When people say widening roads makes the problem worse, they usually cite as an example 'before construction, traffic was congested for X minutes; and after it is congested for X + Y minutes' the problem is that often demand has increased in the meantime (because of population growth, or increased economic activity).
There is certainly some amount of induced demand, and in the right scenarios, that could make things significantly worse. In the Bay Area, if 880 was significantly widened but 101 and 280 were not, it might make it attractive to drive from San Francisco to San Jose by crossing the bay and taking 880 south. This could make 880 less effective for east bay users.
This doesn't become a significant issue if there's reasonable regional planning, and routes are kept in rough balance.
There's a limited budget for transportation projects though, and they can be complex and time consuming, so often roads aren't widened until there is a significant amount of peak congestion. During construction, things get worse, and afterwards, there's still a period of peak congestion. People who expect a road widening to create free flow conditions at all hours will be sorely disappointed. I'll add that on and off ramp design makes a big difference in traffic flow as well; a full clover leaf is good for interchanges with low traffic, but all the weaving makes for poor results in congestion; redesigning to avoid weaving helps flow tremendously without widening, although more/different space can be needed depending on the specific road alignment.
There is potential to argue about what is the best use of transportation money; I'm not familiar with transportation in the DC area though, so I won't weigh in on that.
There is certainly some amount of induced demand, and in the right scenarios, that could make things significantly worse. In the Bay Area, if 880 was significantly widened but 101 and 280 were not, it might make it attractive to drive from San Francisco to San Jose by crossing the bay and taking 880 south. This could make 880 less effective for east bay users.
This doesn't become a significant issue if there's reasonable regional planning, and routes are kept in rough balance.
There's a limited budget for transportation projects though, and they can be complex and time consuming, so often roads aren't widened until there is a significant amount of peak congestion. During construction, things get worse, and afterwards, there's still a period of peak congestion. People who expect a road widening to create free flow conditions at all hours will be sorely disappointed. I'll add that on and off ramp design makes a big difference in traffic flow as well; a full clover leaf is good for interchanges with low traffic, but all the weaving makes for poor results in congestion; redesigning to avoid weaving helps flow tremendously without widening, although more/different space can be needed depending on the specific road alignment.
There is potential to argue about what is the best use of transportation money; I'm not familiar with transportation in the DC area though, so I won't weigh in on that.