“It is an unpardonable sin to misuse numbers for propaganda purposes. The deliberate misuse of statistics to ‘prove’ a point for political advantage often leads to policies which cause serious harm to peoples, organizations, and countries.” 1
“David Sweeney, Mayor Martin J. Walsh’s top numbers-cruncher, told Boston Herald Radio’s “Morning Meeting” that even though the school budget is rising at a rate three times greater than any other city service, “we have no plans to close schools at this time.” 2
 |
| Data from the City of Boston |
We received an email last week from a friend of our blog with an excellent analysis of the recent Medium piece published by Boston’s Chief Financial Officer David Sweeney. We have an interest in the ethical implications of selecting and comparing data sets, so as the Boston City Council is poised to vote on the revised budget, we thought it was important to share this rebuttal of Chief Sweeney’s analysis of the BPS budget in the context of city spending.
1 Spirer, Herbert F., Louise Spirer, A.J. Jaffe.. Misused Statistics. 2nd Ed. Marcel Decker, Inc, NY, NY. 1998. P. 4.
2 Sweet, Laurel J. “Boston Budget Chief: ‘We have no plans to close schools’.” Boston Herald. Thursday July, 16, 2016.
************************************************
David Sweeney’s recent article, Boston Proudly Invests in its Schools, contains many interesting facts and figures about the Boston Public Schools budget. However, a few of the data visualizations make convenient use of three specific mechanisms to exaggerate the growth of the BPS budget in comparison to other city department budgets:
- Using absolute budget growth rather than percentage growth
Many budget increases - in particular salary increases - can reasonably be expected to grow in proportion to the size of the budget. Using absolute dollar increases only serves to make larger departments’ budget increases look bigger. - Selectively using FY16 appropriations vs. FY16 estimated expenditures
When calculating budget growth between FY11 and FY16, Mr. Sweeney uses the Public Safety Cabinet’s FY16 appropriation rather than its FY16 estimated expenditures. However, when comparing FY16 and FY17, Mr. Sweeney does use the FY16 estimated expenditures. The result is that neither chart truly accounts for an increase of more than $20 million in the Public Safety Cabinet’s spending that occurred during FY16. - Selectively including health insurance costs in the BPS budget
When comparing total budget increases from FY11-FY16, Mr. Sweeney uses figures that do not include health insurance costs. When looking at the change in expenditures between FY16 and FY17, however, he includes health insurance costs in the BPS budget but not the other three departments in his chart.
Mr. Sweeney’s article includes the following paragraph and chart:
Looking back over the past five years, investment in BPS has increased much faster than other City departments, growing by 25%, while public safety departments grew by 20%, and all other departments grew by less than 13%. Because Boston has been committed to investing in BPS even as the City’s Charter School costs have climbed sharply, Boston’s total education budget grew by almost 33% over that same time. Maintaining this level of commitment to BPS, with fixed costs rising as well, has not been easy. The City has had to reduce positions in other departments.
 |
| Source: Dave Sweeney, Boston Proudly Invest in its Schools |
First, it is somewhat misleading to follow a paragraph that describes the percentage growth of these departments with a chart that uses absolute growth. Because the school department’s budget is the largest in the city, factors like a modest salary increase negotiated years ago are exaggerated by using absolute dollar figures rather than percentages.
Second, the chart shows that Public Safety spending is staying roughly level from FY15 to FY16, and in fact the Public Safety Cabinet’s original FY16 appropriation of $546,973,900 was a slight decrease from the department’s FY15 expenditures. However, the FY17 budget shows that the estimated Public Safety expenditures in FY16 are actually $567,576,563. It is simply inexcusable for Mr. Sweeney to ignore this increased expenditure of more than $20 million. To put it another way, the Public Safety Cabinet, a department that is significantly smaller than the school department, received a spending increase over the course of FY16 that is larger than the supposedly generous $18.2 million increase proposed for BPS in FY17, and yet this article completely glosses over it.
If we change the chart to use percentage growth since FY11 instead of absolute growth, and add in FY16 estimated expenditures along with the original FY16 appropriation, we see a different story: growth in the BPS budget mirrors growth in public safety spending almost exactly. In fact, with this new spending level, investments in Public Safety spending have actually grown just over 25% since FY11, very slightly faster than investments in BPS have grown.
 |
| Data from the City of Boston: https://www.cityofboston.gov/budget/ |
Now, let’s add in the FY17 appropriations, too, including the $4.7M that was just added to the BPS budget.
 |
| Data from the City of Boston: https://www.cityofboston.gov/budget/ |
As you can see, in FY17 growth in Public Safety expenditures since FY11 will slightly, but clearly, surpass the growth of the BPS budget, even when including the recently announced $4.7M increase to the BPS budget.
Now let’s look at another chart that Mr. Sweeney includes in his article:
 |
| Source: Dave Sweeney, Boston Proudly Invest in its Schools |
You might think that the city’s Chief Financial Officer would be careful to make consistent use of data in his visualizations, but if so, you would be wrong. First, the starting point for this chart is not the same as the ending point for the previous chart; here Mr. Sweeney has conveniently chosen to compare FY17 appropriations to FY16 estimated expenditures. In other words, after looking at the budget up through FY16 in his first chart, he has added more than $20 million to the Police and Fire department FY16 budgets before comparing those budgets to the FY17 appropriations. Since we have already corrected for this “oversight” in the earlier charts, there is no further correction necessary in our charts, but the manipulation of the data is so egregiously deceptive that it is worth reiterating. Second, in the previous chart he used figures that were “net of health insurance.” Not so here - these are the total budget increases, and of the four departments listed only BPS pays health insurance costs out of its own budget. However, Mr. Sweeney is consistent in one important way: he continues to use absolute growth rather than percentage growth as a way of making the BPS spending increases look as large as possible in comparison to the other departments’
Let’s go ahead and correct this chart, too, first by removing health insurance costs from the BPS budget increase:
Already the growth in the BPS budget in FY17 looks more modest compared with these other, smaller departments. Now let’s convert to percentage growth:
 |
| Data from the City of Boston: https://www.cityofboston.gov/budget/ |
 |
| Data from the City of Boston: https://www.cityofboston.gov/budget/ |
Perhaps the city would do well to cast the same critical eye on every department that it has reserved for the schools.
