On the blog Calculated Risk, Bill McBride had a reasonable post criticizing my article that I co-authored with Eliot Spitzer.  McBride argued that we made some errors around the scale of the problem.  I wanted to respond, because I like and respect Calculated Risk and want to make sure Bill understands where we’re coming from.

Spitzer and I claimed that “roughly half of homeowners with a mortgage are effectively underwater, which means they owe more on their mortgage than their house is worth.”

Calculated Risk pointed out that the actual number of underwater mortgages with negative equity, according to Zillow.com, is 28.6%.  He then goes on to argue that ” many of these homeowners were only slightly underwater, and will probably keep making their mortgages payments”, in part because of Obama’s revised program to allow refinancing of those who are underwater (known as HARP).

While McBride’s interpretation of the data is reasonable, we think that arguing that a majority of homeowners have negative equity is also reasonable.  We include people who are near negative equity as effectively underwater.  Here’s what we were talking about, from mortgage analyst Mark Hanson.

On US totals, if you figure average house prices use conforming loan balances, then a repeat buyer has to have roughly 10 percent down to buy in addition to the 6 percent Realtor fee to sell. Thus, the effective negative equity target would be 85%. You also have to factor in secondary financing, which most measures leave out.

Based on that, over 50 percent of all mortgaged households in the US are effectively underwater — unable to sell for enough to pay a Realtor and put a down payment on a new purchase without coming out of pocket. Because repeat buyers have always carried the market as the foundation, this is why demand has not come back. It’s as if half the potential buyers in America died over a two-year period of time.

And Laurie Goodman, of Amherst Securities, projects over 10 million foreclosures by 2018, out of 55 million mortgage holders.  She fears there will be more if prices keep declining, which they could considering that the mortgage market itself is wholly backed by the government and serious servicer side problems have not been fixed.  This is why we wrote our piece, because this isn’t inevitable – it’s a deflationary spiral with foreclosures causing housing price drops causing more foreclosures.

The scale of the problem is enormous, and within the large grey area of slightly above water to underwater, there is a huge reason to worry about it getting worse considering that the market is so dysfunctional.  While McBride made a case for what he thinks the scale is, we think there’s a good reason to take our interpretation as more accurate.  We also believe that authorities have been underestimating the size of the problem for years.

We also assume – unlike McBride – that refinancing programs proposed by the government will have the same effect as earlier housing programs proposed by the government – very little.   Hopefully we’re wrong, and it’s possible that a new program to help spur refinancing will work.  But the administration’s track record and that of the administration before this one suggests otherwise.

McBride made one other criticism, that is that 30 million jobs is too ambitious and that our real jobs shortfall is 5 million.  Again, this is an issue of definition.  He uses the standard unemployment rate and argues that “structural unemployment” is 5%.  We looked at the BLS data and added up all the people who are unemployed, underemployed, and no longer attached to the labor force.  That’s roughly 25 million.  We include the prison population, because we don’t think that imprisoned individuals should be counted as having jobs.  That’s another 2.3 million.

We also do not agree that the idea of “structural unemployment” is a useful way to think about the economy, as the unemployment rate has varied over time.  During World War II, the unemployment rate dropped to 1%.  There’s nothing structural about unemployment, it is often policy and business-cycle related.

Anyway, that’s 27 million jobs needed, which is in fact very close to 30 million.  And there’s a work force in the shadow or black market economy that we don’t know a lot about, and yes, the people doing off the books work (like the nannies and gardeners without green cards sometimes employed by politicians) are people too.

Even if you don’t agree with my jobs number, the idea that we just need 5 million jobs to bring us back to full employment is true only in the most narrow technical sense – if you just look at the number of people unemployed for more than 27 weeks, you get more than 5.5 million people.  That’s dropped a bit in the last three months, but that is a crisis.

I do agree with CR that getting the scale of the problem matters.  I suspect, as these measurements show, part of the issue is that the government isn’t necessarily measuring what matters.  Another part of the problem is that definitions matter, and depending on the definition one uses the scale can be much broader than the conventional wisdom would have us believe.

Anyway, thanks for the comments, Bill.  I’ll keep reading Calculated Risk.

READ MORE: Economic Analysis and Inaccurate Numbers (CalculatedRiskBlog.com)