if the question is efficiency  (see your choice of the word "effective"), authoritarian restricitions of any kind could meet the target in various ways. If instead we want to take a broader view and want to take into consideration costs and benefits to the users (and non users) the most desirable measures, to my view, are directed upstream: I mean on what generates congestion. Which means broadly trip generation and mode choice.

to be concise:

1  - urban planning that minimize the km of trips (this includes density as a transport planning tool)  and favours public transport

2 - fair pricing of private vehicles (this may include congestion charging)

3 - good supply of non private transportations (Public transport infrastructure and operations)

I would generally suggest that the above ranking goes corresponds to a decreasing order of net socio economic benefits. So policy 1 should could often be more desirable than policy 2, and so on... THis obviously depends as well oon the specific context to which your analysis applies.

These are heavily debated questions... other points of view welcome.

Fully agree with Jerome,

It fits with the so-called ASI approach, based on Avoid trips, Shift modes, and Improve technologies.

First of all, I'd try not to expand road infrastructure, since it is known to encourage private transport trips, worsening congestion. 

Then, I'd apply the ASI approach. «Avoid» implies to rethink if current transport demand is needed. New technologies may help to reduce unnecessary mobility, such as tele-working, on-line shopping, etc., whereas an optimisation of closeness between inhabitants and production and leisure could also be adopted.

«Shift» implies a change from unsustainable transport to more environmental-friendly transport systems, such as walking and cycling, or public transport. First of all, cycling and bus infrastructures would need expansion. Secondly, tax distortions and unjustified subsidies should be eliminated, whereas urban road pricing schemes could be applied in order to encourage more sustainable transport choices. Fossil fuel taxation (based on the «polluter-pays principle»), low transit fares and land-use policies limiting urban sprawl could also be considered.

Even though the answer is too general I hope it helped.

Good luck

Massiani's answer looks at policy measures. That may certainly be part of the answer, although it should not be underestimated how complex effective congestion charging can be.  The London example is not very attractive: it is a very expensive system and it is has no guarantees against underutilization, which is also a form of economic loss. The problem is not simply preventing congestion, it's about optimal performance under high load: trying to process as much traffic as possible without falling into congested states, which are known to have a pretty bad performance. Even for a straight segment of road, this is not easy at all. But the problem has to be solved in a network context, because congestion is essentially a network phenomenon: it is easy to prevent congestion in one area and thereby causing it elsewhere. If inflow is reduced into a given area, then upstream the danger of congestion increases. So we want to prevent congestion without causing it elsewhere and without causing serious underutilization.

A lot of research in traffic engineering tries to model traffic and then devise control measures on the basis of traffic simulation. But this is not likely to be a feasible approach because traffic in states near congestion is a chaotic system: very small effects can have very large consequences. In dense traffic, one driver that slows down for whatever reason, can cause a chain reaction which leads to congestion and the flow collapse that results from congestion. Trying to model such a process is just as hard as trying to model and predict the occurrence of collisions in traffic. 

A proposal for how to solve congestion in a network, without causing underutilization and without being dependent on congestion predicting models, can be found in Yubin Wang's thesis, chapter 6 (see attached link).

But don't expect an easy solution to a highly complex problem. And I'm the first to admit that implementation of this approach in operational traffic management centers is an even harder problem than finding just the technical solution to congestion.

https://repository.tudelft.nl/islandora/object/uuid%3Ae466fb81-0d17-44f6-8d92-0c9cb5a73a7f?collection=research

Thank you all for the detailed answers. It seems that combinations are needed from policies to strategies. 

“How to solve traffic congestion” and "why is it so congested" are questions usually asked by my friends, when I tell them that I am a researcher focusing on transportation. It is a very basic question but really difficult to answer in a short time. It is really happy to exchange ideas with all of you on the platform. 

@Jérôme Massiani In Beijing, China, we rotatedly restrict 1/5 cars every day, but large-scale congestion still occurs. It is very interesting to study on the proper number of cars being restricted in terms of efficiency, policy and human rights.

@Yeray Hernandez It is good to know the framework/definition of ASI. It seems that our government is trying to apply it. 

@Jos L. M. Vrancken Yes. Although traffic engineers have proposed a lot of models and strategies, I often wonder how many contributions the models/strategies really made for solve traffic congestion, in particular when a solution is proposed for a isolated point instead of the whole network. Also thank you for the recommended chapter.

No congestion, no development.

Drivers' behavior, I think,  plays very important role in traffic congestion. Maybe autonomous vehicle is the best way to solve it. 

Perhaps the attached self-explanatory graph can be indicative of the reason of congestion, at least in Australia, over the last 113 years. The graph is from Long-term trends in urban public transport BITRE and its link is below. 

https://bitre.gov.au/publications/2014/files/is_060.pdf

Consistent and parallel action in all three aspects that Jérôme mentioned. So simple in theory and so difficult in application.

So, basically, concluding the previous comments: to reduce the number of vehicles. The solutions are different from improvement of public transport and promotion of bikes to fuel pricing and car ownership taxation.

Other measures, such as widening roads, reconfiguration of problematic intersections, development of network, while being beneficial by themselves will only promote more cars to appear, which sooner or later leads to new congestions.

It may be a bit simplistic but I found through the travel behavior projects I undertook in the past that it comes down to the following key elements to a trip - Its Cost, Convenience / Access and Travel Time. Managing these elements simultaneously to reduce traffic congestion is important. How each factor is applied will vary depending on the urban form and location e.g inner city areas will be different to outer suburban and rural areas.

As mentioned in the previous comments cost factors can include road pricing, fuel costs, car prices,paid parking etc. Access and convenience can be affected by the availability and remote location of car parking, making the local road network less permeable to cars but more permeable to walking, cycling and public transport, proximity to employment and services that encourages active transport use  etc and travel time can be affected by factors such as intersection controls, priority given to pedestrians crossing, less permeable local road networks etc

I don't believe there is a single generic silver bullet that can be applied universally to address traffic congestion in all areas. Rather its a coordinated combination of measures that needs to be addressed on a case by case basis at both regional and local levels.

But what ever measures are applied it is vitally important that alternative transport infrastructure and services are in place and work effectively. Without it it  just creates greater frustration and lost opportunity to change travel choice.

Traffic congestion is no easy fix and though not an ideal outcome, if no coordinated action is taken, its likely that congestion  will reach a threshold / trigger point where people are likely to actively seek and demand other transport options anyway.

1. Encourage use of public transportation

2. Do not allow single occupancy car. Full occupied cars only during peak hour.

3. Implementation of staggered time of work 

4. Use of company provided staff bus. 

Transit oriented development is a good idea to reduce congestion.

Integration between land use and public transport system minimizes the need for vehicle movements.

Efficient transit is the answer; make public transportation experience more attractive than driving in congestion.

There are some good examples - Rotterdam, they have a problem rather with a bicycle than car congestion. It is an effect of holistic transport policy, great bike infrastructure, parking policy. Further, plans are to introduce zone "0" where only electric vehicles are welcome. If the policy is coherent and consequent there are the effects.

In my opinion, the most effective way of solving the current and future traffic-related problems, including traffic congestion, is to use new technologies to reshape our transportation systems.

When you consider that traffic jams were common in ancient Rome, road congestion seems to be a property of a healthy transportation system. That much said, many of the others posting here have correctly, in my view, postulated that congestion management is a function of a range of actions, from policy level down to engineering and technology.

Some cities define the congestion problem not in terms of eliminating it, but in terms of relocating it. Boston's famed $14 billion "Big Dig" project was about shifting congestion from the center of Boston to its periphery (this according to the Father of the Big Dig, Fred Salvucci, a professor of mine at MIT way back when), for example.

Many here have written of the role of public transit (or, public transport, depending on your country/region) in reducing congestion. My understanding, as someone who has worked as a planner for over two decades, and who has conducted primary research on the issue, is that transit is only effective at reducing pressure on roadways when it is competitive with driving in the eyes of potential users. From research in multiple cities, we know that the general public is made up of a number of distinct "market segments," or clusters of attitudes. For example, research in San Diego found six such market segments, each of which had a different set of values when it came to trip-making. For some, time was a critical variable; for others, convenience. Either way, since transit is unlike most other social services in that it competes for users in a relatively free marketplace, it's either competitive in a potential rider's view, or it isn't. Too many times, cities pour good money into creating transit infrastructure and services that from day one is not competitive with other options in the eyes of the vast majority of the market.

My research established three *planning* variables that drove the mode choice decision (and one "*political* variable, namely price): network structure, system performance, and customer experience. These are described in a bit more detail in a series of presentations I have posted on slideshare.net under the heading, "A Market-Focused Paradism for Public Transit," in six parts. Network Structure is just that: a person says, "I'm here, I'm going there, where do I need to go to enter the transit system and where does it leave me off?" System Performance is about the role of time in the system, including both traveling and waiting. Customer Experience covers a wide range of elements, but ultimately comes down to this: when a person completes their trip, do they feel like they made the right decision to use transit, or do they feel it was a mistake?

Ultimately, a transit system that connects more people to more destinations, does so quickly, and with an experience that reduces perceived risk throughout the journey is one likely to be perceived as "high value" by the targeted markets, and it is that value that drives not just the decision to use transit instead of drive, but locational decisions as well: where people look for a place to live, where to shop, and possibly even what jobs to pursue. If transit can accomplish this, it seems to attract considerable real estate investment, increasing density levels, which in turn leads to not just more transit use, but especially more walking and possibly bicycling (if separated bicycle facilities are available, such as in most Dutch cities and Copenhagen). This is different than just saying "TOD" (transit-oriented development); if the transit is perceived as too slow, too out-of-direction, not convenient, and not comfortable at the psychological level, then you get higher densities but not necessarily a shift in trip-making from the automobile (this was the situation that Los Angeles was in, at least a decade ago; I do not know if their recent transit construction binge has helped remedy this).

I agree with my colleagues on the implementation of transportation system management tools to reduce traffic congestion. Also the use of ITS concepts in reducing headway between moving vehicles should increase lane capacity and may may prove to be effective. Other innovations may emerge if we start looking at microscopic traffic variables as well.

What is the most effective way of solving traffic congestion?

Agreed there are many methods can be explored as shared by previous researchers / scholars. On the technology improvement perspective, perhaps we can explore the following:

I think that the most efficient way to solve transport congestion is to create enough road capacity (width, quantity). However, this can be done only in smart planned economy. There is a danger of over-investment. Because of asymmetric information, road with too high capacity (let say double of marginally congested) can be planned to get more public money by constructor. Alternatively, incorrect measurement of required capacity can result in under-investment.

Market forces support congestion, because there is minimal investment cost per unit of flow, and thus profits of investor are maximized (both in the case when state pays for the road and when there is a fee per unit of transport).

Introduction of sufficiently high transit fee can indeed reduce congestion, but will not result in an increase of utility of users.

There should be also optimal trade off between public and private transport. If in poor economy this can be regulated by market, while in a rich economy (when everybody can afford a car) there should be individual concern about public utility.

In some rich cities (like Vienna, Austria) not all citizens choose to have a car, because: a) public transport is good and relatively cheap, b) parking place is relatively expensive.

Finally, congestion is typical for large cities. Why do people tend to concentrate there instead of being spread across smaller cities? You can have a look at my new mathematical model with important policy implications: Conference Paper Aggregate and Average Land Rent in Cities

Late to the party! My research considered traffic congestion resolution.

Congestion accompanies commercial success, which is a very fragile balancing act.

Restricting travel is not an acceptable answer, the world has developed due to convenient mobility. Restricting travel will result in commercial catastrophe.

Road charging is not the right answer. it generates a multi-tier system of those who have and those have-not, those who can afford to travel and those who cannot. We could see only the wealthiest people able to travel in much the same way as we watch people with the least fuel efficient vehicles use far more natural resource because they can afford to destroy the environment. Meanwhile we could restrict the travel and convenience of the lower paid working classes that we rely on. There is a similar effect with restricting road use by registration plate number - the wealthier simply buy two cars with odd and even numbers; congestion gets worse because the same number of cars are in use while more cars are parked.

Many public service passenger vehicles operate for long periods with empty seats for routes or sections of routes, they take up more space and use more fuel than cars. PSV work in "joined up areas" of high population and employer densities but are far less successful in smaller cities and practically unused in rural areas.

In the UK we had an idea to reduce congestion by creating "out of town" commercial areas and pedestrianising city centres, it worked - city centre congestion reduced, as did shoppers, municipal income and retail outlet viability. Now we blame On-Line retailers for destroying our city centres and reducing our use of public transport in to those areas.

I do agree with some of the comments (Romanika, Alan, Yuri) but not so much some others.

I concluded that congestion is always going to exist in some form or other while ever you expect commercial success and that you should manage the resources that you have control over (roads and road management), the other parts (humans, vehicles) will manage themselves to a form of benefit / convenience equilibrium.

City pollution is another matter, electric vehicles especially assisted bicycles are my recommendations until we figure out a better way.

Similar to the final comment by Yuri, if congestion is such a problem then stop joining it. Change jobs or employers.

Conference Paper Available Forward Road Capacity Detection Algorithms to Redu...

Developing the public transport system is one of the most effective ways to relieve congestion, especially if travelers are offered better service in terms of (time-cost), comfort and safety.

Also develop the railways network and make optimal planning between the various public transport systems for example (tram-metro-bus-train) to serve all areas, this further encourages the population to use public transport and park their private vehicles.

Finally, to use intelligent systems in an optimal way especially in the control and the regulation of the trafic. Thus to use the ITS to collect the data of the road traffic and the demand of the travelers, according to these data one must make a forecast of the transportation demand in the short and long term.

It will vary across countries. To dig down the challenge, look at the interaction of the state, society and market. Then view congestion as a node in that dynamical system. For non-ergodic spaces: Article Public-Private Partnerships (PPP) on Moulding State Structur...

Optimise traffic-light management

I suggest that using the google map of instant traffic information to let all people know where the traffic jam area to avoid and let traffic police solve the traffic accidents.

Isolate traffic jam event, control the inputs & outputs.

Hello Zhengbing He,

In my opinion, the most effective way of solving / reducing congestion in urban agglomerations is the time shift of the start of operations by the most numerous groups of employees within a given urban agglomeration. Schools, offices and companies usually start and end their activities at similar times, which means that a very large number of people are moving in the same directions, which causes congestion. Congestion is a very complex problem and its solutions require an individual approach for a specific urban agglomeration.

In my opinion, a good city planning is the best way to prevent traffic congestion before it happens since planning is the root cause of this problem.

It’s like this quote: “If the first button in a shirt is put wrong, then every button will be wrong”.

However, if congestion does happen, shifting people from private cars to public transit by promoting transit oriented development, taxing private cars, and improving public transit infrastructures will do (in my opinion).

I stand by my comments in april 2018.

You cannot and should not resolve urban congestion. Congestion is an indicator of commercial prosperity in an area, if the area is properous then many people will want to be a part of it and there will be tidal entry congestion and exit congestion.

Where there have been attempts to resolve congestion by techniques such as e.g. flow reduction or pedestrianisation, these commonly result in reduced attendance due to limited accessibility and reduced convenience. The follow on impact is a spiral of shop closures, reducing commercial prosperity, further business closures and so on.

Urban congestion management is very popular with businesses that can command captive audiences by providing carparking within e.g. a shopping centre, they still create congestion but this is "not their problem".

The other people that benefit from urban congestion management are online outlets. As terrible as it is, Covid-19 has made it possible to observe the results of tremendous global experiments in clearing congestion by reducing social interaction and urban traffic flow. Observe the results of physical business closures and online business expansions, this is what happens when congestion is resolved.

I agree with the comments of James Hardy and I add the following:

There is no consensus on what constitutes a standard, generally acceptable definition of road traffic congestion. Congestion is not a new phenomenon and occurs in all types of environments; congestion is not inclusive of roads networks; it also occurs in rail networks, high rise buildings lifts, supermarket checkouts, and many others situations.

Road traffic congestion occurs when pressure demands for road resources, above the available capacity, are placed upon networks; congestion also occurs due to inefficiencies on the networks (Blunden 1983). Blunden presents two types of traffic congestion: congestion due to pressure of demand, and congestion due to inefficiency. Congestion due to pressure of demand is an ally that must be welcome, as it is a product and a sign of vibrant economic activity. Congestion due to inefficiency is an enemy that must be combated and prevented at all cost. (Blunden, WR 1983, The Tyranny of Transport Congestion - Friend or Foe?, The University of Wollongong).

Hi William, thanks for the comment.

As I sent my previous answer I did think that I had not included the inefficiency congestion type. Sometimes caused by bad initial design, sometimes the result of regional development / evolution rendering a good design unsuitable and sometimes the result of overthinking / attempting to apply technology to control a small part of a chaotic system with a random number of random input variables. These inefficiencies can and should all be alleviated.

As an example of bad design, we have a traffic light system nearby that caused gridlock on the major arterial route even on Sunday afternoons after it was turned on, the solution was to turn the lights off and put bags over the light heads. The area is still busy but not gridlocked. I am not advocating turning all of the traffic lights off, we have other junctions that were improved by traffic lights, but not all.

I am writing few of the solutions which I feel is important.

1. Public transit systems should be so developed that more people will start using it over private vehicles.

2. Mainly for developing countries, lane discipline should be strictly obeyed which even today is not at all followed.

3. All vehicles should be fitted with GPS mandatory which will ease the algorithm of google maps by which better route choice will be provided to the end road user.

Smruti Sourava Mohapatra any more effective measures?

Cycling Culture should be implemented

@Nasim Ullah Khan it depends on the region (country).

Public transport is only effective IF it takes people where they want to go AND when they want to go there. People living in rural areas will always lose in this circumstance, their service use is unlikely to be able to support the service cost which means the service frequency will be reduced or ceased. This in turn makes the service unsuitable / unusable.

There is no such thing as free, the service has to be paid for somewhere (vehicle costs, fuel costs, driver wages). Unless the general public can be convinced to make a continual voluntary contribution, this would be derived from some form of taxation. The tax payers would not see the benefit of paying for a service that does not benefit them and would want greater economic efficiency through reducing unused services and routes such as the rural routes. There is no defining line between urban and rural routes, efficiency will be demanded in each area making suburban areas at risk.

Finally, a public transport service that operates with only a single passenger is likely to be using between 4 and 10 times the amount of (fossil) fuel per passenger mile that would be used by a private car. If the service runs with no passengers for any stage of the journey, the cost per passenger mile is infinite. Very few passengers will be on a journey that begins and ends at a bus depot (not bus station, depot), there will always be stages of infinite cost per mile. Anyone with a concern about environmental damage or global warming would always object to needless energy waste, regardless of how it is derived.