The second article lets various people give their reaction to the first one. The non professionals say what they usually say. But the Berkeley professor of economics answers my question very clearly:
Economic research on this topic is unanimous. There is no question that on net, adding more units tends to lower rents. All existing peer-reviewed academic studies — including work done at Harvard University, the Wharton School at the University of Pennsylvania and by me at UC Berkeley — find that more housing supply results in lower rents and house prices, everything else being constant.
Filtering does work, but as you point out, it takes time: https://oregoneconomicanalysis.com/2016/05/25/housing-does-f...
But just because the best time to build more was 30 years ago doesn't mean now isn't a good time to start.
Also: if you are adding expensive housing it takes some pressure off of the cheaper housing that would otherwise be the 2nd choice of those going into the expensive housing.
https://www.washingtonpost.com/news/wonk/wp/2016/02/12/the-p...
"None of this dismisses the fact that displacement from specific homes happens when low-income housing is literally knocked down to build high-end towers. A good amount of new supply in cities, though, can rise on under-utilized land (former industrial plots, surface parking lots, abandoned properties, etc.). And the cumulative effect of all that new supply can hold down rents across neighborhoods and cities, including for the poor."
Nobody thinks that new housing is bad - in the long term. Where economists start to disagree is how to prioritize low-income housing and market-rate construction in the near-term:
https://www.washingtonpost.com/news/wonk/wp/2016/02/19/how-t...
"To be sure, more supply is needed, but unless it is targeted to those who need it most, it will only help wealthier residents."
"'Filtering,' where older housing units trickle down to lower-income families as they age, can happen in the broader metropolitan context. But it can take decades for filtering do deliver truly affordable units to lower-income households. As apartments age, the rent of a typical unit – not in a hot area - declines an average of 0.31 percent per year so even after 30 years, the rent will have fallen by only 9 percent."
"In gentrifying neighborhoods, filtering does not work at all, because land values and rents rise as the neighborhoods become more desirable and developers bid up land values. So lower-income households must look in other neighborhoods where services and schools are likely to be much weaker. Hence the gentrification process can reconstruct economic segregation."