Posts Tagged ‘Housing Starts,Margin of Error’

Housing Starts Rise In November, But With A High Margin Of Error

December 22, 2010

Housing Starts Dec 2008-October 2010The number of single-family Housing Starts increased in November, adding 30,000 units as compared to October.

The Census Bureau defines a “housing start” as a home on which construction has started.

November’s starts represents a 7 percent increase from the month prior. However, if you see the Housing Starts story online or in the papers, you’ll notice that the press is calling the market gain at 4 percent.

So which result is right? The answer is both.

The government’s monthly Housing Starts data is published as a composite report; lumping activity among 3 separate housing types into a single, group reading.

The 3 housing types are:

  1. Single-family homes (i.e. 1-unit)
  2. Multi-unit homes (i.e. 2-4 units)
  3. Apartments (5 units or more)

The group reading is a fair description of the market and it’s easy-to-understand. As a result, it’s what the press tends to report. However, for home buyers in Utah , it’s the single-family category that’s most relevant.

 

The reason why single-family homes accounted for 84% of November’s Housing Starts is because that’s the type of home that most buyers buy. Few purchase 2-4 unit properties, and even fewer buy entire apartment complexes.

That said, it’s possible that November’s Housing Starts data is wrong. Within the press release, the government placed an asterisk next to the data, indicating that the figure’s margin of error exceeds its actual measurement.

Against a 7 percent gain, the reported margin of error is 13.5%. There is no statistical evidence, therefore, to prove the actual change was different from zero.

If Housing Starts did fall in November, it will help to reduce the housing inventory, which will, in turn, help keep home prices high. For home sellers, this could mean good news. Fewer homes for sale increase competition among buyers.

Housing Starts Rise In August, But By Less Than The Headlines Report

September 23, 2010

Housing starts September 2008 - August 2010The number of single-family Housing Starts rebounded in August, climbing 4 percent from July’s 14-month low.

A “Housing Start” is defined as a home on which construction has started and the August increase represents 18,000 single-family units nationwide.

If you only read the headlines, however, you would think the data was stronger. This is because the Housing Starts data is actually a composite of 3 types of homes — single-family, multi-family, and apartments — but  the press tends to lump them all three together.

As a sampling, here are a some headlines on the story:

  • US Stock Futures Rise After Housing Starts Surge (WSJ)
  • Housing Starts At 4-Month High, Hint At Stability (Fox)
  • Housing Starts Jump 10.5% In August (Marketwatch)

Now, it’s not that the news is wrong, per se, it’s just not necessarily relevant.  Few home buyers  are buying multi-family homes or entire apartment complexes. Most buy single-family and, for the first time since April, single-family starts are on the rise — just not by as much as you’d believe from the papers.

Even still, we can’t be entirely sure that the August Housing Starts data is accurate anyway.

A footnote in the Department of Commerce report shows that, although single-family starts are said to have increased 4 percent, the data’s margin of error exceeds its actual measurement, meaning the data has “zero confidence”.

In other words, starts may have dropped in August, but it’s something we won’t know for sure until revisions are made later this year.