation-building to motivate their
staff members to work harder. Some other things that may affect collection are perception of the government and the quality of taxpayers’ services being given by the institution. OPEN
FORUM
|
||||||||||||||||||||||||
|
||||||||||||||||||||||||
