Fix inconsistency with DateOffset near DST

Author: arthurlw

pandas/_libs/tslibs/offsets.pyx

@@ -283,8 +283,10 @@ _relativedelta_kwds = {"years", "months", "weeks", "days", "year", "month",
 
 cdef _determine_offset(kwds):
     if not kwds:
+        from dateutil.relativedelta import relativedelta
+
         # GH 45643/45890: (historically) defaults to 1 day
-        return timedelta(days=1), False
+        return relativedelta(days=1), True
 
     if "millisecond" in kwds:
         raise NotImplementedError(

pandas/tests/tseries/offsets/test_offsets.py

@@ -1095,9 +1095,11 @@ def test_dateoffset_misc():
 
 @pytest.mark.parametrize("n", [-1, 1, 3])
 def test_construct_int_arg_no_kwargs_assumed_days(n):
+    from dateutil.relativedelta import relativedelta
+
     # GH 45890, 45643
     offset = DateOffset(n)
-    assert offset._offset == timedelta(1)
+    assert offset._offset == relativedelta(days=1)
     result = Timestamp(2022, 1, 2) + offset
     expected = Timestamp(2022, 1, 2 + n)
     assert result == expected
@@ -1227,3 +1229,11 @@ def test_is_yqm_start_end():
 def test_multiply_dateoffset_typeerror(left, right):
     with pytest.raises(TypeError, match="Cannot multiply"):
         left * right
+
+
+def test_dateoffset_days_vs_n_near_dst_transition():
+    ts = Timestamp("2022-10-30", tz="Europe/Brussels")
+
+    offset_days = ts + offsets.DateOffset(days=1)
+    offset_n = ts + offsets.DateOffset(1)
+    assert offset_days == offset_n