An Economic Perspective of a VAT Increase, Part 2: Effects on Tax Revenues and Welfare

Last week's (Part 1) discussion regarding the effect of the VAT rate increase on prices illustrated that from a basic supply-demand perspective, ceteris paribus, base prices (those collected by the producers) and equilibrium quantities should both fall, following the VAT hike.  It was, moreover, briefly mentioned that one immediate side effect of this is a direct loss in social welfare, commonly referred to as deadweight loss.  Indeed, any increase in the tax rate, in this sense, always embodies a certain degree of inefficiency.  This is why economists constantly argue that tax rates adjustments should be approached with extreme caution and utmost consideration to intended goals and potential effects.

What are the chances that the decision of raising Romania's VAT rate from 19% to 24% within one week was preceded by a comprehensive methodological analysis of its consequences?  For the majority of Romanians, the answer comes in the form of either laughter or tears (sometimes both).  Yet, criticism of this decision is not, by any means, new; the common media has exploited the opportunity to denounce the policy move to maximum extent.  Sometimes, the government officials themselves, either as a paradoxical phenomenon or arrogant sarcasm, publicly proclaim the counterproductive nature of this initiative.  However, most such discussions appear to be narrowly cast in terms of poorly understood concepts: usually limited to effects on consumption and encouragement of tax evasion.  On the other hand, certain obvious outcomes -- such as, for example, the fact that a higher VAT rate implies lower profits which immediately leads to lower income tax revenues -- receive remarkably little attention.

Indeed, many questions and details remain unaddressed.  For those wishing to grasp a better understanding of the economic effects of VAT-type (e.g. consumption) taxation, the present discussion once again offers a fairly simple explanation grounded in fundamentals of economic theory.  What exactly drives the inefficiency of a consumption tax?  How does an increase in the VAT rate translate into overall tax revenues, both from VAT and income based collections?  How may a benevolent policy maker (i.e. one who is concerned only with the well-being of the public) approach the decision of a VAT rate increase?  These are just a minor subset of questions requiring methodological answers in order to attempt to understand the misalignment between the motivation driving the current government's decision and the consequences that may be expected.  Unfortunately, one clear conclusion supported by the ensuing analysis is that bankruptcy has now materialized as a rather genuine possibility for Romania in the near future.


Possible Benefits of the VAT

In Part 1 of this discussion, it was shown that the VAT distorts the market equilibrium such that the consumers pay a higher prices and purchase a lower quantity, while the suppliers receive a lower price and sell a lower quantity.  In this sense, each side is forced to give up a portion of their surplus (e.g. welfare).  Part of the surplus taken from consumers and producers forms the tax revenues collected by the government.  However, a portion is lost altogether, creating the deadweight loss of taxation.  Economists argue that the latter is inevitable, and in addition, the higher the tax rate, the greater the deadweight loss.  Nevertheless, almost every known country in the world exercises some form of taxation, and most employ a tax on consumption (e.g. a variant of the VAT).

Hence, taxation must be to some extent beneficial, and in fact, such benefits are easily conceptualized from an economics perspective.  That is, taxes may be regarded in economic terms as simply a form of welfare redistribution.  A typical argument for the use of taxation schemes is that markets are not always sufficiently effective at allocating resources in society such as to meet certain social needs, particularly when certain inhibitors such as asymmetric information, transaction costs, etc. are present.  Some standard examples of these are healthcare, military, and justice.  Another possible reason for taxation is inter-temporal: the deadweight loss generated today may be offset by future benefits arising from government programs that combat an individual's tendency to make decisions based mostly on the present.

As an example, consider a certain government taxing automotive gasoline consumption.  Resulting from this, the price paid at the pump by drivers increases, while gasoline consumption decreases along with the revenues obtained by gasoline producers.  In the short term, both automobile drivers and gasoline producers contribute to the collected tax revenue, but some of the surplus is inevitably lost.  However, suppose the government uses the collected tax for two purposes: (i) improve the efficiency of and subsidize public transportation and (ii) sponsor research of alternative energy sources and auto fuel technologies.

As a result, the surplus lost by consumers in terms of gasoline consumption is potentially offset by new availability of public transport options.  For the gasoline manufacturers, on the other hand, lost revenues in gasoline sales are compensated for by government funding of research and development efforts.  In the long term, pollution is greatly reduced and possibly new, more efficient and perhaps even cheaper alternatives to gasoline (as automotive fuel) are uncovered.  The gains in social welfare induced by the latter may very well offset the initial deadweight loss generated by the tax, in which case, both consumers and producers in the automotive gasoline industry should actually favor the tax-based initiative.


Potential Pitfalls of the VAT

There are two subtle points inherent to the gasoline tax example above that are critical to its successful outcome.  First, in order for the tax revenue funds to used as described, the government must to a large extent be indeed benevolent.  Second, a good deal of rather sophisticated analysis would likely be necessary to determine precisely how much of a tax needs to be imposed.  A failure of either requirement is likely to undermine any favorable outlook of the gasoline tax as it would disrupt the delicate balance between collecting enough tax revenue to finance the necessary public programs, while at the same time, avoiding the imposition of an insurmountable burden on the market, in the initial stage.

The first point is in fact a commonly explored topic in Romania.  In simple terms, if the government is not benevolent, and enough incentive is present, there is no guarantee that the tax revenue will be used for any of the beneficial programs discussed above.  Rather, instead of improving public transportation and encouraging technological research, the funds may simply be allocated for personal use by the politicians entrusted with their administration.  When this happens, the entire tax burden -- that is, all the surplus relinquished by both consumers and producers -- becomes a social detriment.

To be exact, the misappropriated tax revenues do not actually add to the deadweight loss; these funds are not entirely removed from the economic system.  Rather, the problem arises in terms of a grossly inefficient redistribution of welfare.  The surplus extracted from automobile drivers and gasoline producers, in this example, would simply be entirely reallocated to a select few members of society.  Clearly, one may have a hard time convincing the market participants to comply with such a taxation scheme.  In fact, the more consumers and producers believe that politicians are likely to give into the incentives of misusing tax revenues, the more incentive they have themselves of finding ways to evade the taxation.

The second issue, on the other hand, typically receives relatively little attention.  Nevertheless, it is important to recognize that even with a benevolent government, taxation may still not meet its intended objectives.  Specifically, essential to the successful implementation of a taxation scheme are both a precise estimate of the immediate surplus losses, separable into tax revenue contributions and deadweight loss, along with an accurate forecast of potential future benefits that may be expected from the planned social programs financed by the tax.  If either of these fails, policy makers run the risk of either collecting insufficient taxes to support the costs of the intended programs, or generating a deadweight loss that is ultimately insurmountable even by sufficiently funded and efficiently operated initiatives.  In either case, incentives of market participants to comply with such a taxation scheme are once more endangered.

In the context of the latter, let us briefly consider how taxation policy may be approached, while keeping the technical aspect of the analysis on a simplest possible level.  Specifically, suppose that for the purpose of VAT policy, the government primarily considers the optimal rate at which to collect the consumption tax.  Three major categories, in this case, require a thorough analysis:

1. How much tax revenue may be expected at a given tax rate?
2. What are the contributions of consumer surplus derived tax revenues, producer profits derived tax revenues and deadweight loss to the tax burden generated by a given tax rate?
3. How much benefits to producers and consumers, and over what time frame, may be expected from the social programs funded by the collected tax?

Moreover, recall that the answers to the above questions will vary from industry to industry, and have close ties to the elasticities of supply and demand.  Therefore, when considering a VAT rate that is applied equally across many industries, the analysis must also account for variations across industries at an aggregate level.

In Part 1, the comparison of a particular market equilibrium in the presence of a VAT tax to that of a tax-free equilibrium was presented graphically within the basic supply-demand framework.  The plot, reproduced here for convenience, illustrates the entire tax burden partitioned into its three notable components: tax revenues from consumer surplus (yellow), tax revenues from producer profits (orange) and deadweight loss (red).

In practice, each of these areas must be estimated as accurately as possible, for every industry where the tax is to be applied; in general, these areas mostly depend on the tax rate (multiplier) $t$ along with demand and supply elasticities, $\epsilon_d$ and $\epsilon_s$, respectively.  One objective of the policy makers, consequently, should be to ensure that every segment in part is compensated for by the social programs that follow taxation.  Otherwise, the tax scheme is inevitably inefficient, from a social welfare perspective.

Now consider a decision to raise the tax rate $t$.  For the sake of an interesting argument, suppose that the rate depicted in the plot above represents a well thought out (originally) taxation scheme.  Then, what type of reasoning might justify an additional rate increase?  What are the possible economic consequences of such an increase, in this context?  These questions may continue to be explored through the basic supply-demand framework.  The following plot compares two market equilibria, induced by two different VAT rates: the original rate $t_1$ and the increase rate $t_2$ (where the original VAT induced equilibrium replicates the post-tax scenario of the above plot).

Note first the obvious implications: increasing the rate from $t_1$ to $t_2$ yields a continuing rise in final prices (from $t_1p_1$ to $t_2p_2$) along with declining base prices ($p_1$ to $p_2$) and industry outputs ($q_1$ to $q_2$). Observe, however, that the effects on tax revenues are slightly more complex, compared to the scenario of the original tax implementation (relative to a tax-free equilibrium).  Specifically, because equilibrium market quantity decreases, some of the tax revenues under the initial tax scheme are lost.  This is represented by the orange area in the plot above.  On the other hand, the greater difference between base and final prices, yielded by the higher tax rate, produces additional tax revenues, as depicted by the yellow regions.

From this perspective, two important characteristics of the change in tax revenue structure are worth further exploration.  First, it is important to acknowledge the fact that the lost tax revenues (orange area) fall entirely into the additional deadweight loss created by the tax increase (market by the red outline) -- no portion of it is returned to the consumers and producers it was originally elicited from.  On the other hand, the gained tax revenues are extracted solely from additional consumer surplus and producer profits!  The restructuring of the tax revenues itself resulting from the rate increase, therefore, is entirely inefficient.  However, unlike the original introduction of taxation, the implications here are far more profound.

The latter claim is directly connected to the fact that tax revenues do not necessarily increase following a raise in the VAT rate.  So much is clear from the plot above -- if the total yellow region is less than the orange section, then revenues lost to the declining industry output outweighs revenues gained from the inflated price difference, and overall, less tax revenues are collected in the particular market.  Nevertheless, the deadweight loss of taxation increases regardless (i.e. the region outlined in red is guaranteed to exist).  When this happens, the increase in the tax rate induces additional losses in social welfare and a decline in tax revenues, which clearly makes the welfare loss altogether irrecoverable.  In short, the consequences of such an increase may be quite disastrous, just by virtue of the simplest imaginable market forces naturally at play.

Indeed, a decision to increase the tax rate in an already existing taxation scheme warrants even more caution and extensive, methodological consideration than an initial tax implementation.  It is moreover fair to conclude from this analysis that one clearly inappropriate motivation for raising a tax rate is an unexpected drop in tax revenues resulting from an exogenously falling market demand (i.e. due to business cycle fluctuations).  In fact, this is exactly the type of scenario susceptible to the detrimental outcomes discussed above, in addition to reinforcing further downward pressure on the already contracting demand.  An application of the latter to Romania's case is discussed shortly.  First, we conclude the supply-demand base welfare analysis of a VAT increase with a brief look at the implied relationship between tax revenues and the VAT rate.


Effect of the VAT Rate on Tax Revenues

Imagine a hypothetical market, which prior to 1 July 2010 was transacting $q_1=$ 100 units per day at a base price of  $p_1=$ €100 (therefore, at 19% VAT the final price paid by the customers was $\tilde{p}_1=$ €119).  Suppose that supply and demand in this market are such that an increase in the base price from €100 to €101 would lead producers to expand supply from 100 units to 111 units, while the equivalent increase in the final price form €119 to €120.19 would drive down the quantity demanded by consumers to 88 units per day.  Alternatively put, the industry is characterized by supply elasticity $\epsilon_s =$ 11 and demand elasticity $\epsilon_d =$ 12.  Note that the plots above may be representative of such a market.  More importantly, this information alone is sufficient to derive the following conclusion from the basic supply-demand framework: ceteris paribus, the VAT rate increase inevitably lowered the tax revenues obtained from this particular industry.

In fact, any alteration of the VAT rate away from 19% is bound to negatively impact the overall taxes collected, given the specific demand and supply elasticities.  This is because for this industry, 19% is the revenue maximizing rate of taxation, which in the simplest terms means the following.  The total tax collected in a market is the product of three components: the tax rate, the base price and the output, expressed algebraically

\[T = (t-1)pq\]
Therefore, at a 19% VAT rate (e.g. $t =$ 1.19) and equilibrium $p=$ 100, $q=$ 100 the government would collect €1900 in VAT taxes per day from this industry.

All else held constant, this is the maximum revenue attainable in the respective market.  To see why, recall that increasing the VAT rate always decrease the equilibrium base price and quantity.  In the formula above, therefore, as the VAT rate increases one of the factors (e.g. $t-1$) increases accordingly while the remaining two factors (e.g. $p$ and $q$) move in the opposite direction, thereby exposing the overall revenue $T$ to two opposing forces at any particular time.  As it turns out, while $t$ increases up to $t^*=$ 1.19 revenues increase to the extent that the inverse effect on $p$ and $q$ is sufficiently weak in comparison to the growing tax rate $t-1$.  However, continuing to raise the rate in excess of $t^*$ is counterproductive: beyond this point, the downward pressure exerted by the falling base price and market output overwhelms the higher tax rate and yields an overall negative effect on tax revenues.

To visualize the above described effect, consider the plot of the tax revenues generated in the basic supply-demand framework against a given tax rate:

Economists typically refer to this plot as the Laffer Curve (named after the economist who first pointed out the theoretical relationship between tax rates and resulting revenues) and emphasizes a trivial yet often overlooked fact in policy: even under the simplest, hypothetical settings governments can only directly control the rate of taxation, not the actual amount of the collected tax.  The latter, on the other hand, is eventually determined by the chosen rate of taxation and market characteristics, captured in large by the elasticities of demand and supply.

In fact, under the basic supply-demand framework, it can be shown that $t^*$ is determined only by these elasticities through the following relationship:

\[t^* = \frac{1 + \epsilon_s^{-1}}{1 - |\epsilon_d|^{-1}}\]
It is worthwhile to observe that nothing in the above formula, nor the preceding analysis, involves tax evasion.  Rather, natural economic forces of supply and demand are sufficient to support the Laffer effect of taxation.  Of course, just as important is once again the fact that this relationship is specific to each industry.  When a single rate is applied to a variety of heterogeneous markets, the aggregate relationship is more complex.  To that end, one should consider how many and what type of industries in the general economy are comparable to the hypothetical cases.  One interesting question, therefore, becomes how many industries in Romania are comparable in terms of demand and supply elasticities to the hypothetical market examined in this section?


Application to Romania and Conclusions

It is not difficult to imagine the process by which government officials arrived at the decision to increase the VAT rate to 24%.  By an arithmetic calculation, the increase of 5 basis points in the VAT rate infuses RON 3.5-4 billion in additional tax revenues by the end of 2010.  This covers approximately one third of the €2.5 billion gap needed to be filled in order to meet the budget deficit requirements of the World Monetary Fund (WMF) agreement.  The expected additional revenues, in turn, align Romania back on target with the WMF plan, allowing for the release of an additional €913 million in borrowed funds.  Benefits: WMF tranche in the BNR reserve.  Costs: none, really (some people get upset, but it's the crisis, it happens...).

Ok, perhaps it is more complex than this -- there are certainly plenty of theories, some probably at least in part masquerading varying degrees of substance: personal interests are involved, it is all part of the president's grander master scheme targeting power, etc.  Many times public statements are outright paradoxical.  For example, Public Finance Minister Vladescu has been quoted as projecting that "the economic downturn will be with certainty greater after the application of the VAT increase," but if the finance ministry really understands the implications of this, how does it justify the simultaneously expected increase in tax revenues?  At other times, government officials have spoken of effects on consumption, prospective growth in tax evasion, etc. tossed in with a mixture of other popular economic and political catch phrases.  In all, however, the rhetoric seems to only reveal a frivolous misuse of poorly understood, regurgitated technical terms.

The decision was, in short, incompetent.  The consequences may be quite devastating.  However, the waves of criticism currently engulfing Romanian society still lack any sense of legitimate seriousness -- it is still perceived as yet another misguided decision by an incapable, untrustworthy political class that Romania will have to survive, a familiar experience for the common public.  Yet the basic lessons from microeconomics outlined above suggest a far more dire situation.  To see this, one just needs to follow a few simple connections.

Setting aside for the moment the issue of tax evasion, consider once more the hypothetical market of the previous section with $t^*=$ 1.19.  What proportion of industries currently operating in Romania may be characterized by enough elasticity in supply and demand to support the Laffer type of relationship between the VAT rate and tax revenues, with $t^*$ at or below 1.19 (i.e. where the tax revenue maximizing VAT rate is less than or equal to 19%)?  If there are enough to make an impact on the overall amount of VAT tax collected, then clearly no amount of anti-evasion measures, regardless of how successful they are, will deter the inevitable drop in future tax revenues.  Likewise, tax evasion does not preclude the simple fact that the certainty of declining profits resulting from the VAT rate increase (as discussed in Part 1) clearly exerts an immediate negative impact on the revenues collected from income taxes (e.g. cota unica).

Yet, without doubt tax evasion is a well mastered practice in Romania, a claim receiving little opposition amongst the Romanian public or political class.  Indeed, the presence and likely dramatic expansion of VAT related evasion will effectively drive $t^*$ further inward for a great deal of Romanian markets.  More importantly, combating this evasion cannot eliminate the prospect of reductions in tax revenues.  This is because while successful anti-tax evasion measures may reduce the level of evasion, it must be recognized that such instruments are themselves costly.  That is, the financing necessary to support the cost of monitoring, verifying, investigating and enforcing tax collection activities must essentially be derived from the same budget that the VAT tax is targeted to support.  To that end, there is little dispute to the claim that the cost necessary to achieve any noticeable deterrence of tax evasion at this point in Romania is formidable, and hence, significant contraction of tax revenues resulting from an increased incentive to evade taxation is inevitable.

At the same time, microeconomic fundamentals ensure that the VAT increase invariably generates deadweight loss in all industries.  This is true even if one was to take the counter-factual position that the initial taxation scheme (at 19%) was well enough designed and implemented such as to be justifiable from a social welfare perspective.  The conclusion also holds regardless of the elasticity levels in the market demand and supply.  Consequently, Romania's economy now has all the proper ingredients for the disastrous scenario previously described: falling tax revenues together with increasing deadweight losses across all industries, leading to irrecoverable losses in welfare.

One does not need to be a Ph.D. economist to conjecture as to what outlook this forms for investors, local and foreign.  Now take into consideration the environment of contracting demand, an inaccessible banking sector and the regulatory uncertainty highlighted by the erratic, haphazard way in which this latest fiscal measure was introduced; hints are already abound regarding an all out overhaul of the taxation and fiscal policies, targeted for the start of 2011, as well.  Clearly, the pressure on a risk-averse investor (meaning one who does not derive separate pleasure from the pure gambling aspect of the venture) to look towards opportunities elsewhere has just taken an immense leap.

If this in fact materializes, the medium and long term path for Romania's tax revenues is downward bound, both in terms of VAT and income based collections.  The choice to increase the VAT rate by 5 basis points in this shotgun fashion is not the sole catalyst for the potential destructive outcome.  However, following a long series of adverse exogenous influences accompanied by capricious and inept fiscal responses, the latest may have very likely secured Romanian economy's place on precisely this very evolutionary track.  Unfortunately, its final destination is certain bankruptcy.

Of course, we will not actually know if this is in fact where we are headed until we get there -- foreseeing or even identifying such a phenomenon affirmatively requires a fairly serious degree of empirical research (I believe this actually exists on a considerably high level within BNR, but is precluded from being related externally).  We can keep our fingers crossed meanwhile that it is not yet the case.  The problem, it seems however, is that Romania's legislature is fairly determined to get us exactly there.  Unfortunately, for this they do have the competence to succeed.