The Other Other Side of Question 3

With just over two weeks remaining before election day, the supporters and opponents of the various candidates and ballot questions are scrambling to make their cases and persuade their undecided neighbors.

Particularly during economically difficult times, the ability to persuasively claim  that your project or proposed policy change will have a positive impact on the economy can be very compelling for obvious reasons.

Yesterday,  Suffolk University's Beacon Hill Institute (BHI) weighed in on the potential economic impact of Ballot Question 3 -  A proposal to reduce the Commonwealth's sales tax from 6.25 percent to 3 percent.

Their press release was entitled "The Other Side of Question 3: More Jobs and Less Unemployment" suggesting that their analysis of the impact found that the passage of Question 3 would have a net positive effect on the economy.

A closer look at their release and their brief report suggests that this headline is more than a little misleading....

Based on their brief description, what BHI did is to use a model to develop a "dynamic" estimate of the economic effects of what would occur in the event Question 3 is approved.   In their report they claim that:

•  27,199 private sector jobs would be created and 9,885 public sector jobs would be lost (net positive job impact of 17,314.

• $2.05 billion in state tax revenue would be lost.  $33.4 million in local tax revenue would be gained (as a result of the positive effects of the additional wages produced by the new jobs).

When evaluating these kinds of claims, I typically ask myself what would really happen if the policy change in question were in effect....in this case if the state suddenly found itself down another $2 billion in tax revenue.

While BHI briefly reflects on how the state might make ends meet with $2 billion less in revenue, they are far from specific and not very compelling.  In their concluding paragraph they suggest that job losses could be minimized with wage cuts for state employees, increased health care and pension contributions and service privatization.

Left unanswered is the question of whether sufficient savings could be plausibly achieved in this manner.  Additionally, even if such changes were implemented, they would directly reduce the income of state workers thereby reducing their level of economic activity and yielding, at best, a neutral economic result.   

The Mass Taxpayers Foundation (MTF) recently released its take on this issue.  While the size of the revenue impact of Question 3 is in dispute -- MTF estimates a $2.5 billion impact, BHI's estimate of the impact is closer to $2 billion,  MTF's analysis of the impact of Question 3 on the state's discretionary spending is not.   It is based on a review of the way the state actually spends its money, is methodologically transparent and, unlike the BHI analysis that uses a proprietary model, can be double-checked or replicated by anyone who is inclined to do so.

As MTF rightly points out, fully half of the state budget is legally mandatory (MassHealth, Chapter 70, debt service, pension costs, unemployment benefits and other federal requirements).  That means that the other half of the budget would bear the entire brunt of the revenue cuts that would result from the passage of Question 3.

Even if we accept BHI's $2 billion estimate, combined with a widely recognized $2 billion structural deficit the state already faces, discretionary state spending would need to be reduced by $4 billion (or approximately 25%) to arrive at a balanced budget in the event Question 3 becomes law.

This begs the question of what areas of the budget are discretionary and how much money would need to be cut from each in the event Question 3 passes?

According to MTF, the approximately $16 billion in state discretionary expenditures and cuts would be distributed over a wide variety of state programs and services including:

• Local Aid to Cities and Towns ($400 million+ in additional cuts)*
• Public Higher Education ($200+ million in additional cuts)*
• Public Safety ($600+ million in additional cuts)* 
• Transportation ($50+ million in additional cuts)*
• Economic Development ($75+ million in additional cuts)*
• Human Services ($1+ billion in additional cuts)*
• Capital Spending (reduced, potentially higher debt service cuts if bond rating is lowered) 

* Assumes about an across the board 25% cut in discretionary FY11 expenditures.  Reductions in the size of the cuts in one area would necessarily require increasing the size of cuts in another.

It is easy to imagine several additional "dynamic effects" of these cuts not considered by BHI's model, most notably the impacts of the cost shifting from state government to households and local governments that would likely occur in the event these cuts were made.

The impacts on household budgets of likely increases in property taxes in some communities, school fee hikes and higher user fees for courts and other public services are not systematically considered by the BHI model.

Further, it stands to reason that these cuts would disproportionately impact families with special needs (those with a developmentally disabled,  mentally ill or elderly family member), those who are out of work and those attending a public institution of  higher education.  These households would either need to pay out of pocket for these services (thus reducing their disposable income available to spend on other things and reducing economic activity) or do without these services.

Additional impacts such as the need for some family members to leave the workforce to care for a elder or disabled family member no longer covered by state services would also have negative economic effects.  This is to say nothing of the profound human costs that would result from some of these cuts in programs and services.

So, what can we learn from the BHI report?  Two conclusions seems warranted and both are non-controversial in my view.

1) The proposed sales tax cut would have some positive effects on the state economy.
2) It will also dramatically reduce tax revenue.

Significantly, what we don't learn is whether on balance the implementation of Question 3 would have a net positive impact on the state economy.  Notwithstanding the misleading title of the report suggesting otherwise, there are clearly numerous reasons to believe the net impact of Question 3 may be negative in purely economic terms.

While it is not always possible for a given analyst to definitively answer all sides of a given question, it is standard practice for analysts (as opposed to advocates) to discuss the limitations of their work and transparently acknowledge the implications of issues they could not consider or systematically address.

It is unfortunate that BHI made no effort to account for any or even acknowledge any of these additional potential impacts in their analysis despite their claims that their "dynamic model" captured the range of Question 3's economic impacts.

The problem here is what a methodologist would term omitted variable bias and what Saint Thomas (and my Mom) would call a sin of omission.  Bottom line, to the extent that BHI failed to consider or even acknowledge additional factors that could have a material impact on the analysis, their analysis is misleading.

As someone who has led a number of these types of projects,  I am acutely aware of how technically difficult it can be to conduct these studies in a manner that carefully considers both the economic benefits and the costs of a given project or proposal.  These sorts of challenges should make all policy analysts humble and policymakers and the public skeptical about all claims about the costs and benefits of major changes in public policy.  

Regardless of how one feels about the role of government in society, I think we can all agree that the price of getting our decision on Question 3 wrong will be very high.  What we need in this situation is an honest debate about how to cope with the profound fiscal challenges we are facing while promoting a high quality of life and standard of living across the state.  Viewed through this lens, yesterday's Beacon Hill Institute release leaves a lot to be desired.

It is completely legitimate and healthy for us to have a vigorous public debate about the role that the state and the nation should play in providing a social safety net for its people.  This is a debate that is currently occurring across the nation and it is a very important one. 

If policy analysis is to inform rather than distort these critical public discussions, analysts must clearly distinguish between what is (facts) and what should be (values).  In the matter of Question 3 the facts are clear.   What remains to be decided on November 2nd is a question of values.

Let's hope we get it right...