Household spending in Spain 2008-2014

Javier takes a look at the Spanish National Statistics Institute's household budget survey in the post-bubble era.

Basing to the year 2008, the most obvious observation is the degree to which household consumption fell off from 2008 to 2014. Ignoring the very odd results for the highest income group (and perhaps attributing them to sampling error, or something), there's a fairly direct relationship between earnings and change in spending. The group making 3000 to 5000 euros per month saw only a minimal decrease, whereas those making under 500 euros managed to lower their expenses by a full 20 per cent.

The next chart shows the percent change between 2008 and 2014 of the total household budget represented by various outlay items. Fixed (more or less) costs like home rental and auto fuel went up for all groups but, as is to be expected, the lower the income the higher the effect. Interesting is the trajectory of spending on food for the two lowest income groups. It would be easy to conclude that paying the rent and keeping the car running were both accomplished at the expense of the nutrition budget - and yes, tobacco stands out and may be a result of ever increasing taxes proving themselves to be regressive.

Evidently, car purchases suffered at all levels as did spending on restaurants.

Looking at the highest income group from the same perspective, we see a different mix. Auto expenses coincide on the positive side for the two but house rental does not. And the wealthy evidently attempted to maintain their lifestyle by spending more of their budget on holidays, for example. On the negative side, we see the same pattern of the reduction of the weight of basic variable costs of food and medicine relative to the average.

Comparing 2014 to 2008 changes in relative expenditures for the very lowest household income group versus the national average produces the next chart. Again, the necessities of life predominate on the plus side - notwithstanding the appearance of 'leisure and culture' or 'telephone'. The skimping, for its part, took place where it could - less and lower priced foods and medicines, etcetera.

But hidden in the data for the lowest income levels is a huge change in the composition of the group. The chart on the left shows the 2008-14 percentage change in expenditures by income. Total spending of the very lowest income level increased by over 70 per cent over the period - a snapshot of the widespread downward economic migration result of the recession - while average spending went down only marginally.

The combination of this with a 12 per cent drop in spending per 'unit' (e.g. household member) would be a result of larger family sizes at what were higher income levels prior to the crash. This corresponds to the large increase in the rental share of spending. The nouveau poor live, by necessity, in larger homes.

Javier García Echegaray and Charles Butler

More than a drop in the bucket, but still...

José takes a quick look at the decimation of household incomes over the course of the Great Recession.

Comparing Spanish household income to the 4th quarter of 2008 - when employment earnings marked their all-time high of 140 billion euros - the chart on the left of the principal components of the 26 billion euro net income drop (approximately 10 per cent of GDP), shows families getting hit from almost all sides over the course of the downturn.

The 14 billion euro decline in salaries and wages is testimony to the huge increase in unemployment in Spain. Interesting is the average wage per each of the 2.2 million lost jobs of 6,480 euros. The crisis demolished the low end of the pay scale.

Other victims include income from property rentals and interest or dividends. The first is obviously a result of over-supply stemming from the building boom and subsequent crash. But the case of investment income (and other current transfers, to some extent) is more than a little ironic, lower returns being to a great degree product of monetary measures to alleviate the effects of the recession. It is easy to conclude that older workers and the retired are the most affected.

Increases in social benefits summing to 9 billion euros have only managed to claw back about 25 per cent of the gross income loss.

Broken down, the major contributors to these benefits show non-contributory pensions providing about half of the total increase. Much of this effect derives from a recent administrative decision. The 400-odd euro monthly subsidy for the long term unemployed over the age of fifty has been reassigned from the unemployment to the pension column - a tacit admission that many of these people will not manage to re-enter the labour force. On the plus side, this payment may serve as a bit of an encouragement to not go back to work. A hidden youth employment subsidy, in effect.

The increase in old age pensions mainly reflects Spanish demographics although a certain amount of corporate downsizing was effected through early retirement.


Charles Butler and José Domingo Rosello

Small mercies

Javier and I have put together four charts to illustrate some of the weaknesses that accrue to an economy that has trouble generating large firms.

Chart 1 shows the ratio of the percentage of total gross value added produced by firms of different sizes to the percentage of total firms belonging to each category. Clumsy and maybe counterintuitive because the important determinant is the denominator, the message is clear. The countries on the left, including Spain, have lots of small and inefficient enterprises and a few large, productive businesses. The remaining graphics show how this kind of organization evolves in the real world of a long-lasting recession.

Large and medium-sized firms were able to defend their profitability far better than small and micro companies.

Looked at from the point of view of revenues, the legions of Spanish companes employing fewer than ten persons, took the brunt of the damage - although this very remarkable difference may point to other, non-measurable, deficiencies of this type of structure. Not least of these would be their greater agility when it comes to defrauding the tax man, the social security system and the gatherers of statistics. Simply put, the marginal benefit to the government of chasing down cheaters of this size is scant and the resultant data may understate their resilience.

In terms of ability to collect debts, firms with more than 250 employees held a clear advantage over the competition and microfirms did the worst. The results for the other two groups clearly do not conform with the expected.


Charles Butler and Javier García Echegaray

Petroleum and Spanish CPI

Javier took a look at the historical relationship between Brent petroleum prices in euros and Spanish CPI - and came up with a simple model.

The base case assumes the average euro price for Brent crude for the first ten days of  March - 54.10€ - holds throughout 2015. The dashed red line in the chart on the left shows the the model's expected contribution of the price of crude to consumer price inflation for this scenario. The solid red line shows the same real data to date.

With the euro/dollar exchange rate held constant at early March levels, here are the model's projections for Spanish end-of-year CPI based on ten dollar per barrel changes in the price of Brent petroleum.

Average inflation rates for 2015 would look something like this:

36 €	-1.1%
45 €	-1.2%
Base	-0.9%
63 €	-0.8%
72 €	-0.6%
82 €	-0.5%
91 €	-0.4%

Javier García Echegaray and Charles Butler

Spanish IP - more than meets the eye

The lackluster performance of Spain's industrial production index over the last few months has been in stark contrast to the notable improvement in the country's economy. We don't buy the INE's IP data... and here's why.

Here you can see the notable difference between the manufacturing PMI and industrial production. Overlaid is the annual per cent change in electricity demand from industry. The sector is reporting strong activity in the PMI survey and electricity use is confirming it. The IP seems to be referring to a different country.

Again, revenues from large industrial firms (as reported to the country's tax collector, the AEAT) do not confirm the published IP.

The auto industry provides an example of expected correlations between same-sector production, electricity use and employment.

The same data for the machinery and equipment sub-group, for their part, show that the lights are on and there are people in the building - but they don't seem to be doing much. At this economic juncture, that's an unlikely combination.

As to why this is taking place, the most obvious candidates would be inappropriateness of the sample data the INE bases its estimates on and/or problems with the methodology they use to extrapolate industry-wide statistics from those survey results.

A friend of the blog will, in fact, be meeting with some people fairly high up in the INE organization next week. He has promised to bring the matter up. Stay tuned.






Javier García Echegaray and Charles Butler

Inside Spanish employment stats

We look at Spain's February jobs report.

Led by services employment, all sectors except agriculture showed employment growth. The olive crop, the main determinant of winter farm jobs, was very small owing to drought conditions over the growing season.

Service jobs comprise 80 per cent of new employment versus the sector's 68 per cent weight in the total economy. The obvious doubts about productivity should be weighed against the likelihood that many of these new posts reflect surfacing from the large underground economy.

Construction employment is still about two-and-a-half million off the peak - a peak it will never see again. And industry continues to disappoint, if not actually get worse.

Extrapolating (with the usual caveats about that practice) from recent data, services employment will surpass pre-crisis levels in August of this year.

In fact, as we suggested above, productivity seems to be decreasing. But employment figures suggest a repeat of Q4's 0.7 percent GDP growth.

Urged on by the government incentive of drastically reduced contributions, self-employed social security affiliations were the first to go positive back in mid-2013. However, since the beginning of 2014, the growth rate of the number of new autónomos has stagnated or decreased. Salaried employment gains, on the other hand, appear to be maintaining an uptrend.

The percentage share of part-time work over total employment appears to have stabilized or is decreasing slightly.

The percentage share of temporary contracts is increasing. This, given the incentives in Spanish labour legislation, is to be expected in an economic upturn from a low employment base.




Javier García Echegaray and Charles Butler.

The secret life of the Spanish economy

Javier's excellent Spanish-language coverage of the latest GDP figures provides the charts for this entry. Commentary all mine, though.

The big percentage mover behind Spain's Q4 GDP gains was gross fixed capital formation, led by transportation equipment and an improved construction sector.

The government sector (as well as being the only negative contributor) is principally responsible for Spain's negative GDP deflator. The fixed capital formation component (up 5.7 per cent year-on-year in the latest GDP release), shows positive prices for the first time in exactly four years.

Household spending came in at 3.1 per cent over a year earlier. The chart shows spending running well ahead of total wages, raising doubts as to its sustainability. Possibly a symptom of the resumption of recession-delayed purchases, much will depend on the continued improvement of the labour market.

Total goods and services exports improved 3.6 per cent over Q4 2013. Merchandise export order backlogs indicate that strength will continue in the near term.

Lastly, increased employment in the midst of economic growth seems to be taking its toll on Spanish productivity. A good argument can be made that this reflects, as did the enormous efficiency gains over the downturn, the effect of changes in the country's large grey employment market. You can read my very brief description of how this works in a June 2011 post.



Charles Butler, with charts from Javier García Echegaray.