Yesterday, i stumble upon some of my country macroeconomic data, and i found some interesting pattern on quarter-on-quarter gdp growth, almost every Q4 and Q1 has a negative growth. After checking another country's data, i can't find the same pattern. later i learned that my country data was probably non-seasonally adjusted data, since when i search for non-seasonally adjusted(NSA) GDP for another country, they also show similiar pattern.

My question is :

Why did NSA GDP usually drop on Q4 And Q1, what causes this pattern?

Did this also mean we're usually in recession in Q2 (Since Q4 and Q1 growth are negative)? After the Q2 2020 macroeconomical data released, there are a lot news about our country entering technical recession, even if our Q1 yoy growth was still positive.

Are seasonally adjusted GDP growth are the same as Year-on-year GDP growth?

What data do we use to define recession? Seasonally adjusted one or Non-Seasonally adjusted one?


Output usually falls during certain quarters/months and for many countries it is in the winter. Tourists usually travel for vacations in summer when kids also get long summer holiday. Farm work is highly seasonal and there is more of it in summer than winter (as pointed out by @dismalscience in an excellent comment construction is seasonal too) etc. As a result GDP will fluctuate seasonally and this is in fact the reason why we seasonally adjust series for many research purposes. Although usually these fluctuations are not big enough to cause recessions.

Depending on geography and in which industries country specializes it will be affected differently by seasonal variation in output. There is no a priori reason why it should happen in Q1 and Q4 but for most countries it will be either during winter or summer. Since you do not mention which country you talk about there is not much what can be said about that.

If real GDP dropped in two consecutive quarters then usually it would qualify as a recession (for example that is definition given by Blanchard et al. Macroeconomics in glossary on pp 569), but that is not the only definition used in literature - there simply is no single agreed upon definition by all. Research on calling recessions often uses seasonally adjusted real GDP in order to identify recessions (e.g. Hamilton, 2011), but most definitions do not specify whether it should be seasonally adjusted or not.

  • $\begingroup$ The NBER doesn’t use the 2-quarter rule. On the page you linked: “Why doesn't the committee accept the two-quarter definition?” $\endgroup$ – Brian Romanchuk Oct 12 '20 at 17:49
  • $\begingroup$ @BrianRomanchuk sorry, my bad, I read an old source that linked there, and said they used that as a technical rule, but not anymore but I did not realized that and just used link to primary source without double checking. I already corrected the text $\endgroup$ – 1muflon1 Oct 12 '20 at 18:26
  • 2
    $\begingroup$ In addition to agriculture, construction is also highly seasonal (and in most developed economies it’s a much larger share than agriculture). $\endgroup$ – dismalscience Oct 12 '20 at 19:44
  • $\begingroup$ @dismalscience why construction is highly seasonal? BTW, i live in indonesia, which is a tropical country. Q4 and Q1 are in rainy season, at Q4 agriculture output dropped ~20%, which account for ~2.5% drop in GDP, other sector also stagnating at Q4(household consumption growth is only 0.1%, low demand for goods maybe explain the stagnating growth for other sector).the weird one is the decline at Q1, agriculture has a positive growth ~10%, but government spending dropped ~40% (Healthcare, education, construction, dropped probably because this). Why did government spending dropped a lot on Q1? $\endgroup$ – Fahrul Oct 13 '20 at 1:30
  • 2
    $\begingroup$ @Fahrul In colder climates, concrete becomes uneconomical to work with when the temperature falls below about 5°C. Based on what I know (which is limited), rainy seasons slow construction activity as well (because some things are more expensive to do when it’s raining). I pulled up the construction GDP series from Badan Pusat Statistik Indonesia, and it appears to peak quarterly in Q4 and dip significantly in Q1; that seems to line up imperfectly with what you observe. $\endgroup$ – dismalscience Oct 13 '20 at 15:33

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.