do traffic cops cause traffic jams?

everyone who has ever been to dhaka on a normal workday knows that we have the world’s least enviable traffic jams. on any given day, it can take up to half an hour to travel just a couple of kilometers. my daily commute, which clocks in at only about 7 km each way, takes me anywhere from 20 minutes on a good day, to almost an hour and a half.

urban planning experts complain about the lack of roads. drivers complain about slow-moving rickshaws. passengers complain about the driving skills of others. but everyone seems to agree that one of the major causes of traffic jams is the capacity and activities of traffic police in dhaka.

earlier this year, i decided to test this for myself, as part of a course on econometrics that i was taking at the time. being the pessimist i am, i naturally beleived that, the more traffic police there were on the streets, the more traffic there would be, and therefore the longer it would take to travel a given distance.

i figured that it would be a sufficiently easy model to test, and that getting data would be relatively easy. however, it’s quite a general and basic model in itself, and so there may be flaws in the analysis. the results are quite interesting and informative, but if you would like to bore yourself with the full gamut of econometric analysis, you can see it all here.

the results indicate that, on average, adding another traffic cop to the streets would reduce the average travel time by about 75 seconds. On the flip side, an additional car on the road would increase travel time by 5 seconds.

before we get in to what this could possibly mean, let’s talk about how i did this research. first, i picked a specific route for this study – asad gate to wireless, mohakhali, and the way back, via bijoy sharani and mohakhali rail crossing. this route was selected for several methodological reasons:

  • rickshaws are not allowed to travel on this route, and so the impact of these slow-moving vehicles on traffic jams was controlled. 
  • there are no schools along this route, which means that congestion on this route is also not due to school traffic.
  • parking is prohibited along most of this route, and so most vehicles are supposed to be in motion on this route.
  • this route is one of the most notorious jam-packed roads in the city, particularly in the evenings.
  • there is always generally a high concentration of traffic police on duty along the route.

i collected data on the time taken to traverse the route, the number of traffic cops along the route, and the average traffic gathered at two points during my trip. the last of these was collected at great risk to life and limb: once my car was stopped at the traffic light, i jumped out and counted the number of cars waiting at the light, and then jumped back in to my car once traffic started moving again. data was collected at both morning and evening peak traffic hours. morning rush hour here is defined as between 7 am and 9:30 am, and evening rush hour ranges from 5 pm to 8 pm. to randomize the data, i travelled at different times every day during these two blocks.

what do the results mean? well, to deal with the simple things first, the second result makes sense. given the limited road capacity, more traffic on the streets would obviously lead to more traffic and therefore increased travel time. 

the traffic cop phenomenon is harder to explain though. after all, most of their activity seems to standing at traffic intersections all day, languidly waving to cars to pass or stop. except, of course, if you have an accident or are being mugged. in which case they are most likely to disappear faster than a rabbit after a gunshot.

to explain why this occurs, it’s necessary to extrapolate from our data and use our logic. i surmise that this is because of the fear factor: if there is a traffic cop at the intersection, you are more likely to follow traffic rules and drive carefully. ergo, the more traffic police, the higher the likelihood of following traffic rules.

but that’s a simplistic piece of logic at best, and given the tendencies of bangladeshi drivers, it’s also rather hard to believe. instead, the simpler explanation may be that, as the morning progresses from 7 am onwards, two things happen simultaneously: more traffic cops come on duty, and traffic slowly decreases from the school rush to the commuting rush to the general traffic trends. similarly, as the evening progresses, more and more traffic cops come on for the evening shift while the commuter rush starts to decline.

but are ingenious police shift timings the only cause for this phenomenon? perhaps, but that is quite pessimistic. although travelling down the dhaka streets and watching the activities of the police force would reinforce one’s assumption that they are incompetent, there is also a need to be more objective. i would postulate that there is an extension of the “fear factor” hypothesis: since the primary task of the police is to reinforce the traffic light system, it’s likely that, if there are traffic cops, drivers are more likely to obey traffic lights, thus ensuring smoother flow of traffic and more effective traffic management. thus, there is less traffic, and simultaneously lower travel times.

this research, however, is by no means comperehensive and conveniently ignores several other important variables. for example, it does not take into account how many policemen were actively guiding traffic, as opposed to standing around chatting or seeking bribes while pretending to fine people (i know, gross generalization). similarly, it assumes that all traffic counted is travelling in the same direction, instead of turning or travelling in other routes.

finally, it ignores the fact that there are other underlying reasons for traffic, other than policemen or the number of cars. for example, one of the main reasons traffic jams occur in dhaka is that important intersections are closed off haphazardly across the city, meaning that cars that want to turn at a certain intersection often have to congregate at a limited number of turning points, thus backing up traffic. after all, traffic moves in flowing patterns, and blocking the flow inappropriately can cause it to back up indefinitely. think of water flowing into a bucket through a pipe with lots of holes. if you start plugging up the holes arbitrarily, the amount of water in the pipe will start to increase, until both the pipe and the bucket are flooded. therefore, ignoring traffic mnagement as a primary cause of the traffic jams is virtually impossible.

using these results for policy recommendations along the lines of increasing the number of traffic cops can be dangerous and will probably not solve traffic problems. however, it is undeniable that more cars will cause more traffic, and so traffic reduction must focus on controlling the number of cars or managing them better, so that they don’t clog up the limited pipes that make up dhaka’s streets.

political procrastination

the roads of dhaka are eerily empty for 8:30 pm. strangely, there are few traffic cops on the road – just a lone soldier dutifully manning the road 27 intersection. no sergeants astride their trusty motorcycles, no battalion of police positioned idly at major traffic points. few cars are on the streets – busses abound, but otherwise there are few other vehicles. thin queues of people seem stranded at bus stops. storekeepers rush to shut their shutters, half an hour before standard closing times. no noise, no incessant honking of horns, no road rage-infested drivers trying to push others off the roads. a lone beep from my car, at an errant rickshaw, seems to ricochet off the concrete walls.

my mobile flickers softly and vibrates twice. new message received. opening the inbox to read the missive seems to take forever, with the status bar seeming to sway lazily from one side of the screen to the other. finally it opens. “govt wanted to defer JS vote to dec 28…but now plans to hold polls dec 18 due to lack of consensus among parties,” it reads. i curse under my breath – is this the beginning of the end?

the drive turns out to be refreshingly but strangely quick. i’m home in 15 minutes, whereas the regular commute takes up to an hour and a half every day. sprint up the stairs to catch the news – have talks broken down? is this the political armageddon that we’ve been waiting for?

nothing but replays of the five adviser’s press briefing, and scenes of bnp leaders trooping in to their office. special correspondents eagerly wait outside every politician’s house, but no one has anything to report. stay tuned, they tell me, we’ll be back live with bnp’s press briefing.

the clock winds down. 48 hours expire, but more time is needed. the leaders are deeply embroiled in conversation, we learn, so stay tuned. meanwhile, nothing to do but watch tv commercial after commercial, while the same old news scrolls across the bottom of the screen, billed as “breaking news”.

at last, some of the leaders decide to brief the press. this is it, the moment of judgement for the future of the nation. khaleda’s nowhere to be found, surprisingly. perhaps her speechwriter couldn’t produce another classic government-bashing, blame-shifting, military-praising masterpiece? no matter. everyone’s favorite delwar has his own written page that he reads from.

no mention of 48 hour deadlines. no mention of boycotting or participating. no mention of anything but empty angry rhetoric aimed at the caretaker government. 

and, just as quickly as it had emerged, the anxiety and intrigue disappear. no, that’s wrong. the anxiety and intrigue don’t dissipate, but are replaced with anger and frustration. no concrete decision, no announcement, no sense of finality. only the usual talk about “the people” accompanied by a nearly desperate final plea for the withdrawal of the state of emergency. have the seven four demands now boiled down to only one?

miraculously, delwar agrees to answer the press’s questions. the first question is, traditionally, extremely wrong, and gives him space to vent his generic frustrations with the government. meanwhile, muzahid smiles creepily at delwar’s side. why doesn’t anyone ask him about the party’s decision?

finally some intrepid reporter talks about the people waiting at home for a decision. delwar evades the question, as usual, saying something irrelevant about free and fair elections. it takes several more tries for a journalist to ask the actual question: will bnp participate in the elections on december 18?

but delwar’s just too clever to answer that question that easily. we’ll meet the alliance tomorrow, he says, and then decide. 

and just like that, the press conference is over. no clear answer, no definitive direction, just another attempt by the bnp to buy some more time to prepare for the election. what did they do for  the past 48 hours, i wonder. couldn’t they have drawn up their contingency plans in all that time? why did they have to wait past the deadline to decide what to do?

the mobile flickers again. “nomination deadline extended by 3 days,” says the breaking news alert. great. that just gives them 3 more days to waver and flitter about. another three days of uncertainty for everyone trying to figure out if this country will just disintegrate into anarchy come 2009. 

i’m not a political person, as anyone who reads this blog will know. but this cat-and-mouse game of demands and deadlines is not helping bnp’s cause. they can’t seem to set a deadline and make it stick, and can’t seem to arrive at a decision one way or the other. meanwhile, the fantastic five run around town on the daunting cantonment-dhanmondi-cantonment commute to figure out a way to placate everyone, and extend deadlines to allow these people to continue to waver incessantly without a concrete decision.

if i was bnp, i would be highly concerned about my public image. spouting conspiracy theories off the top of one’s head as a means of buying time doesn’t necessarily translate into votes, at least not mine. and that’s all it boils down to – fear that poor preparations will lead to a crushing electoral defeat. but why are their preparations poor anyway? did a certain former “technocrat” actually end up doing any work, or was he too busy with his personal hobby, filing false cases against random influential people?

am i going to vote in this election? no. i plan to be somewhere else, on a much deserved vacation. but i, and all of the jonogon that delwar and all the others keep talking about incessantly need to know – nay, deserve to know – if bnp will participate in the election. if bnp will not participate, fine, but we the people have the right to know. if they will, even better.

it bothers me greatly that not even the top leadership of bnp seem to know what the ultimate decision is. there seems to be enough people regaling khaleda’s ears with monologues on the benefits of either course of action, but all that that seems to be doing is making it her own personal decision at the end. and that is truly scary. it all boils down to two choices for her: lose the election and lose face and power for the next five years, or boycott the election and destroy the country in the process. and, given her previous choices in life, and her potential need for revenge against the hasina of ’96, i fear she may choose to do the latter. 

in my opinion, the only reason that decision was not made tonight may be that she doesn’t trust her alliance partners quite as much as she claims to. i think she fears that, if bnp boycott the election, jamaat may go ahead and participate anyway, thus dissolving their alliance. jamaat could easily do that – look at 1996. given the recent dissension within her ranks, this would in effect her party to just another footnote in the country’s history. quite hard to launch an anti-government campaign against the government when even your partners are in bed with the other side, isn’t it?

in the end i think it boils down to a case of rats deserting a sinking ship. bnp will continue to hold out till the 23rd without giving a firm answer, and then eventually give in, once jamaat’s threats to participate no matter what finally sink in. then the other alliance will gleefully make use of their indecisiveness and crush them in the elections. alternatively, she might stand firm and refuse to participate. in which case i have a feeling that my fellow jonogon will wholeheartedly ignore her boycott calls and cast their vote for anybody they wish, making voter turnout remarkably high. that’ll ruin any claims of power and support that bnp can muster next year, and ensure that she is unable to destablize the country much. i think it’s just punishment for the charade that they have led us through over the past week.

one way or the other, the bnp now have to figure out a way to lose the election while still saving face. otherwise, given their recent performance, they may very quickly become just another minority party in a jamaat grand aliance.

vote virgins

with the support of a sponsor who wishes to remain anonymous, i conducted a small survey of first-time voters in bangladesh. the survey was conducted from november 4 to november 23, 2006, and covered a sample of 10,000 young first-time voters. the sample consisted primarily of students of various private universities, with some public university and college students thrown in. the survey was based in dhaka. respondents were asked whether or not they would be voting or not. i’m not delving into the methodology here, because it will take up too much space. however, enough safeguards were used to ensure the validity of the data collected.

at the outset, before i delve into the results, i would like to offer up some caveats. first, the sample is in no way large enough to be comprehensive and is not designed to present a picture of the majority of first-time voters. instead, it is confined specifically to those who are undertaking tertiary education, in order to provide a picture of what this group of citizens actually thinks.

secondly, as this research was conducted early in the tenure of the current caretaker government, the events we have seen occuring over the past few months has little to no effect on the mindsets of the respondents. although respondents were only asked if they would be voting this time around, we have tried to limit the impact of the current political climate on the responses. for this reason, the team actively avoided surveying those who were already politically aligned within their institutions.

finally, this survey was not designed to find out who the respondents would vote for: complexities including privacy issues and additional administrative permission from the institutions clearly meant that such an analysis would be impossible given the timeframe and budget of the research.

what, then, should this data be used for? the reason i was interested in conducting this research was because of a hypothesis that i had developed – that educated people were growing increasingly sick and tired of the political stalemate arising out of choosing between two major parties and the current political climate, and that they were becoming apathetic regarding their right to vote. while the results do not overwhelmingly support this hypothesis, they should be indicative of a general trend in this direction.

now, on to the much-awaited results!

  • of those surveyed, 57.42% indicated that they would not be voting in the upcoming elections. the remaining 42.58% indicated that they would be voting in the upcoming elections.
  • 27.93% of those who said they would not be voting aren’t or aren’t sure whether they are registered as voters.
  • 12.76% of those who said they would be voting aren’t or aren’t sure whether they are registered as voters.
  • 12.82% of the first-time voters who are not certain if they are registered as voters said that, if they were, they would vote in the coming election. the remaining sample said that they would not vote even if they were registered.
  • only 2.57% of those who said they would not vote said that the implementation of awami league’s electoral reform demands would make them willing to vote.

i realize these results aren’t groundbreaking, but should be a good indication of the way people are beginning to think in bangladesh today.

sorry for the delay in posting these, but it was quite difficult to get permission from the sponsor to post these until they had checked the data for themselves.

also, for the regular readers, sorry for the huge gap in postings. i’ve been spending most of my free time with a family member who was very sick, but that ended with his death last night. regular transmission should resume shortly.