Fast Cars and the Magical City of Tonawanda
Getting a speeding ticket in the State of New York can be a upsetting — and expensive — experience.
Drivers convicted of speeding often face penalties and fines, and repeated or excessive offenses can result in the loss of a license. But in some places in New York State, drivers issued a speeding ticket may see a very different outcome than one would typically expect: a smaller fine, the avoidance of long-term penalties, and — oddly enough — additional money in local government coffers.
Last year, I wrote a post detailing some data science techniques that used a data set from the State of New York on traffic violations. This is an incredibly rich data set and it’s awesome that the state makes this available as open data. However, the issuance of traffic tickets only tells part of a larger story — since the release of this initial data set, the state has now released data on traffic ticket convictions. Comparing both of these data sets allows us to see the full picture of what happens when a traffic ticket is issued in the State of New York, from issuance all the way through to adjudication.
An examination of these data sets show that — in certain areas of the state — tickets issued for speeding are much more likely to be negotiated down to a lessor offense that allows drivers to avoid a penalty from the Department of Motor Vehicles (DMV), and actually provides a backdoor benefit for local governments.
A Lesser Offense
There is a longstanding practice in New York State that allows local courts where traffic violations are adjudicated to accept a plea bargain for a lessor offense from some drivers that are issued tickets for speeding. There doesn’t seem to be any hard and fast set of rules governing this behavior, but what I have found through research suggests that at the judge or prosecutor’s discretion motorists that don’t have prior speeding violations and/or weren’t speeding excessively may be allowed to plead down to a non-moving violation. These lessor offenses typically aren’t reported to DMV so they don’t result in points against a driver’s license. The fines owed for these non-moving violations may be lower than the fines for speeding, and because of the nature of the offense all of the fine revenue goes to the local jurisdiction, not the state.
This would seem to create a perverse incentive for local jurisdictions to allow drivers to plead down speeding tickets, and some proposals have been made to alter the practice. Funding municipal operations through fines and penalties has led to serious issues in other parts of the country, and we should be concerned that this practice may result in an uneven (and unfair) application of the law. Whether or not a motorist should be allowed to plead down to a lesser offense should be based on factors other than whether the local jurisdiction wants or needs the revenue to fund their operations.
Now — thanks to the State of New York’s open data efforts — we have data that can be used to determine how widespread this practice is, and make some recommendations for what policy makers should do going forward.
Looking at the Data
Once the data from the state’s portal is loaded into a local RDBMS (more on this in the technical notes below), we can run queries against it. For the purposes of this analysis, we wanted to determine the conviction rate for speeding tickets for each local court in the convictions data set. To do this, we need to count up the number of tickets issued for speeding violations that get assigned to each local court and compare it the number of convictions for speeding offenses for that court. In other words:
speeding tickets issues / speeding convictions = conviction rate
In running queries against the data, it probably makes sense to limit our analysis to jurisdictions where a non-trivial number of speeding tickets are issued. If pleading down tickets to lesser offenses creates an unwanted incentive for local officials, this is probably more likely to occur in places the local revenue generated from this practice is more significant.
So, if we restrict our analysis to jurisdictions where 10,000 or more speeding tickets were issued over the four year period covered by the state’s data, we get the following:
| Court | Tickets issued | Convictions | Rate |
| TONAWANDA CITY COURT | 11393 | 851 | 0.0747 |
| HAMBURG TOWN COURT | 12358 | 1128 | 0.0913 |
| TROY CITY COURT | 10641 | 1206 | 0.1133 |
| SOUTHAMPTON TOWN COURT | 20766 | 3940 | 0.1897 |
| AMHERST TOWN COURT | 23009 | 4391 | 0.1908 |
| SUFFOLK DISTRICT COURT | 10644 | 2153 | 0.2023 |
| CLARKSTOWN TOWN COURT | 14125 | 2878 | 0.2038 |
| YONKERS CITY COURT | 14030 | 2929 | 0.2088 |
| GUILDERLAND TOWN COURT | 12734 | 2881 | 0.2262 |
| SALINA TOWN COURT | 11102 | 2732 | 0.2461 |
| CHEEKTOWAGA TOWN COURT | 18785 | 4977 | 0.2649 |
| TONAWANDA TOWN COURT | 28097 | 7629 | 0.2715 |
| SYRACUSE CITY COURT | 14226 | 3927 | 0.2760 |
| NEWBURGH TOWN COURT | 14057 | 4103 | 0.2919 |
| COLONIE TOWN COURT | 20001 | 6858 | 0.3429 |
| HARRISON TOWN COURT | 20516 | 8017 | 0.3908 |
| POUGHKEEPSIE TOWN COURT | 11850 | 4752 | 0.4010 |
| ULSTER TOWN COURT | 16768 | 6731 | 0.4014 |
| YORKTOWN TOWN COURT | 11185 | 4682 | 0.4186 |
| GREENBURGH TOWN COURT | 13459 | 5924 | 0.4402 |
| WALLKILL TOWN COURT | 16720 | 8051 | 0.4815 |
| BEDFORD TOWN COURT | 14736 | 7629 | 0.5177 |
| CORTLANDVILLE TOWN COURT | 15333 | 8165 | 0.5325 |
| CHENANGO TOWN COURT | 12901 | 7853 | 0.6087 |
| BROOKLYN SOUTH TVB | 54186 | 38590 | 0.7122 |
| SUFFOLK TVB | 38345 | 29971 | 0.7816 |
| BUFFALO TVB | 16866 | 13329 | 0.7903 |
| ROCHESTER TVB | 24929 | 19952 | 0.8004 |
| MANHATTAN NORTH TVB | 31924 | 26618 | 0.8338 |
| QUEENS SOUTH TVB | 19444 | 16354 | 0.8411 |
| QUEENS NORTH TVB | 94626 | 79970 | 0.8451 |
| RICHMOND TVB | 61138 | 53742 | 0.8790 |
| BRONX TVB | 59097 | 52902 | 0.8952 |
At the bottom of the list — places that have the lowest conviction rates for speeding tickets issued — we see the City of Tonawanda (not to be confused with the Town of Tonawanda, which is also on the list) located in Western New York near Buffalo. The rate of conviction on speeding tickets in the City of Tonawanda is about 7.5%. The Town of Hamburg and the Town of Amherst — which also have very low conviction rates for speeding tickets — are also in Western NY.
Overall, the list includes areas with major highways and significant traffic levels — Clarkstown, Yonkers, Suffolk and Southampton are all downstate and Long Island, and Troy and Guilderland are in the Capital Region — which are places where we would expect to see lots of traffic tickets issued. If we look at how the practice of pleading down traffic violations plays out in terms of local revenue, we can see that it may provide significant support for some jurisdictions. In the City of Tanowanda, fine revenue represents about $475,000 annually — about 6.5% of total city revenue (excluding state aid). Not the largest revenue source by any stretch, but certainly enough to make a difference.
At the top of the list — places that have the highest rates of convictions for speeding tickets are the various Traffic Violation Bureaus. These are run by the state DMV, so there is no incentive to plead down speeding tickets to lesser offenses and generate local revenue. In fact, I don’t think plea bargains are allowed at these facilities. This means that for some areas — like Western NY — drivers may see the final adjudication of their speeding ticket handled very differently depending on where it is issued (or, more accurately, depending on the court it is assigned to).
Around the City of Buffalo there are a number of jurisdictions — Tonawanda, Amherst, Hamburg, Cheektowanga — that have very low conviction rates for speeding tickets. But within the City of Buffalo (where speeding tickets are adjudicated by the Traffic Violation Bureau) the conviction rates are much higher.
Where Do We Go From Here?
The current situation, where drivers issued speeding tickets are allowed to plead down to lesser offenses and generate local revenue, seems far from ideal for a number of reasons.
The practice of supporting municipal operations through penalty and fine revenue can create perverse incentives and undermine people’s confidence in the impartiality of government officials. Because there are no hard and fast rules governing how different jurisdictions allow drivers to plead to lesser offenses, there is no way to make sure that this opportunity is being evenly applied. Do men and women plead at the same rate? Do minorities plead at the same rate as whites? The data currently available don’t allow us to evaluate these questions, but the quasi-formal nature of the practice should raise concerns.
Moreover, the incentive created by this practice should be particularly concerning from a traffic safety perspective. The penalties and fines established for speeding are put in place for a reason — they work, by reducing the incidence of fatal accidents and loss of life. Permitting speeders a chance to plea to a lesser offense allows potentially dangerous drivers to remain on the road with little consequence.
Unfortunately, not all state policy makers see it this way. In Buffalo — where this analysis shows the imbalance between conviction rates is stark — state legislators have in recent years moved to allow speeders in the city to plead to a lesser offense, the same opportunity afforded ticketed drivers in surrounding towns and villages.
This is a move in the wrong direction.
Local governments — particularly in Upstate New York — are strapped financially, burdened by crumbling infrastructure and face a host of challenges in funding their operations. But we should not allow the practice of pleading down speeding tickets to be used as a way to provide a financial crutch to New York’s ailing local governments. Policy makers should eliminate this practice altogether.
Supporting local governments shouldn’t require making our highways and roadways less safe.
Technical Notes: Here’s How I Obtained and Used the Data
The logic for calculating conviction rates is contained in a couple of stored procedures — one to determine the conviction rate for speeding tickets for a given year, and the other to calculate it over the four year period that both data sets cover (this is the one used to generate the table shown above). This is obviously not to most elegant or efficient SQL, but for the purposes of this analysis it will do the job.
All of the data used in this analysis is available from the State of New York’s open data portal and cover the period from 2011–2014. Specifically, I make use of the traffic ticket data set here, and the traffic ticket conviction data set here. Since I want to combine data from both sets, I opted to download both and load them into a local RDBMS so I could run SQL queries on them.
The traffic ticket data set is rather large — over 14 million rows. It contains one record for each ticket issued along with descriptive information about the violation, when it was issued and the court it was assigned to for adjudication. For the purposes of this analysis, we are interested in tickets issued for speeding, so we confine our queries to a subset of all tickets with the set of descriptions listed here. This list was generated by selecting violation descriptions with the word SPEED in them — a little simplistic but not a bad way to start. A few of the ticket categories seem out of place — like UNLAWFUL SPEEDOMETER and VEH WITH PERFORMANCE SPEED LESS THAN 55 MPH REG-9 — so for now, we’ll drop them from our analysis.
The conviction data set contains aggregate counts of convictions for specific violation types (in this analysis, I focused on this set). Since these are aggregate counts, we can’t tie specific tickets to an outcome in a local court (at least not with these data sets). However, we can look at each court and sum up the types of convictions that occur for a given time period. Since the ticket data set includes information on the court each ticket was assigned to, we can create a view of how many tickets issued for speeding in the State of New York were plead down to lesser offenses and in which jurisdictions the practice is more prevalent.
I should note that the conviction data set needed a bit of tweaking before I could conduct this analysis. The format for the Court_Name field was slightly different from the ticket data set, and this was the field I wanted to use to match ticket counts to conviction counts. Using OpenRefine, I was able to modify this data set and add a new column named Standard_Court_Name that used the same naming convention as the ticket data set. As such, the table structure for this data set in my local database differs slightly from the structure of the data available on the NYS data portal.